BG2 with Bill Gurley, Brad Gerstner & Aaron Levie | Software Valuations, Earnings, Immigration

00:00:00.000 | I can absolutely see sometime over the next four quarters,

00:00:03.840 | we're going to hit a zone of disillusionment

00:00:06.080 | because everybody's pigpiled in here.

00:00:08.140 | They think it's all happening now.

00:00:09.940 | It's going to take a little bit longer,

00:00:11.440 | just like the internet did, just like cloud did.

00:00:14.020 | (upbeat music)

00:00:16.620 | - Bill, we made it to episode two.

00:00:28.300 | - We did. Aaron, I'm super excited you decided to join us.

00:00:32.340 | I've known you for a while

00:00:34.200 | and I've heard other people talk about you.

00:00:36.540 | And there's two things I just really love.

00:00:38.500 | One, you always have an independent point of view

00:00:41.100 | and you're not afraid to state it,

00:00:42.940 | which is not true for everyone,

00:00:44.780 | especially people in the CEO slot.

00:00:46.920 | But more importantly, people refer to you as an old soul.

00:00:50.480 | Now, I looked that word up.

00:00:52.460 | It's actually, it's not derogatory at all.

00:00:55.280 | And it doesn't even mean you're old.

00:00:57.320 | It means you're wise beyond your years.

00:00:59.660 | Now, I know you're not 40 yet,

00:01:01.660 | so you're two decades short of me.

00:01:04.400 | So I'm not calling you old,

00:01:05.900 | but I do think people look up to you, respect you,

00:01:08.900 | think of you as a mentor in the valley.

00:01:11.400 | And so we'd love to hear what's on your mind

00:01:13.980 | and what you're thinking about.

00:01:15.400 | - I appreciate that.

00:01:16.240 | I think, actually, I have more gray hair

00:01:17.780 | than either of you guys.

00:01:19.280 | So I think I'm going to say--

00:01:22.580 | - That's what happens when you run a software company

00:01:24.980 | through Zerp.

00:01:27.260 | - They did not tell me this when I got started.

00:01:30.160 | So I wish this would have been on the label,

00:01:33.260 | but yeah, no, I'm super excited to be here.

00:01:35.860 | Obviously, a ton of stuff happening

00:01:37.420 | in the industry right now,

00:01:38.380 | and I'm excited to dive into whatever

00:01:40.460 | you guys wanna talk about.

00:01:41.460 | - We have to kick off with software, right?

00:01:44.160 | You're one of the thought leaders in the space,

00:01:46.780 | and certainly we know you well,

00:01:48.100 | and most people in Silicon Valley know you well,

00:01:50.420 | but maybe just a level set,

00:01:52.900 | quick reminder for everybody.

00:01:54.380 | What does Box do?

00:01:55.700 | What inspired you to get it started,

00:01:57.400 | and then how it's evolving as you guys exit Zerp?

00:02:01.260 | - (laughs) Actually, the Zerp thing,

00:02:04.100 | I know there's a lot of Zerp stuff going on,

00:02:08.540 | but we actually never really even had

00:02:10.720 | that much of a Zerp benefit.

00:02:12.540 | I can certainly get into the different eras.

00:02:14.620 | There was about a three-year period

00:02:16.860 | where we raised a ton of capital and we burned it,

00:02:19.780 | and we did have a little bit of a Zerp benefit,

00:02:21.560 | but it was actually, on paper,

00:02:23.540 | exactly the right thing to do at that point.

00:02:25.660 | I got diluted a ton in the process as a result,

00:02:28.020 | but it was actually, strategically,

00:02:29.700 | the most important thing to be doing.

00:02:31.940 | But history of the company, super quickly,

00:02:34.180 | and what we do, so we built a platform

00:02:36.020 | that helps companies manage their most important data,

00:02:38.280 | their financial documents, their marketing materials,

00:02:40.260 | their contracts, their strategy content,

00:02:43.320 | anything that can turn into a piece of content

00:02:45.540 | we manage and store and secure.

00:02:47.500 | We have about 115,000 customers,

00:02:50.060 | 70% of the Fortune 500,

00:02:51.940 | and we started the company with a really simple premise.

00:02:54.020 | This was in 2004, 2005, when we got the idea.

00:02:57.420 | It was right at the start

00:02:58.620 | where you could put data in the cloud.

00:03:00.180 | It wasn't called the cloud then,

00:03:01.380 | and we did crazy bridge rounds and all this stuff.

00:03:03.940 | I think we pitched Bill four times at that point,

00:03:06.820 | handled it, got rejected, probably appropriately,

00:03:11.240 | and, or maybe it was Peter Fenton,

00:03:13.780 | but we can blame it on him,

00:03:15.300 | and then pivoted to the enterprise,

00:03:18.380 | and then just got super lucky,

00:03:19.780 | which was there was this hyper wave of cloud with AWS,

00:03:24.780 | iPhone, and then iPad,

00:03:26.780 | and what happened was everybody's legacy technology

00:03:29.600 | no longer worked in this era of mobility

00:03:32.740 | and cloud computing for managing their content,

00:03:35.140 | and so we were able to kind of ride that wave

00:03:38.140 | where we just had a better, cheaper, faster,

00:03:40.020 | more secure platform than what else was on the market,

00:03:42.380 | and then that was sort of the period

00:03:45.380 | where we just poured the fuel on the fire

00:03:49.380 | on hiring sales reps, hiring engineers,

00:03:51.780 | building out a platform,

00:03:52.780 | and then scaling up to basically where we are today,

00:03:55.580 | so 15 years of history in a couple seconds.

00:03:59.340 | - One thing I'd love to talk about,

00:04:01.820 | especially with you here, Aaron, is software multiples,

00:04:05.580 | and so you've lived through a lot of different

00:04:09.420 | spots on the curve,

00:04:12.740 | and I've always said that Silicon Valley

00:04:15.180 | has the crudest kind of least intelligent view of valuation.

00:04:19.260 | They always rush to price to revenue

00:04:22.380 | because it's easy,

00:04:23.700 | and because, quite frankly, it's easier to be optimistic,

00:04:27.060 | and then the other thing that happens in Silicon Valley,

00:04:30.140 | all founders kind of tie themselves to a single number,

00:04:33.620 | so 10 times revenue or something,

00:04:35.880 | and Brad has some data,

00:04:37.860 | and on the first slide,

00:04:39.900 | I think the average multiple for a SaaS company is now six,

00:04:44.900 | which is actually not that bad.

00:04:47.020 | - Yeah, that's great.

00:04:48.180 | - But even to say the average is misleading

00:04:51.700 | because the dispersion is so great.

00:04:54.160 | - I mean, you guys have done more content

00:04:56.700 | and thought leadership on this

00:04:57.580 | than probably any two people in the world.

00:05:00.340 | Not all software companies are created equal.

00:05:02.980 | Not all technology companies are created equal.

00:05:05.060 | I think we clearly have been through a period,

00:05:08.900 | probably due to just the influx of capital,

00:05:11.460 | maybe this is the Zerp element, Brad,

00:05:12.780 | that you're kind of referring to,

00:05:13.940 | is if you're one click or two clicks

00:05:16.700 | removed from Sand Hill Road,

00:05:18.620 | and Sand Hill's equally to probably blame at certain points,

00:05:21.540 | you're sort of like, oh, tech-enabled everything

00:05:23.660 | should all get a six or 10x revenue multiple

00:05:26.460 | when the underlying economics of this business

00:05:29.020 | could be literally 100x difference.

00:05:31.180 | You could have a five or 10% margin company

00:05:33.340 | versus a 90% margin, 95% margin company,

00:05:37.540 | and, but yet the outside capital inflow

00:05:40.860 | is treating those as exactly the same revenue multiple,

00:05:42.940 | which is obviously diabolically crazy.

00:05:45.380 | Our own journey has been a little bit really simple

00:05:47.940 | because we've always sort of traded

00:05:50.140 | somewhere between four and 8x.

00:05:51.660 | We've never had the crazy high multiple.

00:05:53.260 | We haven't also been totally sort of forgotten about

00:05:57.700 | at any point, and I think that's a result of,

00:06:00.540 | we've always had basically between 70 and 80% gross margin,

00:06:04.040 | somewhere in the 100 to 120% net retention rate.

00:06:08.160 | And so the economics kind of always work out that,

00:06:10.680 | okay, you kind of know the underlying contours

00:06:12.760 | of the business model, but yeah, I mean,

00:06:15.120 | I think we've been for too long treated all software

00:06:18.160 | and all technology indiscriminately as sort of,

00:06:22.480 | like it should be an 80% gross margin company

00:06:25.280 | with a triple-digit net retention rate,

00:06:28.640 | and that's just not the case.

00:06:29.800 | So, I mean, I kind of almost flip it back to you guys

00:06:31.820 | on like, how do you think, do we come back

00:06:35.040 | to forgetting about the differences,

00:06:37.160 | or is this now a much more sort of normal environment

00:06:40.400 | that sustains?

00:06:41.240 | - I think that really is the critical question.

00:06:43.040 | We have a few more charts here.

00:06:44.280 | To your point, if you pull up this next chart,

00:06:46.800 | the blue line represents the multiple of revenue.

00:06:49.960 | Here, we are actually looking at this growth adjusted,

00:06:54.160 | and you can see that we had this major spike

00:06:58.240 | during the period of '20 and '21, right?

00:07:00.680 | So, this is normalizing for growth.

00:07:03.000 | We fell all the way back down to this 10-year average,

00:07:05.560 | and now we're above it again on a growth-adjusted basis.

00:07:09.940 | If you go to the next slide, we said, okay,

00:07:12.700 | let's look at it from a free cashflow

00:07:14.400 | multiple perspective, Aaron, and what you've seen here

00:07:18.800 | is that a lot of companies like Box have said

00:07:22.520 | during this period of '20, '21, whoa,

00:07:24.960 | we gotta focus on profitability, not so much on growth,

00:07:28.080 | because we may not raise that next round of financing.

00:07:30.500 | So, everybody got a lot more profitable.

00:07:32.320 | So, on a free cashflow basis, you know,

00:07:35.320 | the multiples are actually trending

00:07:36.800 | right around that 10-year average.

00:07:38.920 | You see this next slide,

00:07:40.080 | which I think points out really well,

00:07:42.720 | this is the median free cashflow margin

00:07:44.880 | of the same basket of software companies.

00:07:46.960 | So, you see these companies were, you know, remember,

00:07:49.520 | you remember back in 2015, it's hard to raise capital

00:07:52.680 | if you were a software company.

00:07:54.980 | You know, certainly in '13 and '14, it was.

00:07:57.360 | So, everybody had to fund themselves.

00:07:59.120 | Then we entered this period again, you know,

00:08:02.240 | around Zerp where everybody said, oh, I guess it's just

00:08:05.980 | growth at all costs and who cares about profitability?

00:08:08.600 | And now we're seeing the return to that.

00:08:10.720 | But the final chart, at least on this bit

00:08:13.600 | that I thought would be interesting, you know,

00:08:15.720 | everybody talks about the rule of 40.

00:08:18.360 | So, this is a dispersion, this is a scatterplot

00:08:21.160 | just of all the software companies, you know,

00:08:23.720 | against that rule of 40, you know,

00:08:26.040 | and you see where, you know, Box, you know,

00:08:28.440 | stands on that currently, you know, like, you know,

00:08:31.200 | just to calibrate, you have one of the highest

00:08:33.520 | free cashflow margins, I think, you know,

00:08:36.680 | in the software universe, around 29%,

00:08:39.640 | but currently are kind of that mid single digit growth rate.

00:08:42.800 | And I think this is the point that Bill talks about

00:08:45.280 | that I'm super interested in, because you have to deal

00:08:47.800 | with the animal instincts of the market.

00:08:50.240 | And there's certain phases of the market

00:08:52.200 | where they're valuing growth really highly

00:08:54.200 | and certain phases of the market

00:08:56.040 | where they're valuing profitability really highly.

00:08:58.760 | It sounds like you just kind of, you know,

00:09:01.080 | ignored Zerp a bit, didn't get caught up in this,

00:09:03.840 | but when you looked around, right,

00:09:05.560 | Frank Slootman said to me, probably in 2021,

00:09:08.560 | most software companies in Silicon Valley are walking dead

00:09:11.320 | and they don't even know it, right?

00:09:13.280 | When you looked around in 20 and 21 and look around today,

00:09:16.960 | how do you think your fellow CEOs and founders

00:09:20.840 | are balancing that profitability versus growth trade-off?

00:09:24.920 | - A lot to unpack here.

00:09:26.680 | So, first of all, I mean, empirically,

00:09:28.800 | the data is showing that most public company CEOs

00:09:32.400 | have dealt with this relatively well.

00:09:34.680 | They have sort of responded to the crisis

00:09:37.400 | and, you know, Brad, especially, obviously,

00:09:39.120 | you put out great content that I think has catalyzed

00:09:43.760 | that across a number of CEOs as well.

00:09:46.720 | And I think there's been a clear wake-up call to,

00:09:49.280 | you know, let's just say, you know,

00:09:50.800 | my contemporary group of, you know,

00:09:53.520 | 2010 onward SaaS companies where, you know,

00:09:57.400 | we benefited from relatively cheap capital,

00:10:00.480 | you know, relatively high valuations.

00:10:02.360 | We could use that capital to, again, grow at all costs.

00:10:05.560 | I would, you know, if I represent, you know,

00:10:08.600 | probably many of the names on this group,

00:10:10.200 | I think that was the exactly right economic decision

00:10:12.240 | at the time.

00:10:13.080 | There was sort of a market share war.

00:10:15.040 | You were acquiring customers that maybe had a,

00:10:18.480 | sort of a five or 10-year, you know,

00:10:20.280 | kind of lifetime value curve associated with them,

00:10:22.760 | maybe even more.

00:10:24.160 | And so, of course, you want to gobble up

00:10:25.600 | as much market share as humanly possible.

00:10:27.200 | There's a lot of stickiness to the platform.

00:10:28.760 | There's some kind of network effects

00:10:30.480 | in some of these enterprise software players.

00:10:32.400 | - And that's what you were referring to earlier, Aaron,

00:10:34.680 | when you said doing the right thing at the time.

00:10:37.120 | - Yeah, I think that, I just wish I could have negotiated

00:10:40.080 | maybe slightly better terms, you know,

00:10:42.040 | from a dilution standpoint, but, you know,

00:10:44.560 | there's little I would have done dramatically different

00:10:46.800 | in terms of putting our foot on the gas.

00:10:48.720 | - If you could go back, yeah.

00:10:49.680 | - Yeah, like, with the perfect benefit of hindsight,

00:10:51.960 | there's some things in maybe demand gen

00:10:54.240 | that were unprofitable that we'd like to have not done.

00:10:56.320 | There are some decisions, like, you know,

00:10:59.040 | it just, to all the founders out there,

00:11:00.600 | like, there are lots of little things,

00:11:02.240 | and feel free to call me, like, you can just,

00:11:04.400 | you can actually have better terms of your billings

00:11:07.600 | with customers that brings in more cash flow up front,

00:11:09.880 | which means you raise less money.

00:11:11.480 | Like, all of these very boring things

00:11:13.960 | actually just mean you can be more capital efficient

00:11:16.440 | and still grow at the same rate.

00:11:18.000 | And, you know, we had a fantastic board.

00:11:20.000 | They were pushing us and governing us appropriately

00:11:21.960 | on these things, but there's lots of little tiny,

00:11:24.560 | you know, single decisions that can be the difference

00:11:26.600 | of an extra five or 10% dilution in some cases.

00:11:29.680 | - Wow.

00:11:30.880 | - And, you know, if I could go back,

00:11:32.920 | I maybe would be incrementally more thoughtful

00:11:34.400 | on a bunch of those.

00:11:35.240 | But the underlying element of grow at all costs,

00:11:38.080 | get market share, you know, get to 100,000 customers

00:11:41.000 | I think was pretty important.

00:11:42.280 | And I think so is true for the Slack and Okta

00:11:45.160 | and Zooms of the world.

00:11:46.400 | Now, some of those names, you know, sort of,

00:11:48.760 | I just would, you know, mostly blame Wall Street.

00:11:51.840 | Some of those names got, you know,

00:11:53.840 | so hyper accelerated into the future on their valuations,

00:11:58.280 | you know, way outside, you know,

00:11:59.720 | what the market was actually gonna be able

00:12:01.800 | to actually support.

00:12:04.400 | And so then that's why you have this, you know,

00:12:06.200 | crazy volatility of multiples, you know,

00:12:08.720 | versus either revenue or free cash flow.

00:12:11.240 | And then now, obviously, and, you know,

00:12:13.400 | I think companies sort of invested, you know,

00:12:16.600 | maybe as a result of getting that response,

00:12:18.880 | you know, from Wall Street,

00:12:20.120 | and then now are obviously pulling back.

00:12:22.080 | But I think that from a public company standpoint,

00:12:25.120 | I think the message is clear.

00:12:27.080 | Everybody understands.

00:12:28.160 | I think the private market's

00:12:29.320 | probably still a little bit different

00:12:30.320 | 'cause you can do this under the covers.

00:12:32.360 | You don't have, you're not gonna get judged

00:12:34.320 | on a quarterly basis on, you know,

00:12:35.880 | how much have you reduced your expenses.

00:12:38.560 | And, you know, there's probably less risk

00:12:40.160 | of getting a letter from Brad, you know,

00:12:42.480 | in those kinds of environments.

00:12:44.240 | - That is a huge risk these days.

00:12:46.040 | - Come on, come on, come on.

00:12:48.720 | - But ultimately, so I think what we, you know,

00:12:52.400 | have is probably, you know, dozens,

00:12:54.120 | maybe hundreds of companies where they do have,

00:12:56.760 | they probably could not file an S-1 today,

00:12:59.840 | you know, with their current economics,

00:13:01.600 | but they might actually be able

00:13:02.560 | to either grow into their numbers,

00:13:04.600 | or they can do some of these changes behind the scenes,

00:13:06.880 | and then ultimately have a bit more of a, you know,

00:13:10.320 | pristine set of financials

00:13:11.560 | when they finally go public.

00:13:13.160 | So I think we've been afforded the amount of time,

00:13:16.680 | partly because of the AI wave, you know,

00:13:19.200 | partly because of some kind of, you know,

00:13:21.200 | soft landing in the economy where like,

00:13:23.400 | there's gonna be a lot more successful outcomes here

00:13:26.080 | than I think we would have probably thought

00:13:27.440 | a year or two ago.

00:13:28.880 | - Hey, Aaron, a quick question for you.

00:13:31.240 | Obviously, there are a lot of software companies

00:13:33.040 | in the Valley.

00:13:34.640 | Based on what you've been through,

00:13:36.000 | if you were advising them on what metrics matter most,

00:13:40.000 | what's at the top of the list

00:13:41.480 | for them to pay attention to internally?

00:13:44.280 | - There's probably a few that are just like literally

00:13:46.000 | the underlying business model economics.

00:13:48.680 | Like I'm a big believer in gross margin.

00:13:50.680 | Your gross margin will ultimately determine

00:13:53.640 | your operating margin.

00:13:55.080 | There's almost no way that you can kind of make those two,

00:13:58.560 | you know, kind of get out of sync.

00:14:00.560 | And so if you're subsidizing something,

00:14:03.240 | or, you know, or in just such a commodity business,

00:14:06.560 | and you're, you know, 40, 50, 60% gross margin,

00:14:09.160 | like there's just like no way

00:14:10.640 | you're gonna have an operating margin

00:14:11.760 | that looks like a software company.

00:14:13.200 | Understanding your gross margin,

00:14:14.640 | managing to gross margin, I think is super important.

00:14:17.880 | You know, all forms of LTV CAC are probably good.

00:14:21.000 | I don't know what the latest, you know,

00:14:23.280 | everybody, every two years is a new term

00:14:24.800 | in the industry that is used.

00:14:26.800 | But like something that just shows

00:14:28.320 | that you can acquire a customer profitably,

00:14:30.880 | and whether the payback is a year or two years

00:14:32.960 | or three years, almost doesn't matter

00:14:34.400 | as much as do you, you know, ultimately generate

00:14:39.080 | a long-term, you know, sticky customer.

00:14:42.240 | I think cashflow is super important.

00:14:45.920 | I've definitely like, I've gotten religion on cashflow.

00:14:50.560 | And I think companies getting the cashflow sooner

00:14:54.840 | is a really good move.

00:14:56.480 | I think it pushes the business to be super focused.

00:14:59.200 | There was a lot of, again, kind of sloppiness

00:15:01.960 | around the edges that we had in our hyper growth years

00:15:05.480 | that if we maybe had had cashflow

00:15:07.600 | as a more top of mind metric,

00:15:09.800 | we probably could have executed the same results,

00:15:12.520 | but around the margin, we would have been, you know,

00:15:14.680 | just doubling down the things that were working,

00:15:17.000 | the incremental international region

00:15:18.720 | that was sort of like, you know, we were, you know,

00:15:21.400 | just betting on we probably would have waited to invest in.

00:15:24.440 | And then ultimately again, you know,

00:15:26.520 | burn less cash, gotten to the same point.

00:15:29.000 | So I think cashflow is, you know,

00:15:30.720 | I definitely encourage founders to raise less

00:15:34.240 | than I would have five or 10 years ago

00:15:36.000 | and focus more on, you know,

00:15:37.600 | self-sustaining, you know, business models.

00:15:40.040 | - The comment you made about earlier

00:15:42.360 | about managing your cashflow,

00:15:44.800 | and I presume the same,

00:15:46.240 | you're meaning the same thing on gross margin management.

00:15:49.400 | I've just found very few companies in the Valley

00:15:52.840 | pay attention to this.

00:15:53.880 | Collections is another one that's in this area.

00:15:56.920 | So, you know, a company will be at 70, 80 million

00:16:01.040 | and they don't have any good processes around collections.

00:16:05.280 | And so their DSOs are just slipping out.

00:16:07.760 | They could be way sooner

00:16:09.320 | and they're just not doing the work.

00:16:10.960 | - I mean, they're like,

00:16:12.400 | you guys should rally to get collections people

00:16:15.120 | to get paid way more or something.

00:16:16.520 | So like, just, there are some of these functions

00:16:19.680 | that are so high leverage that if you just nail it,

00:16:22.560 | you know, the deal desk legal team

00:16:25.280 | who's negotiating the contract terms.

00:16:27.640 | I mean, these things are like,

00:16:28.880 | they literally are trajectory defining in your cashflow.

00:16:31.800 | - Earlier in the podcast,

00:16:33.840 | you mentioned net dollar a couple of times in a row.

00:16:36.600 | How important is that?

00:16:38.680 | - I think incredibly important.

00:16:39.800 | You can make, you know, any business model work

00:16:42.000 | if it's 90 or 100 or 110,

00:16:44.160 | but like, there's no question that if you can be,

00:16:46.680 | you know, in our best years, you know,

00:16:48.640 | in the hyper growth year, we're 130, 140, 150.

00:16:51.600 | That comes down at some point.

00:16:53.760 | And, but, but, you know, anything, I mean, just,

00:16:56.840 | I mean, you guys have all the math on this,

00:16:58.240 | but just like as high of a number as possible, so.

00:17:01.360 | - You know, one of the things you said earlier,

00:17:03.040 | I mean, there's no doubt Silicon Valley has contributed,

00:17:07.840 | I think, to the confusion here, right?

00:17:10.040 | And one of the things that, you know,

00:17:11.760 | the rules of thumb that we, we've talked about everybody--

00:17:14.680 | - Wait, who are you blaming?

00:17:16.200 | - No, I'm saying the investors have contributed, right?

00:17:19.720 | - VC.

00:17:20.560 | - And at the end of the day, you know,

00:17:22.120 | like Bill has this aphorism that, you know,

00:17:25.240 | venture capital ultimately follows the public markets,

00:17:28.200 | right, and founders follow venture capital.

00:17:30.320 | And the reality is that part of the reason Bill and I,

00:17:33.560 | I think, are such advocates for getting

00:17:35.080 | into the public market sooner as a founder,

00:17:37.920 | whether, you know, like Benioff did and Bezos did

00:17:40.160 | and so many others,

00:17:41.440 | is because it exacts a bit of a discipline.

00:17:43.480 | But Bessemer came out with an analysis last week.

00:17:46.440 | We have a chart on this I want to pull up

00:17:48.360 | called Rule of X, right?

00:17:50.120 | And I would love to see us get rid of Rule of 40

00:17:52.600 | and start talking about Rule of X,

00:17:54.480 | because what Rule of X does is it acknowledges something

00:17:59.200 | that all investors know to be true,

00:18:00.880 | which is growth and margin is not treated equal, right?

00:18:04.680 | And so, you know, for everybody, you know,

00:18:06.920 | who's may not be as familiar with Rule of 40,

00:18:09.160 | it's this idea that if you have a 20% free cashflow margin

00:18:12.120 | and a 20% growth rate,

00:18:13.800 | then that you're at a Rule of 40, right?

00:18:17.040 | And I think in the case of Box, Aaron, you know,

00:18:19.520 | if you take your, let's call it 20, 30% free cashflow margin

00:18:23.840 | and, you know, and 5% growth rate, even, you know,

00:18:27.520 | that would be a, you know, a 35%, you know,

00:18:30.360 | 35 on the Rule of 40.

00:18:32.320 | What this chart shows is that, you know, again,

00:18:35.600 | during the ZURP period, right,

00:18:37.080 | growth was a huge multiplier.

00:18:39.400 | Like that's all people cared about.

00:18:41.160 | But if you look at it over time,

00:18:42.840 | the average has been two to three.

00:18:45.120 | And I asked my team to actually plot the regression

00:18:47.840 | on this now, just so I could see

00:18:49.600 | what today are we valuing the most?

00:18:51.880 | And that's this next chart,

00:18:53.400 | which basically shows growth is being valued

00:18:55.880 | at about three times, right?

00:18:58.120 | In the Rule of 40 calculation, you know,

00:19:01.200 | what margin is being valued at.

00:19:03.280 | So I think one of the things I would like to see change

00:19:06.400 | in the nomenclature from the investment community

00:19:08.840 | in Silicon Valley is that we get over this enslavement

00:19:12.200 | to Rule of 40.

00:19:13.080 | At the end of the day, you said it.

00:19:15.480 | For a public market investor, when I look at Snowflake,

00:19:18.200 | or I look at Palo Alto Networks, or I look at Azure,

00:19:21.560 | I'm looking at what their free cash flow

00:19:23.640 | is going to be in 2025.

00:19:25.760 | I'm applying a multiple to that,

00:19:27.400 | and I'm discounting back to where we are today.

00:19:30.200 | Whether you're Rule of 40 or a multiple of revenue,

00:19:32.680 | those are shorthands for getting

00:19:34.600 | at the multiple of free cash flow.

00:19:37.200 | And so I'm just curious, you know, does this make,

00:19:40.120 | you know, when you see this Rule of X,

00:19:42.720 | I assume it's instinctive to you that you get,

00:19:46.240 | you know, how this changes over time.

00:19:48.720 | - Yeah, I mean, A, 100%, B, I would love to move on

00:19:53.160 | from this sort of, you know, random combination

00:19:55.840 | of two metrics that doesn't tell you a full story.

00:20:00.320 | I guess the question would be, you know,

00:20:05.640 | in this model, if you have a 40% growth company

00:20:09.520 | at exactly 0%, or whatever, it doesn't have to be 40,

00:20:12.240 | but like, but 0% cash flow, and whatever the growth rate is,

00:20:17.240 | how do you use your judgment to decide

00:20:19.320 | when will that company be able

00:20:20.760 | to actually produce the cash flow,

00:20:22.360 | and do the underlying economics, you know,

00:20:24.400 | sort of support that?

00:20:25.600 | And so then you don't have sort of these money pit,

00:20:28.560 | you know, burning companies getting valued

00:20:31.520 | in the same way as a snowflake, let's say,

00:20:34.600 | or a workday or whatever, who like,

00:20:36.720 | you kind of deeply understand it's going to be

00:20:38.320 | a 30 or 40% margin company at some kind of point

00:20:41.400 | in the future.

00:20:42.240 | So you have to have some way to kind of normalize for that,

00:20:44.120 | because again, not all growth, you know,

00:20:46.280 | rates are created equal as well.

00:20:48.760 | And I don't exactly know how you do that

00:20:51.400 | other than just individual company analysis

00:20:53.800 | to just think through, like,

00:20:55.040 | what are the pricing pressure dynamics

00:20:57.480 | in this industry going to be?

00:20:59.560 | You know, what part of this particular product

00:21:01.920 | gets commoditized, where is there going to be,

00:21:04.040 | you know, long standing, you know, kind of moat,

00:21:07.960 | or, you know, for lack of a better term.

00:21:10.560 | And unfortunately, none of these models,

00:21:12.160 | you know, sort of capture that

00:21:13.240 | when you look at a regression line.

00:21:14.560 | So, I don't know how to solve that.

00:21:16.760 | - But, I mean, literally at the peak of the madness,

00:21:19.640 | people were saying, oh, you know,

00:21:21.200 | you can do software investing

00:21:22.680 | just using quant analysis, right?

00:21:24.440 | Just say, you know, slap 100X on whatever the ARR is,

00:21:27.840 | and, you know, this is super easy.

00:21:29.480 | And, you know, I remember back in 2013,

00:21:32.000 | when we invested in Snowflake,

00:21:33.800 | nobody was investing in software.

00:21:37.680 | And part of the view at the time was, you know,

00:21:40.400 | like, we were very focused on database software,

00:21:43.720 | because it was the only software

00:21:45.320 | that we had really seen scale to a billion dollars in revenue.

00:21:48.560 | And a lot of this application software

00:21:50.720 | would hit a ceiling, right?

00:21:52.840 | Somewhere around 70 million, 80 million, 90 million,

00:21:55.640 | and never got to the margins that were promised.

00:21:58.280 | But, you know, you could look at, you know,

00:22:00.600 | Teradata or Netezza, or, you know, or IBM or Microsoft,

00:22:05.040 | the database market was always one of those markets

00:22:07.480 | that was incredibly large,

00:22:09.400 | and that could get to these mature margins.

00:22:12.240 | - Yeah, I mean, this obviously makes your business

00:22:16.280 | actually have differentiation in it.

00:22:18.040 | But like, you know, I would see either IPO filings

00:22:21.320 | or, you know, analyst reports of,

00:22:23.480 | okay, you have a company growing at, you know,

00:22:27.280 | 30, 40% or something,

00:22:29.640 | and then you'd apply the same revenue multiple

00:22:33.080 | on that company as, let's say, a Snowflake or whatnot.

00:22:36.440 | But one has a 90% retention rate or something,

00:22:40.720 | which means that sales and marketing is required forever

00:22:45.720 | to refuel the customer base

00:22:48.000 | because of the churn rate of that product.

00:22:49.920 | And it means you'll never actually converge on,

00:22:53.280 | you'll never actually converge on cashflow

00:22:55.480 | at the level of a company with, you know,

00:22:57.240 | let's say a 110 retention rate or whatnot,

00:23:00.760 | because the moment you stop spending on sales and marketing,

00:23:04.080 | that customer will drop,

00:23:05.240 | the company will drop their customers

00:23:06.760 | and they won't have any cashflow.

00:23:07.680 | So, but like, but like from afar,

00:23:09.560 | they were like plotting on the regression line, you know,

00:23:12.240 | like they should be funded or, you know,

00:23:14.720 | like the valuation should be a Snowflake.

00:23:16.240 | So that's, I don't know how, you know,

00:23:18.360 | I mean, the market eventually kind of gets that right

00:23:20.640 | on a per name or category basis.

00:23:22.680 | But I do think, like,

00:23:23.960 | if you don't really understand the dynamics

00:23:26.240 | of these categories,

00:23:28.160 | it gets easy to sort of not understand like,

00:23:31.360 | well, how will this company actually produce real cashflow

00:23:33.960 | in the future in three or five or 10 years out?

00:23:36.280 | - Yes.

00:23:37.120 | - Hey, one quick question for you, Aaron.

00:23:38.840 | I've noticed over many years of investing in startups

00:23:42.480 | that a lot of times the product market fit

00:23:45.080 | of the first product will get you,

00:23:47.480 | especially a really good product market fit,

00:23:49.400 | will get you to a hundred million or 200 million

00:23:51.920 | or some really large level,

00:23:55.200 | but then start to peter out.

00:23:56.960 | And then you have to figure out other growth drivers.

00:24:00.520 | And so I imagine, you know,

00:24:02.600 | just based at the scale you're at,

00:24:04.120 | you've had that kind of thought process,

00:24:06.360 | like how do we drive growth for a more mature startup

00:24:10.480 | or a company that's a little bit further along?

00:24:13.160 | Any hints for everybody?

00:24:15.440 | - Yeah, it's so case by case,

00:24:16.840 | 'cause it all depends on probably

00:24:17.920 | how big your first category is.

00:24:20.400 | And so, you know, as a founder,

00:24:23.120 | you always think your first category

00:24:24.560 | is gonna be just insanely massive.

00:24:25.960 | And so you can kind of run the clock on that.

00:24:28.520 | And we were really,

00:24:30.800 | I think we actually paid a lot of attention to this

00:24:32.640 | around kind of three, four or 500 million in revenue.

00:24:35.320 | We said, okay, what's our next act?

00:24:37.640 | And I think McKinsey did this analysis like 10 years ago

00:24:42.640 | around when your second act has to happen

00:24:47.480 | in your growth curve

00:24:48.320 | or else you basically just peter out.

00:24:50.720 | So we got like pretty obsessed with that.

00:24:52.240 | And so our second act was really our platform business,

00:24:56.800 | which turned us from just being an application

00:24:58.960 | that you bought at an end-user license

00:25:00.960 | to then you buy actually like our platform utilization.

00:25:04.720 | And that certainly worked, that gave us a boost.

00:25:07.840 | That has had different characteristics

00:25:09.440 | than just getting us to the billion

00:25:12.000 | as quickly as we wanted.

00:25:13.000 | So we had to get into security and workflow.

00:25:14.880 | So it's like, we're beyond like second act

00:25:18.760 | and it's more like we've got multiple

00:25:21.160 | kind of additional capabilities that we monetize.

00:25:24.320 | But you do have to,

00:25:25.520 | there's definitely like a finessing

00:25:27.080 | to your product strategy, your go-to-market strategy.

00:25:30.440 | Like we used to be like super SKU oriented

00:25:32.880 | and then we had to rebundle things.

00:25:34.800 | And so your whole company is sort of,

00:25:37.040 | riding these shifts in some kind of concerted way.

00:25:42.040 | And it's interesting like part of this,

00:25:44.600 | part of this as like a founder

00:25:48.160 | and just your own kind of like acknowledgement

00:25:51.720 | of where you are in the journey is sort of saying,

00:25:54.640 | okay, like there's two kinds of companies maybe out there.

00:25:57.120 | Let's just like two kinds of successful companies out there

00:25:59.920 | for getting the ones that don't work or get acquired.

00:26:03.440 | There's, wouldn't it be great

00:26:05.120 | if everybody was Google or Meta, that would be great.

00:26:07.000 | Like you just grow 50% a year forever

00:26:09.440 | and then 10 or 20% a year.

00:26:12.040 | But then, and that's fantastic.

00:26:13.640 | Like I'd love to have a trillion dollar company.

00:26:16.200 | Wouldn't love to be in Congress like Zuck is,

00:26:18.400 | but other than that.

00:26:20.360 | And then there's,

00:26:21.760 | but then there's like actually like extremely good,

00:26:25.000 | valuable companies that you're,

00:26:26.800 | you just are building a real business.

00:26:28.720 | And like, you don't like,

00:26:30.720 | you don't think about them as Google or Meta or whatnot,

00:26:32.960 | but they produce an insane amount of value for the world,

00:26:37.360 | for their customers, for their cultures and employees.

00:26:40.840 | And so we started studying the Autodesks

00:26:43.680 | and Adobes of the world,

00:26:45.160 | which is like, okay, well like what's a compounding model

00:26:47.160 | where you're growing 10 or 20%

00:26:49.920 | and your free cashflow is 30 to 40%.

00:26:52.960 | And you just do that for like literally decades.

00:26:56.200 | There was a company we studied a ton that,

00:26:59.800 | it almost further proves the point

00:27:01.520 | in the past couple of weeks, Ansys.

00:27:03.440 | We studied these guys a year or two ago

00:27:05.000 | and it's a $30, $40 billion,

00:27:07.160 | and that being $30 billion company,

00:27:09.600 | they just own their space of CAD,

00:27:12.520 | simulation, industrial software,

00:27:15.480 | maybe $2 billion revenue, plus or minus,

00:27:18.040 | 40, 30, 40% margin.

00:27:20.120 | Like that's a fantastic business.

00:27:21.920 | Like everybody would love

00:27:22.960 | a 20 to $30 billion market cap company.

00:27:25.360 | And they don't have to deal with the sort of daily changes

00:27:30.360 | of Wall Street of, okay,

00:27:31.960 | we're valuing growth at this percent more or whatnot.

00:27:34.240 | They're just like,

00:27:35.080 | let's just build a very profitable business,

00:27:37.360 | like dominate our market and scale that.

00:27:39.840 | And I think that, you know,

00:27:41.040 | I think it'd be great if the Valley also got, you know,

00:27:43.960 | more enthralled by those types of names.

00:27:45.920 | - I love that idea of studying other businesses.

00:27:49.640 | That's another thing that I think people don't do enough of,

00:27:52.760 | but going out and getting the peer data,

00:27:55.040 | looking at how others did it, finding those experts.

00:27:58.160 | - Have you looked at like Cadence?

00:28:00.320 | - Not deeply.

00:28:02.560 | - Okay, of course Brad has,

00:28:04.320 | but like we should all be talking about Cadence more.

00:28:07.120 | Like we should be talking about Cadence.

00:28:08.920 | We should be talking about Synopsys.

00:28:10.600 | We should be talking about Ansys.

00:28:12.000 | Like these are crazy,

00:28:13.640 | like literally like 15 years ago, Cadence was a,

00:28:16.680 | I'm going to make up a number,

00:28:17.520 | so please just like do all your charts or whatever.

00:28:19.280 | Like, I don't know, three, $5 billion company.

00:28:21.400 | It's a 50, $60 billion company today.

00:28:23.520 | Like, and just dominate a vertical and just keep building

00:28:27.240 | and just do it profitably.

00:28:28.520 | And I think if we got as enamored by that as building Meta,

00:28:33.240 | you know, first of all,

00:28:34.160 | we'd have a lot less stress all the time.

00:28:36.200 | But I think we'd just be building, you know,

00:28:39.040 | more attractive and valuable companies.

00:28:41.040 | - Speaking of dominating in Meta,

00:28:43.400 | several of the large companies reported last week, Brad.

00:28:47.440 | And if I think about the public markets,

00:28:50.640 | like prior to, right prior to 2024,

00:28:53.520 | everybody's talking about Magnificent Seven, Own 2023,

00:28:57.960 | and everybody's talking about cost reduction in the Valley

00:29:00.680 | is something you're very familiar with

00:29:02.760 | from your back and forth with Meta.

00:29:04.840 | So what did we learn this week that,

00:29:08.480 | and how does it affect the playing field?

00:29:10.160 | - Well, yeah, I mean, so you and I talked about last week,

00:29:13.960 | you know, 2023, like all these things were up a lot,

00:29:17.200 | but it was really reversion to the mean.

00:29:19.240 | I mean, we started 2024 still trading below

00:29:22.240 | kind of the 10-year average.

00:29:23.600 | That's how debilitating '22 was, right?

00:29:27.920 | But I said this year,

00:29:28.920 | like I thought this year would be more normalized.

00:29:30.960 | Like I think return expectations, you know,

00:29:33.360 | are more in kind of that 15 to 25 range,

00:29:35.640 | not, you know, the return to target of 80

00:29:37.600 | like we had at the start of '23.

00:29:39.800 | But, you know, I went into Q1

00:29:41.480 | kind of holding my breath a little bit to see, right,

00:29:44.200 | what in fact, whether or not we were seeing

00:29:46.120 | any of this acceleration.

00:29:47.320 | So, you know, we had Google, we had Meta,

00:29:49.680 | we had, you know, Amazon and Microsoft report last week.

00:29:53.280 | And I had a couple, a few big takeaways.

00:29:58.040 | The first was, you know, the Mag Seven's

00:30:01.040 | really not the Mag Seven

00:30:02.080 | and maybe the Mag Four right now, right?

00:30:03.880 | Because we have, you know, a few of these companies,

00:30:06.720 | Apple's flat on the year, Tesla's down on the year,

00:30:09.680 | Google's basically flat on the year, you know,

00:30:12.040 | and you have four companies that are-

00:30:13.480 | - We could have gotten ChadGBT

00:30:15.200 | to do a really nice illustration of that dynamic,

00:30:19.080 | the seven with three of them falling behind.

00:30:21.760 | - I think- - We still can, I guess.

00:30:24.280 | - We'll insert that post-fact.

00:30:25.800 | You know, but last week, what we saw at a Meta,

00:30:29.640 | and I think this is really significant, you know,

00:30:32.000 | obviously as a big beaten race,

00:30:34.520 | stocks up over 30% year to date,

00:30:36.680 | they're clearly benefiting, right?

00:30:38.760 | They are at the tip of the AI spear.

00:30:41.800 | Their video engagement was up 25% year over year,

00:30:45.320 | which is just like hard to get your head around.

00:30:47.600 | There's only one way to do that,

00:30:49.400 | that's through the application of AI.

00:30:52.120 | We're seeing their monetization also up a lot, you know,

00:30:56.000 | but here's the thing that I think is so impressive, right?

00:30:59.360 | They nearly tripled their earnings per employee, right?

00:31:02.600 | Their head count in Q3 of 22 was 87,000,

00:31:05.680 | this quarter they reported 67,000.

00:31:08.160 | So they're doing this while they're like,

00:31:10.600 | they really believe that Flatter is faster.

00:31:13.480 | And I think there are a lot of people

00:31:14.600 | who are cutting 9% or 12% or this or that, you know,

00:31:18.680 | and it's not a fundamental change

00:31:20.600 | to the ethos of the business.

00:31:22.320 | It's just a little bit more CYA, you know,

00:31:25.080 | and I think that, you know,

00:31:26.640 | what you see with Mark is a real commitment

00:31:28.840 | to transform that business,

00:31:31.080 | and it's showing up in, you know, in the results.

00:31:33.320 | And, you know, I think I tweeted something about,

00:31:36.520 | you know, Reality Labs is still losing,

00:31:38.600 | I think, $20 billion.

00:31:40.640 | And if you look at that on a per share number, okay,

00:31:43.560 | and that's about $6 a share, you know,

00:31:47.920 | in losses for the company due to Reality Labs.

00:31:50.240 | And if you apply a 20X multiple to that,

00:31:52.440 | that's about negative $120 a share

00:31:55.840 | that's being attributed to the valuation of Meta

00:31:58.880 | as a result of their--

00:31:59.720 | - Hold on, hold on, Jack, that's only true

00:32:03.160 | if you put zero future free cash flow

00:32:06.760 | from that unit in your analysis.

00:32:08.880 | - Yeah, no, I-- - Which maybe you're doing.

00:32:10.520 | - Well, I would like to think that everybody

00:32:12.120 | was doing that analysis, Bill,

00:32:13.440 | but you and I both know that most sell-side analysts

00:32:16.000 | are slapping a 20X multiple on earnings for this company,

00:32:18.880 | and, you know, that's what suffices for research these days.

00:32:22.120 | And I would just say, in a consolidated view,

00:32:25.680 | that is pretty crazy to me,

00:32:27.000 | and Zuck even mentioned on the call--

00:32:29.200 | - Well, but also, I mean, alternatively, Bill,

00:32:32.800 | like, it's not, like, it's also not crazy

00:32:34.760 | because how should you figure out, like,

00:32:37.760 | what the actual profit potential is of Reality?

00:32:40.400 | Like, we have no evidence at the moment

00:32:42.560 | of what that looks like.

00:32:43.400 | So, everything else would just be a made-up number.

00:32:46.400 | - Fair. - So, I don't know.

00:32:48.560 | - But any non-zero, you know,

00:32:50.120 | one of the things that Mark said during the call

00:32:52.640 | is that the Meta AI glasses with Ray-Ban,

00:32:55.280 | they sold twice as many as they expected to sell, okay?

00:32:58.440 | I think we're gonna have this explosion of AI devices

00:33:00.760 | we're gonna talk about a little bit later.

00:33:03.280 | But he said that was the first clear example

00:33:05.280 | of some of this convergence

00:33:06.960 | between the work that they're doing in Reality Labs, right,

00:33:10.000 | and what they're doing in the rest of the business.

00:33:11.840 | And so, you know, again, as an investor,

00:33:15.520 | I assign greater than negative $120 a share value

00:33:20.520 | to Mark Zuckerberg investing $20 billion a year

00:33:24.360 | on something that's in the zip code of AI,

00:33:28.440 | you know, and augmented reality.

00:33:30.520 | The second thing we saw this week, Bill, Google,

00:33:33.600 | you know, I mean, it was a solid quarter,

00:33:35.720 | but the stock is flat to down now on the year.

00:33:39.440 | I think you're seeing increasing concern

00:33:41.440 | about the core business, you know,

00:33:44.080 | from threats, you and I've talked about this,

00:33:46.400 | chat GPT, perplexity, just go through the list,

00:33:49.360 | you go into any room,

00:33:50.400 | clearly answer engines are becoming more relevant.

00:33:53.120 | Well, you know, when you said, you know,

00:33:54.400 | chat GPT could have made the chart,

00:33:56.040 | you didn't say BARD could have made the chart

00:33:57.920 | or Google could have made the chart.

00:34:00.400 | So I would say Google Cloud was pretty uninspired.

00:34:04.000 | And I think, I don't think they've gotten real

00:34:06.640 | on the efficiency and the fitness.

00:34:08.360 | You know, I have a slide here,

00:34:09.600 | I asked my team to pull together.

00:34:11.960 | This slide plots, you know, what they've done on headcount

00:34:16.640 | and what that translates into

00:34:18.320 | in terms of earnings per employee

00:34:20.480 | over the course of those, you know, same six quarters, right?

00:34:24.000 | And so Facebook has actually reduced headcount significantly

00:34:28.160 | and their earnings per employee has exploded.

00:34:30.640 | And Google kind of, you know, they're fine,

00:34:32.800 | but they really just haven't made many changes.

00:34:35.240 | Now, if this was just about earnings per employee

00:34:38.200 | in the short term,

00:34:39.040 | then you see your differential

00:34:40.480 | in terms of share price performance right there.

00:34:42.800 | I think a much, much bigger issue here

00:34:45.840 | is the implications this has for future product development.

00:34:49.400 | The implications it has for, you know,

00:34:51.720 | real focus out of the business.

00:34:53.560 | And I just think this flatter is faster is very real.

00:34:57.360 | I think that Meta is very focused

00:34:59.280 | and I think that Google is still trying to sort out,

00:35:02.680 | you know, how to get focused around some of this stuff.

00:35:06.440 | And then finally, cloud software, you know,

00:35:09.320 | after a year of belt tightening,

00:35:11.160 | you know, what we really heard out of Amazon

00:35:14.600 | and Azure last week is that kind of core workloads

00:35:17.880 | are back to their prior growth rates, right?

00:35:19.920 | So that CFOs are largely through the, you know,

00:35:23.360 | the belt tightening on the core.

00:35:25.080 | And we saw real evidence, right,

00:35:27.720 | of AI workloads beginning to kick in.

00:35:30.600 | Both at Amazon, as well as at Azure.

00:35:32.920 | I think Azure, I've got a couple of charts here,

00:35:35.960 | you know, that the team threw together.

00:35:37.440 | The first is just, you know,

00:35:38.800 | you can see the reacceleration on the right-hand side

00:35:42.960 | in terms of this is all three

00:35:44.440 | of the cloud providers put together.

00:35:45.880 | So this is their growth, aggregated growth rate.

00:35:48.360 | And on the right-hand side,

00:35:49.360 | you can see you finally saw that turn, that reacceleration.

00:35:52.640 | Now, mind you, these things are reaccelerating

00:35:55.400 | in the case of AWS from $100 billion run rate, right?

00:35:59.240 | So like, you know, you're reaccelerating.

00:36:01.960 | These are massive growth rates at massive scale.

00:36:05.160 | On the left side, what you see

00:36:06.480 | is they actually set a record quarter,

00:36:09.200 | 15 billion in net new ARR across these three platforms.

00:36:14.200 | So I think, you know, that to me is, you know,

00:36:19.520 | kind of the real story that came out of, you know,

00:36:23.480 | out of the big cloud providers.

00:36:25.460 | And, you know, one of the questions I think

00:36:28.520 | that we now have to answer, you know,

00:36:31.720 | is does that read through apply

00:36:33.360 | to everybody else in software, right?

00:36:35.760 | Like, you know, is this something that's unique

00:36:39.700 | to the big cloud providers?

00:36:41.000 | And then the second question that I had in my mind,

00:36:43.160 | you know, was like, when do we just hit the wall on this?

00:36:46.520 | Right?

00:36:47.720 | I, you know, I had a friend to say to me the other day,

00:36:50.640 | they thought that three years from now

00:36:52.320 | that the cloud providers,

00:36:53.720 | the revenues would actually be smaller

00:36:55.840 | than they are today, right?

00:36:57.160 | Which I can't get my head around.

00:36:59.200 | When I look at this, you know,

00:37:01.080 | I talked to Adam Solepski who runs AWS,

00:37:03.240 | and he said, "Brad, this is a multi-trillion dollar market,

00:37:06.920 | and that does not include AI."

00:37:08.880 | Okay, so yeah, $100 billion is a lot,

00:37:11.880 | but it's a multi-trillion dollar market,

00:37:14.040 | and it doesn't include AI.

00:37:15.400 | So Aaron, I can't get back over to you.

00:37:16.960 | That's, you know, that's my brain dump

00:37:18.720 | on what happened in the public markets this week,

00:37:20.700 | but on the software side in particular,

00:37:23.040 | you know, were you surprised at these numbers?

00:37:25.360 | And how do you think about,

00:37:27.140 | is this unique to what's going on with the cloud providers?

00:37:31.120 | You know, or are you starting, you know,

00:37:34.000 | to, you know, hear other software companies

00:37:37.280 | talking about how, about similar trends?

00:37:40.340 | - Yeah, so first of all,

00:37:42.600 | you might wanna short your friend's portfolio,

00:37:45.760 | or do the, I've never heard that metric of,

00:37:49.480 | in three years from now, we're gonna be smaller,

00:37:52.400 | but that's an interesting take.

00:37:54.340 | I honestly am so done with trying to guess

00:37:58.260 | when this thing runs out.

00:37:59.400 | Like, it is, I think we just have to honest,

00:38:02.560 | like, you basically are just,

00:38:04.920 | it is an uncapped, unknowable total amount

00:38:08.280 | of addressable market in this space.

00:38:11.040 | There are clear differences between the hyperscalers

00:38:14.880 | and the rest of software simply because

00:38:17.040 | a company can just move their core data center infrastructure

00:38:20.640 | from their, you know, their current colo

00:38:23.280 | and server providers to the cloud.

00:38:25.600 | And, you know, some software providers

00:38:27.600 | will make money in that transition,

00:38:28.760 | but that's basically 90% hyperscaler benefit.

00:38:32.400 | And so that doesn't sort of all show up

00:38:34.280 | in the rest of software and SaaS in the same way.

00:38:37.020 | So there's some kind of tale of two cities

00:38:39.160 | of these massive infrastructure migration

00:38:42.480 | where the hyperscalers are just, you know,

00:38:44.620 | again, uncapped market opportunity for these guys.

00:38:48.760 | And then there's a lot of surrounding services

00:38:50.240 | like you would imagine that as this goes up,

00:38:52.080 | the snowflakes and confluence and whatnot

00:38:55.120 | also see some benefit 'cause there's lots of data services

00:38:58.080 | that you need around that.

00:38:59.800 | But yeah, we're not seeing any slowdown of, you know,

00:39:02.800 | companies that we talk to just continuing

00:39:04.960 | to be moving to the cloud.

00:39:07.280 | I, you know, get the fortune of talking to, I don't know,

00:39:10.240 | you know, dozens of CIOs a month.

00:39:13.480 | And there's nobody that is sort of like

00:39:15.860 | done with their cloud journey.

00:39:17.340 | There's nobody even at 90%.

00:39:19.740 | And this is across every sector of the economy.

00:39:22.420 | So we're still weirdly relatively early.

00:39:25.180 | And then to Adam's point, like, yeah,

00:39:26.660 | like that's even pre-AI.

00:39:27.820 | AI is just in the experimentation phase

00:39:30.340 | for all intents and purposes in the enterprise right now.

00:39:33.620 | And so we don't even have the AI TAM, you know,

00:39:36.860 | quite understood from a software

00:39:38.900 | or infrastructure standpoint yet.

00:39:40.400 | - Brad made the comment that in his view

00:39:44.040 | from checking around that CIOs had moved past

00:39:47.600 | their kind of shrink and reassessment.

00:39:50.120 | Do you agree with that?

00:39:52.000 | - I wouldn't want anybody to like trade

00:39:54.960 | on my like qualitative anecdotes, but-

00:39:58.120 | - Yes, you're right, Aaron and Bill.

00:39:59.480 | As a reminder to everybody,

00:40:01.160 | just our opinions, not investment advice.

00:40:04.040 | - But you know, there's like,

00:40:06.160 | there's very clearly this interrelated, you know,

00:40:08.680 | system of like, you are like listening to Powell's commentary

00:40:13.220 | and then deciding at your management committee meeting,

00:40:16.940 | you know, should we lean into budget, you know?

00:40:19.260 | And so, you know, if we're only hearing

00:40:22.020 | doom and gloom 18 months ago, inflation is exploding,

00:40:25.500 | interest rates are about to go crazy,

00:40:28.460 | massive layoffs happening in tech,

00:40:30.100 | that's like a scary environment to then be investing,

00:40:32.660 | you know, and leaning into whether it's a software company

00:40:35.860 | or an insurance provider in, you know, Minneapolis,

00:40:38.280 | like we're all, you know,

00:40:39.480 | kind of thinking about this the same way.

00:40:41.240 | Fast forward to today, you know, okay,

00:40:43.640 | inflation coming down clearly,

00:40:45.320 | interest rates either at the peak

00:40:46.920 | or starting to come down at some point,

00:40:48.820 | like I feel like, okay, we could invest a bit more,

00:40:50.960 | which means that that incremental application

00:40:53.080 | we wanted to launch that had some data consumption

00:40:56.200 | as a part of it, you know, the extra focus

00:40:59.600 | of the engineers working on optimization versus innovation,

00:41:02.780 | you can kind of tune that knob a little bit.

00:41:04.760 | And so I think that's the sort of qualitative

00:41:07.140 | not captured in the, you know, econometrics,

00:41:10.220 | you know, kind of elements of this

00:41:11.180 | that are actually happening now in businesses,

00:41:13.020 | which gives me a little bit of, you know,

00:41:14.860 | incremental confidence that that means CIOs

00:41:17.260 | are gonna be putting on more growth initiatives

00:41:19.400 | going forward.

00:41:20.240 | - Hey, Aaron, you just hit on something so important.

00:41:23.000 | You know, perhaps one of the greatest sources of alpha

00:41:26.220 | for me and certainly my mentors in this business

00:41:29.300 | over a long period of time is this idea

00:41:31.100 | of positive and negative reflexivity, right?

00:41:34.100 | That in fact, when there is all doom and gloom,

00:41:38.160 | that causes people to behave in a certain way, right?

00:41:42.360 | And when confidence begins, you know,

00:41:44.480 | like as we started entering this year,

00:41:46.600 | as confidence is turning and the second derivative

00:41:49.340 | on interest rates is down, et cetera,

00:41:52.360 | it leads to the positive reflexivity on the other side.

00:41:54.820 | So it actually leads to acceleration

00:41:57.680 | because people feel comfortable in doing that.

00:41:59.720 | And you need to invest ahead of that, right?

00:42:03.000 | And oftentimes to really get ahead of it,

00:42:06.060 | you have to be buying when there is, you know,

00:42:08.460 | the proverbial blood in the streets,

00:42:10.000 | when there is that doom and gloom

00:42:11.740 | and people say it can't ever, you know,

00:42:13.740 | it can't ever be different.

00:42:15.260 | - Well, yeah, yes.

00:42:17.060 | I mean, I can only imagine though,

00:42:18.020 | there is some timing element to that

00:42:19.340 | because you have to know when you've reached the bottom

00:42:20.740 | of the doom and gloom. - Of course, of course.

00:42:21.980 | - But that aside, which I don't actually wanna like

00:42:25.300 | veer this politically at all, but like,

00:42:26.900 | as just like a bookmark, this is,

00:42:29.060 | I get very confused why we are so passionate

00:42:31.380 | about changing the government right now.

00:42:33.900 | Like, they're like, they're nailing it.

00:42:35.400 | Like, why pull out a Jenga piece out of nowhere

00:42:39.200 | and then see what changes.

00:42:41.680 | But again, you know, different podcasts,

00:42:44.080 | but like, we are so lucky right now

00:42:48.060 | that we somehow landed this thing.

00:42:50.160 | And sure, there's, you know, a bunch of incremental issues

00:42:53.560 | on the margin that we're dealing with,

00:42:55.200 | but like, wow, we should like be, you know,

00:42:57.720 | not taking this for granted too much.

00:43:01.000 | So that's a software take.

00:43:03.120 | I know you have opinions on this other one as well.

00:43:05.520 | Right, there's, you know, there's a spicy battle out there

00:43:09.440 | about, you know, Google and 10 Blue Links and, you know,

00:43:12.680 | and what happens next.

00:43:14.200 | And, you know, I tweeted over the weekend, you know,

00:43:16.920 | this isn't a question about like knowing, right?

00:43:19.520 | Larry and Sergei always knew that 10 Blue Links

00:43:22.280 | was like a waypoint in route to answers, right?

00:43:25.600 | Like anybody who was in the business

00:43:27.560 | of information retrieval,

00:43:29.360 | or they've been around artificial intelligences

00:43:31.440 | and out in front of everybody else.

00:43:34.560 | And so the idea here is that, you know,

00:43:37.480 | I think it's not a matter of will,

00:43:40.400 | I don't think it's a matter of knowing, right?

00:43:42.720 | I just think it's this innovators dilemma, right?

00:43:45.920 | That Google faces, which is that 99% of their profits

00:43:49.200 | or 110% of their profits comes from an advertising model

00:43:53.880 | around 10 Blue Links that is just at its core

00:43:56.680 | seems somewhat at odds with what Perplexity is doing,

00:44:00.280 | with what ChatGPT is doing, the age of answers.

00:44:03.400 | And so I would just love to, you know,

00:44:05.880 | kick it around for a second.

00:44:07.440 | How, if you're running Google today, right?

00:44:10.280 | And you're faced with this innovators dilemma around this,

00:44:13.120 | like what should they be doing here?

00:44:16.000 | - Well above my pay grade, fortunately,

00:44:18.720 | I am just a lowly enterprise software.

00:44:20.600 | (laughing)

00:44:22.400 | But, you know, so as a reminder, obviously for everybody,

00:44:25.400 | like innovators dilemma is about business model,

00:44:28.200 | you know, difficulties.

00:44:29.160 | And we always think it's a technology problem.

00:44:31.360 | It's always about your business model,

00:44:33.480 | doesn't make the new thing attractive to you.

00:44:36.200 | And so you avoid doing it until you get disrupted.

00:44:38.800 | You know, I might take after, you know,

00:44:43.440 | thinking about this a little bit, maybe not enough,

00:44:45.240 | is I don't think there's a totally 100% classic

00:44:49.360 | innovators dilemma issue with the move to AI with Google.

00:44:53.320 | I think on the other end of this,

00:44:55.840 | the business model could be as good if not better.

00:44:58.640 | And Brad, you think about this a lot of like,

00:45:00.760 | if you actually had Google literally tell you

00:45:03.000 | the thing to buy and they got a transaction on that,

00:45:07.760 | it would actually be a similar business model.

00:45:10.560 | They would just get you to that answer even faster

00:45:12.680 | and you'd still transact and you'd still do commerce

00:45:14.760 | and they are still the distribution engine

00:45:18.280 | for all of the people that need to advertise somewhere.

00:45:20.880 | Like everybody has to find a customer

00:45:23.200 | and so you still need an interface

00:45:25.000 | to get my product into that customer's hands.

00:45:28.280 | And so if Google's AI is doing that better

00:45:31.040 | than the 10 blue link model, that's great.

00:45:33.240 | I actually think that the more complicated thing

00:45:35.360 | is probably more of a product management,

00:45:37.360 | user behavior experience issue,

00:45:39.280 | which is how do I navigate my customer base?

00:45:41.880 | How do I navigate my user base to this model

00:45:44.480 | in a way where in the process,

00:45:45.880 | I don't get totally hammered by that journey.

00:45:48.400 | And I don't know how to do that.

00:45:50.440 | Like Bart is clearly their way of experimenting to do that.

00:45:53.320 | But like, their big problem is like,

00:45:55.920 | you go to a search box, you type in a search

00:45:57.680 | and they give you a lot of results

00:45:58.960 | and Perplexity and ChatGBT

00:46:01.200 | are just totally different product interfaces.

00:46:04.080 | And so they can't clearly make

00:46:05.480 | the Google homepage just do that.

00:46:07.700 | So they have to kind of--

00:46:08.540 | - Doesn't that sound, doesn't that rhyme in your ear?

00:46:10.760 | I mean, 1999, Google was just a different interface

00:46:14.920 | relative to the portals that existed at the time.

00:46:17.920 | And you can say, why the hell didn't Yahoo just copy Google

00:46:21.440 | because they had an existing business model

00:46:24.040 | that required them to sell real estate on this big port.

00:46:28.720 | - There was, probably you're 90% more right,

00:46:31.720 | but there was also a literal tech differentiation

00:46:34.920 | that Google had, and Yahoo did not have PageRank.

00:46:37.840 | And so Google just literally gave you better results.

00:46:40.800 | And so the user base migrated.

00:46:42.460 | I mean, short of maybe some internal

00:46:45.160 | kind of political issues,

00:46:46.800 | like Google's technology is probably

00:46:48.720 | not in the way of this issue.

00:46:50.200 | And so now it's actually like a, as a user.

00:46:52.520 | Now, I will leave like one big X factor,

00:46:55.640 | which is just like, maybe our brains are the limiter here.

00:46:58.200 | And like ChatGBT just owns that little slot in our brain

00:47:01.880 | of you ask a question, get an answer.

00:47:04.200 | And like, now we have like a whole

00:47:05.520 | like brain rewiring problem

00:47:07.200 | that Google's gonna have to contend with.

00:47:09.440 | I don't know, I mean, like that's above the,

00:47:11.960 | that's above what a PM on the Google search homepage can do.

00:47:15.080 | So this one is, you know, pretty complicated.

00:47:19.840 | Bill, were you raising your hand?

00:47:21.600 | - I didn't want to interrupt.

00:47:23.240 | So for the college game day fans,

00:47:25.400 | I'll do a lead course, so not so fast.

00:47:27.720 | I'll take the other side of this one.

00:47:29.140 | - Okay.

00:47:30.460 | - I think it's a horrific problem.

00:47:32.320 | - Oh wow.

00:47:33.160 | - Because one is the user interface

00:47:38.820 | is subject to disruption.

00:47:40.460 | So the whole idea that I'm gonna throw you 10 links,

00:47:45.060 | it's really more than 10, right?

00:47:46.340 | 'Cause there's four ads on top and four ads on the bottom

00:47:49.860 | and you gotta search through and find what you want.

00:47:52.200 | I'm a big sports fan.

00:47:53.280 | I often search for the roster of a team.

00:47:55.740 | That's not even at the top.

00:47:57.400 | - No, yeah.

00:47:58.240 | - It's like five links down.

00:47:59.400 | I have to find it and get around the ads.

00:48:01.760 | So it's like, so that's one.

00:48:04.360 | Two, you already mentioned the business model.

00:48:07.120 | Their business model is to throw their customers

00:48:10.300 | in a cage match and let them compete

00:48:12.840 | with one another to the death.

00:48:15.060 | And that drove revenue sky high

00:48:19.980 | because it created Prado, like a super optimal payment.

00:48:24.580 | And no one wants to see four or five ads.

00:48:27.900 | You're not gonna get to the same revenue per visit

00:48:31.580 | with a transaction model that you do with the ad model

00:48:35.100 | because you already mentioned Cactail TV.

00:48:38.180 | People will pay using marketing math,

00:48:41.000 | 40, 50, 60% of first purchase

00:48:44.400 | with a transactional integration, they wanna pay 5%.

00:48:49.080 | So you have a 10X reduction if you were to partner

00:48:52.560 | on a transaction enablement than you do on a marketing lead.

00:48:57.400 | And so the customer wants neither of those on the screen.

00:49:02.040 | - Yeah, I'm gonna push back a little bit on the...

00:49:05.380 | So when I said transaction, it didn't necessarily mean

00:49:08.320 | that you turn it into a CPA model, but maybe you do.

00:49:12.080 | And I understand the changes in economics there.

00:49:14.600 | I guess what I'm saying is that when I do a Chachabity

00:49:17.120 | question of, "Hey, give me a recipe for this thing."

00:49:20.680 | It's actually a gap that they don't then let me

00:49:22.800 | go buy the things that...

00:49:24.920 | - But if you were to go buy 'em, you wouldn't buy 'em

00:49:26.920 | from a marketing model, you'd buy 'em

00:49:28.360 | from a transactional model.

00:49:29.600 | Otherwise you're gonna say, "I'd like to book a flight

00:49:32.720 | to Chicago and get a hotel room."

00:49:34.760 | And instead of doing it, the Google AI is gonna say,

00:49:39.760 | "Would you like to hear deals from Hotels.com

00:49:43.740 | and Expedia and Booking.com?

00:49:46.540 | Can I read you through the deals they're offering?"

00:49:49.160 | Like that's not...

00:49:50.640 | - Brad, you're the expert on this.

00:49:52.260 | Why can't Google get $40 for that flight

00:49:54.940 | or $30 for that flight instead of the $2

00:49:57.780 | for clicking through to Expedia?

00:49:59.700 | - Well, I think to Bill's point,

00:50:02.220 | if you look at the traditional commission models,

00:50:04.140 | and I think the iconic example is in travel,

00:50:07.520 | the traditional commission model, for example,

00:50:09.860 | in hotels is 10%.

00:50:12.380 | And let's just say the average daily rate is 100 bucks.

00:50:15.020 | So they're willing to spend on a commission model, 10 bucks.

00:50:18.180 | But in a marginal advertising model

00:50:21.420 | where they're bidding up all the way to their gross profit

00:50:25.780 | on in order to get that next customer

00:50:29.260 | to drive that growth rate, which I think...

00:50:31.380 | And if you're a startup, you certainly go above it.

00:50:34.380 | You may spend two or three X on that individual purchase.

00:50:38.340 | It certainly has extracted more rents.

00:50:40.500 | But I think the other piece here

00:50:42.420 | is that if we really just telescope way out here,

00:50:47.220 | like we are heading to answers and actions, right?

00:50:51.300 | And even the people who invented the damn Blue Link model

00:50:54.580 | don't believe that that is where this thing ends.

00:50:56.780 | So like, and all I'm saying is the idea

00:51:00.020 | that Google can replicate a 99% monopoly

00:51:03.900 | and take all of that pool of profits with it

00:51:06.660 | into this new world after letting chat GPT become the verb,

00:51:11.660 | at least at the start for what this new world is.

00:51:15.620 | It just, that to me is almost impossible to believe,

00:51:18.820 | but the consensus view has continued to be

00:51:21.540 | that they're going to dominate this new thing

00:51:23.880 | the way they dominated the world of search.

00:51:26.060 | And I just think if they pull that off,

00:51:28.140 | I hope I'm a shareholder along the way,

00:51:29.940 | that will be one of the greatest jujitsus

00:51:32.020 | in the history of capitalism.

00:51:33.660 | I'll be remarkably impressed.

00:51:34.980 | One last thing.

00:51:36.220 | There's one last problem.

00:51:37.820 | In addition to everything I brought up,

00:51:40.060 | because their core business model

00:51:42.460 | is to throw their customers in a cage match.

00:51:45.380 | They don't have the culture internally

00:51:50.380 | to partner in a friendly win-win way.

00:51:53.540 | Every company I've ever had that sat down

00:51:55.940 | to do a deal with Google has been shown a contract

00:51:59.420 | that you would never, ever, ever, ever consider.

00:52:02.500 | And if we're going to get to a place

00:52:04.680 | where I just tell Google my favorite travel site is this,

00:52:08.180 | my whatever, they're going to have to work out a deal

00:52:11.420 | that's reasonable on both sides.

00:52:13.660 | And I don't think they're capable of it.

00:52:15.380 | I don't think they're capable of sitting down

00:52:18.100 | and it becomes a cultural problem

00:52:20.700 | because if every meeting you have

00:52:23.280 | with an external third party,

00:52:24.680 | you bring this kind of I'm in charge attitude,

00:52:28.500 | getting that out of a company is very, very, very hard.

00:52:32.420 | - Yeah, I mean, I see something else.

00:52:35.260 | - You don't have to agree with that.

00:52:36.380 | - No, I know that there's great anecdotes

00:52:39.820 | of that problem in tech, but at the end of the day,

00:52:44.820 | the entire world is advertising on Google.

00:52:47.580 | So they clearly have made,

00:52:48.900 | they clearly figured out how to partner

00:52:50.660 | with the CMO of every company on the planet.

00:52:53.260 | And so if our CMO got a call that said,

00:52:56.940 | "Hey, from Google, we can get you more customers

00:52:59.940 | when they ask a question about content management software."

00:53:03.460 | We're still going to say like,

00:53:05.060 | we're going to follow where the distribution is.

00:53:07.060 | So at the end of the day,

00:53:08.260 | and so then that's why I still think it's all,

00:53:10.620 | like the name of the game is still,

00:53:11.980 | can they navigate their customer base

00:53:14.100 | to a new user experience paradigm?

00:53:16.540 | Well, obviously fast enough before,

00:53:18.020 | you know, it's the classic can Amazon become a studio

00:53:21.180 | before the studios become Amazon or Apple or whatever.

00:53:23.300 | - I mean, the company with the buzz is perplexity,

00:53:26.060 | not even chat GPT.

00:53:28.260 | And so like barge not in the top two, right?

00:53:32.140 | - Yeah, but then you get into a different issue,

00:53:33.940 | which is like, obviously they should just buy these things.

00:53:36.580 | And then, you know, Lenacon doesn't let that happen.

00:53:38.980 | So, you know, how do you like,

00:53:42.100 | like if they bought perplexity, call it Google assistant,

00:53:45.060 | like, I don't think we'd be having this conversation

00:53:46.700 | in the same way, but they obviously, you know-

00:53:48.380 | - That's a fair point.

00:53:49.700 | - Yeah.

00:53:50.540 | - Which is like the Instagram WhatsApp.

00:53:54.140 | - I'm just saying that's like Instagram WhatsApp,

00:53:57.620 | what he's saying and what metadata.

00:53:59.820 | And so Google's like not allowed to do that right now,

00:54:02.620 | which is pretty interesting.

00:54:03.460 | - Yeah, it's totally crazy.

00:54:04.300 | I mean, it's crazy.

00:54:05.140 | - Which actually, one last thing,

00:54:06.980 | it's similar to Microsoft in the '96, '97 timeframe

00:54:11.980 | where they were, you know, in the penalty box

00:54:15.500 | and couldn't do those similar things.

00:54:17.500 | - Yeah.

00:54:18.340 | - I would say it's not only the government's not allowing,

00:54:20.340 | but you gotta remember with Instagram and WhatsApp,

00:54:22.780 | it was not, they were not facing an innovator's dilemma.

00:54:25.700 | Those deals did not cannibalize their core business

00:54:28.380 | in the way that perplexity would cannibalize,

00:54:30.580 | at least in part, the core business at Google.

00:54:32.620 | - Some of the searches.

00:54:33.660 | - Let me shift us forward.

00:54:35.020 | You know, you brought up, you know,

00:54:36.900 | a subject on our episode one, Aaron,

00:54:40.340 | we talked about, you know, these interesting investments

00:54:45.100 | made by one of my partners had called MANG.

00:54:48.020 | So investments by Microsoft, Amazon, Nvidia and Google

00:54:51.420 | into the big model businesses.

00:54:53.980 | Bill suggested, you know, that at a very minimum,

00:54:57.100 | that's low quality revenue.

00:54:59.380 | But, you know, this credit for investment thing

00:55:01.420 | as we've seen historically in the past,

00:55:03.900 | it wasn't lost on me that the very next day

00:55:06.780 | that the FTC came out, I think we have a tweet on this.

00:55:10.340 | It was on Squawk, you know, and said,

00:55:12.580 | "Hey, we're gonna look into the nature

00:55:14.180 | "of these relationships, you know,

00:55:16.680 | "between, you know, the open AIs and Microsoft's, et cetera."

00:55:21.100 | So Bill, just to close that one out

00:55:22.780 | before we jump into our third topic,

00:55:24.420 | you know, what do you think the outcome is of this inquiry

00:55:27.180 | that, you know, that the FTC announced, you know,

00:55:29.700 | which was right in line with some of the things

00:55:31.340 | that you were talking about?

00:55:33.220 | - My concern, which was really a concern for the industry

00:55:36.020 | more than just those players,

00:55:37.820 | is that those deals pervert and distort the market.

00:55:40.940 | And when that happens, people do things you wouldn't expect.

00:55:44.500 | Just like Zerp, all of a sudden,

00:55:46.460 | competition doesn't look like you would think it would

00:55:48.860 | because those are happening.

00:55:50.380 | And just real quick, the one thing that,

00:55:53.060 | I don't think Lena was,

00:55:54.780 | she never mentioned accounting or anything like that,

00:55:57.220 | but the bright lights, you know, being shined at it

00:56:00.460 | might prevent more of those deals in the future,

00:56:03.060 | which I think would be healthier for the environment.

00:56:06.020 | - The thing that I actually don't understand

00:56:07.220 | for you guys is, so if I'm a big tech company,

00:56:11.380 | I have an entirely different incentive structure

00:56:14.180 | for these rounds than, obviously, a venture capitalist,

00:56:17.460 | and yet they're pricing these companies.

00:56:20.060 | - Totally.

00:56:20.900 | - So that's the distortion effect

00:56:21.740 | for either those rounds directly

00:56:23.780 | or for everybody else in the space.

00:56:25.740 | It's like, yes, X company got a $20 billion valuation

00:56:28.900 | from a non-economic actor.

00:56:31.540 | That does not make your revenue or your multiple

00:56:34.460 | have any correlation to that

00:56:35.860 | because the acquisition premium

00:56:37.620 | or the ultimate cashflow of this company

00:56:40.060 | doesn't have any relationship to the credit model

00:56:43.060 | that just got a deal done.

00:56:44.540 | - I think it's even trickier

00:56:45.820 | because based on what I've read,

00:56:49.220 | secondary transactions are kind of expected now

00:56:54.220 | in the hiring, in that world, that LLM competition world.

00:56:59.100 | And if you can't get a financial investor

00:57:02.380 | to invest alongside, 'cause these are non-cash credits,

00:57:06.740 | then you don't have the cash to do the secondaries

00:57:09.660 | and you can't keep up.

00:57:11.220 | And then that's tricky.

00:57:12.420 | And on our last show, Brad talked about

00:57:16.180 | why these big companies might have an incentive

00:57:19.580 | that's other than ownership at the right price.

00:57:22.620 | So that, as you're saying, the valuation's not real.

00:57:26.220 | And there's all kinds of problems this can create,

00:57:30.140 | especially if you're running the company.

00:57:32.100 | - Yeah, I think, unfortunately, there's an inevitable,

00:57:35.700 | there's gonna be an inevitable,

00:57:37.380 | that the music will stop at some point

00:57:38.900 | on this particular dynamic.

00:57:41.820 | OpenAI aside, just because they actually

00:57:44.260 | do have the traction,

00:57:45.260 | they probably do have the real revenue.

00:57:47.260 | - Right.

00:57:48.100 | - But I don't love to see the broadening of this,

00:57:51.820 | just from a, again, healthiness of this model.

00:57:54.940 | - I mean, Bill referred to it a little bit,

00:57:57.420 | like the SoftBank effect from '20 and '21,

00:57:59.980 | where capital was used as a weapon of economic destruction,

00:58:03.380 | capital was the kingmaker and all of this,

00:58:05.620 | rather than allowing the product

00:58:07.260 | and the fundamental performance of the business to go.

00:58:09.860 | But I'm gonna just shift this a little bit, Aaron,

00:58:13.020 | into perhaps a little bit broader discussion about AI.

00:58:17.060 | You had a tweet that caught my eye,

00:58:20.100 | where you just said, this is,

00:58:21.860 | you've been around for a while,

00:58:23.380 | you're a deeply respected technologist.

00:58:26.580 | You said, hey, we got this confluence of events

00:58:28.660 | going on in the world right now,

00:58:30.020 | confluence of technologies that make this, in many ways,

00:58:32.980 | the most extraordinary of times.

00:58:35.180 | And so, as we dig into AI a little bit with you,

00:58:38.100 | maybe a little context for that tweet,

00:58:40.420 | and then I would love for you to go inside out.

00:58:42.700 | Like, where are you actually seeing the traction

00:58:46.500 | for AI in your business?

00:58:48.140 | What are your guys' top three priorities

00:58:50.260 | in terms of leveraging AI?

00:58:52.260 | And how do you think those reflect

00:58:53.700 | the priorities of your peers?

00:58:56.180 | - Sure, yeah, I mean, this tweet specifically

00:58:58.700 | was related to a bunch of stuff,

00:59:01.620 | including Vision Pro or whatnot.

00:59:03.060 | Just, I just think it's an incredibly exciting time,

00:59:06.660 | purely economics aside, to just be building technology.

00:59:10.500 | We are literally given these platforms to build on

00:59:13.820 | that are doing incredible things

00:59:15.220 | that quite literally a decade or two ago

00:59:17.460 | was just not possible.

00:59:19.700 | So, that's more of an emotional, psychological thing

00:59:22.420 | of just like, hey, incredibly exciting time.

00:59:24.620 | As it relates to the opportunities,

00:59:26.500 | and maybe then AI specifically,

00:59:29.740 | we are, so, maybe just like three seconds on Box,

00:59:35.660 | and then I'll broaden it.

00:59:36.620 | So, what we're doing is we have

00:59:39.500 | hundreds of billions of files that are stored in Box,

00:59:42.020 | and every single one of those files

00:59:45.180 | has more value inside of the file

00:59:47.420 | than what currently the customer

00:59:48.900 | is kind of getting benefit from,

00:59:50.460 | 'cause you have to like open up the file,

00:59:52.460 | look at it, read it, watch it,

00:59:54.540 | to get actual like real commercial economic benefit

00:59:57.740 | of that content.

00:59:58.900 | With AI, you now have little bots that can run around

01:00:01.380 | and do things on that content

01:00:02.700 | to generate more value for you.

01:00:04.940 | You could get a decision in your business faster,

01:00:06.980 | you could summarize a contract and accelerate a workflow,

01:00:10.660 | you could extract data from something that's unstructured

01:00:13.940 | to automate a process that was un-automatable before.

01:00:16.420 | So, that's why it's incredibly exciting for us.

01:00:19.300 | The dynamic right now is, I think we have,

01:00:24.780 | we are still, even though we're a year and a quarter

01:00:27.220 | into Chachapiti phenomenon,

01:00:29.060 | we're still literally in the earliest of days.

01:00:30.700 | I think most enterprises right now

01:00:32.220 | are just in the experimental period.

01:00:34.860 | They are trying to figure out where to plug AI

01:00:38.260 | into their overall stack.

01:00:40.580 | I don't think anybody has sort of fully figured that out,

01:00:43.740 | you know, commonly across kind of normal corporations.

01:00:46.940 | And that means that it's anybody's game right now

01:00:49.780 | in terms of the winners and losers of AI.

01:00:52.420 | I think, you know, I probably prefer incumbents in software

01:00:57.420 | on the margin right now,

01:00:59.740 | because it's so important to have the customer data

01:01:02.940 | in an environment that is already trusted,

01:01:05.060 | is already secure.

01:01:06.420 | They don't have to move the information back and forth.

01:01:09.340 | They often already have the workflow.

01:01:11.300 | So, if I'm thinking about who wins in CRM AI,

01:01:14.980 | it's to me Salesforce, as opposed to a startup.

01:01:17.420 | ITSM AI, it's ServiceNow versus a startup.

01:01:20.660 | That's like the quick thing,

01:01:22.340 | because there's not as much innovator's dilemma

01:01:25.180 | in the kind of pure software categories.

01:01:27.900 | In fact, this is just really like an advantage

01:01:30.060 | for anybody, again, who has users, data, and workflows.

01:01:33.500 | AI is like this dream come true,

01:01:35.460 | because now we can just literally offer more value

01:01:37.660 | to our customers.

01:01:38.660 | So, that's sort of like in the classic incumbent categories.

01:01:42.580 | I'm extremely bullish for startups.

01:01:44.300 | They're just not gonna probably be about disrupting,

01:01:46.860 | you know, known categories of software.

01:01:49.420 | There's probably much more about known categories

01:01:52.140 | of like the economy.

01:01:54.020 | And so, I think if we spent like the past 20 to 30 years

01:01:57.340 | putting software, making it so, you know,

01:02:00.460 | software is a layer above what humans do.

01:02:03.340 | Now we're at a point where software will just do

01:02:05.420 | what the human did.

01:02:06.740 | And that creates a whole new vector of opportunities.

01:02:10.580 | And I think, you know, the classic way to look at this

01:02:12.860 | is sort of like, you know,

01:02:14.180 | figure out which parts of knowledge work

01:02:16.060 | can convert into tokens.

01:02:17.780 | And then those are the areas

01:02:18.900 | where there's new software opportunities

01:02:21.140 | that we did not have software for before.

01:02:23.140 | And then those are the businesses.

01:02:24.540 | So, I think there's, you know,

01:02:25.780 | there'll be trillions of dollars made in AI.

01:02:28.500 | But I think it's gonna be in verticals.

01:02:32.340 | It's gonna be, you know, kind of helping augment

01:02:34.180 | the people element of the work.

01:02:36.220 | It's gonna be in the infrastructure

01:02:37.460 | and the scaffolding around it.

01:02:39.860 | I'm a little bit bearish, not bearish,

01:02:42.100 | but just like more like on the margin,

01:02:44.780 | not as excited about the economics

01:02:46.580 | of the actual pure AI model providers.

01:02:49.700 | I think it's tough to be in a space where Zuck,

01:02:52.260 | you know, at any day could just open source

01:02:54.540 | the thing that you've been working on

01:02:55.780 | for, you know, three or five years.

01:02:58.700 | And then all of a sudden, you know,

01:03:00.740 | now there's just this leapfrog moment of technology.

01:03:04.340 | And, you know, the thing that I go back to is like,

01:03:07.620 | there's rarely been technology I think we've seen

01:03:10.540 | where the actual, like the thing you're building,

01:03:13.140 | the asset you're building is almost perishable

01:03:15.700 | as a piece of IP.

01:03:17.300 | Like the thing literally like can just become obsolete

01:03:20.780 | like a second later, and you can't update it.

01:03:22.860 | You don't, like, I can't take the obsolete thing

01:03:26.100 | and make it a little better.

01:03:26.940 | I have to like rerun the training run

01:03:29.180 | and then like do the new model.

01:03:30.620 | And so I actually have like converted, you know,

01:03:32.940 | CapEx dollars into this like thing.

01:03:35.620 | And at any moment, something else could be better

01:03:37.660 | than that thing.

01:03:38.500 | And I have no, I can't pivot around that.

01:03:40.020 | Like I'm stuck with this perishable asset.

01:03:42.260 | So then you kind of just say, okay,

01:03:43.780 | well then OpenAI, Microsoft, Google, you know,

01:03:47.860 | Facebook are basically like where you place the bets

01:03:51.140 | on who makes the models.

01:03:52.580 | And then everybody else

01:03:53.420 | should just be building software essentially.

01:03:55.460 | - What about internally?

01:03:56.540 | I'm curious as someone that runs a large public company,

01:04:01.540 | how do you feel about like what's the right expectation

01:04:06.660 | for programmer productivity improvement?

01:04:09.180 | Are you measuring it?

01:04:10.180 | Do you care?

01:04:11.900 | Where in your org are you seeing impact early?

01:04:16.720 | - Yeah, so, you know, programming obviously

01:04:18.940 | the first place with Copilot.

01:04:20.860 | And I don't know, we have not done

01:04:23.460 | a specific measurement internally.

01:04:25.100 | You know, I'll walk, you know, through the office

01:04:27.220 | and see what's on people's screens.

01:04:29.420 | And, you know, AI is certainly actively being used

01:04:33.100 | for in Chachaputi to optimize a code, you know,

01:04:36.700 | the code that somebody is working on, you know,

01:04:39.740 | how do I change this, you know, SQL query

01:04:42.380 | to be more efficient, that kind of stuff.

01:04:44.500 | Copilot obviously for writing code.

01:04:46.500 | Obviously the estimates are, I would probably just agree

01:04:49.420 | with the estimates of, you know, 30, 40, 50%.

01:04:51.500 | - That's what I heard.

01:04:52.460 | Drunk said 7X yesterday.

01:04:54.900 | Anyway, move on.

01:04:56.300 | - Drunk said that?

01:04:58.420 | - I think he did.

01:04:59.260 | - Yeah, okay, I have not seen somebody say that.

01:05:01.740 | - I don't think that was accurate.

01:05:03.300 | - Okay, got it.

01:05:04.140 | - Yeah, I like the 70% number.

01:05:06.900 | - Yeah, so I'm probably more in the, you know,

01:05:09.300 | 30 to 50% camp in it, like in the best case scenario

01:05:12.500 | right now with where we're at.

01:05:13.820 | And I think still coding is the number one use case.

01:05:17.740 | And there's actually, I think this actually

01:05:19.540 | is an important element of thinking through AI.

01:05:22.220 | Like coding is still working better

01:05:24.460 | than most other use cases.

01:05:26.100 | And it's because you basically have a workflow

01:05:29.180 | where you're in a text interface that is just linear,

01:05:32.980 | where most of the sort of knowledge of the field

01:05:37.940 | is all public and open source for the most part

01:05:40.340 | and available for training runs.

01:05:41.980 | And the next line prediction, you again,

01:05:44.460 | have the perfect interface for next line prediction.

01:05:46.820 | Most knowledge work actually doesn't look like that.

01:05:49.220 | Like we want it to, we want to believe.

01:05:51.060 | - Timer structured.

01:05:51.900 | - Yeah, and so, you know, even the legal work,

01:05:55.820 | I'm still like talking to the client

01:05:58.140 | and then like comparing something with somebody else.

01:06:01.300 | And so this idea that you're gonna get a GitHub

01:06:05.140 | co-pilot effect for all forms of knowledge work,

01:06:08.260 | at least in the near term, it's way too early

01:06:10.380 | to be kind of jumping to that conclusion.

01:06:11.860 | - Programming is like a more precise language than language.

01:06:15.260 | - Yeah, and then literally it's automatically

01:06:17.100 | testable instantly.

01:06:18.780 | And so you just don't have those characteristics

01:06:20.660 | for a lot of other work.

01:06:22.220 | So I think it's gonna take longer than maybe we'd like

01:06:24.860 | for AI to sort of show up in the average knowledge workers

01:06:28.220 | sort of day at the level that GitHub co-pilot did.

01:06:32.060 | You know, if you could automatically write

01:06:33.580 | all of my emails, you know, faster,

01:06:35.700 | I still would have to read each one.

01:06:37.300 | I still have to process like,

01:06:38.900 | do I agree with that thing that is being said?

01:06:41.740 | I still need to understand the substance of the content

01:06:44.700 | from the sender.

01:06:46.340 | I can't have AI just sort of jump in and replace me.

01:06:49.540 | So I think we're probably a little bit early

01:06:51.780 | in the general knowledge work sort of transformation.

01:06:55.580 | And the things I'm much more excited about

01:06:57.340 | are where can you take a process

01:06:59.980 | where a person does that kind of rote task

01:07:02.900 | over and over again,

01:07:04.180 | and we can swap that out with AI and improve that workflow.

01:07:07.500 | And that's where the opportunity is

01:07:09.420 | and certainly what we're spending our time on.

01:07:11.140 | - I mean, it reminds me so much,

01:07:13.260 | I saw this chart this week,

01:07:14.500 | maybe we can pull up from Morgan Stanley.

01:07:16.620 | And it shows how we systematically underestimate

01:07:21.940 | the size of super cycles, right?

01:07:24.300 | Over a reasonable period of time.

01:07:26.460 | When I think about what's going on here,

01:07:28.260 | of course, they show,

01:07:29.500 | they went back and did some math against initial estimates

01:07:32.540 | and said, you know,

01:07:33.700 | super cycles are about 40% underestimated at the start.

01:07:38.340 | So, you know, the initial forecast for the PC

01:07:40.580 | or the initial forecast for internet 152 million users

01:07:43.980 | in 2000 ended up being 361 million users.

01:07:47.340 | When I think about what's gonna happen here,

01:07:50.660 | I can absolutely see sometime over the next four quarters,

01:07:54.820 | we're gonna hit a zone of disillusionment

01:07:57.420 | because everybody's pig piled in here.

01:07:59.500 | They think it's all happening now.

01:08:01.300 | It's gonna take a little bit longer,

01:08:02.820 | just like the internet did, just like cloud did, right?

01:08:05.860 | To hit its stride.

01:08:07.300 | When that happens, those who were, you know,

01:08:10.700 | who are always against the super cycle,

01:08:13.060 | call it a fad, everything else,

01:08:14.700 | they're gonna say, see, I told you so, you know,

01:08:17.060 | it's not really happening the way you all thought it was,

01:08:19.740 | you know, you made bad bets, et cetera.

01:08:21.940 | But my sense here is, you know, and I've said,

01:08:24.860 | the AI is gonna be bigger than the internet itself

01:08:26.860 | in terms of impact on economic productivity.

01:08:29.940 | But I do think sometime in the next several quarters,

01:08:32.620 | we have this zone of disillusionment set in a little bit.

01:08:35.740 | It doesn't happen quite as fast as we all thought.

01:08:38.900 | But then when we look back three to five years from now,

01:08:43.060 | you know, you had a tweet about, you know,

01:08:45.220 | the dramatic reduction in costs of these models.

01:08:49.180 | So you can imagine chat GPT five or six,

01:08:52.060 | and it's costing us 90, you know, percent less than today.

01:08:56.260 | I don't even think we can get our heads around

01:08:59.020 | how that's going to change everything.

01:09:01.660 | - Well, so most of, and just to underscore this,

01:09:04.620 | most, and again, we're like,

01:09:08.140 | we're deep in sort of the use cases of,

01:09:11.900 | I have a really long contract,

01:09:13.460 | I wanna read the contract,

01:09:15.220 | I wanna automatically present data

01:09:16.820 | from that contract with AI.

01:09:18.460 | And so think about like, just all of those things,

01:09:21.860 | you know, an invoice, a contract, a presentation,

01:09:25.420 | most of the things that customers are asking us for,

01:09:28.100 | what they want to do with their content

01:09:30.820 | is 100% possible.

01:09:33.580 | And just, there's a curve, which is,

01:09:36.180 | is AI cheap enough to make that use case

01:09:40.140 | be worth transforming to a software-based way

01:09:42.500 | of doing that versus human-based way.

01:09:44.540 | And so that's like exactly what you want in a market,

01:09:47.660 | because you know the costs are gonna come down,

01:09:50.060 | and then you'll reach that convergence point.

01:09:52.820 | So like, we're increasingly not limited

01:09:56.460 | by the architecture of the technology.

01:09:58.740 | And now we're just limited by like,

01:10:00.100 | can Jensen make these things cheaper?

01:10:02.540 | And can the, you know, can the engineers at OpenAI

01:10:05.300 | have more efficient model algorithms?

01:10:08.140 | And that's great,

01:10:09.340 | because then you can just ride that curve.

01:10:10.820 | Like, we are way more inundated with use cases

01:10:14.700 | that companies want to do with AI today

01:10:16.900 | than simply just like the scalability of these architectures

01:10:21.380 | from a cost standpoint.

01:10:23.100 | - Brad, I'll give you one other reason

01:10:24.460 | why I think you might be right

01:10:25.900 | about the trough of disillusionment.

01:10:27.860 | You know, I saw this great interview once with Spielberg

01:10:31.180 | when he was talking about JAWS.

01:10:33.060 | And he said, you know, I didn't show the shark

01:10:36.580 | until like 75% of the way through this thing.

01:10:39.980 | And the reason he said he did that

01:10:41.620 | was because human's imagination is much grander

01:10:45.020 | than anything I could possibly put on the screen.

01:10:47.900 | And I, you know, we had AI before LLMs.

01:10:52.220 | Like, I heard Nikesh bring this up once.

01:10:55.380 | And so, but in our modern lexicon,

01:10:57.780 | we think of AI as chat GPT and what the LLM did.

01:11:01.100 | That's what really shocked us.

01:11:03.260 | And I just worry that the one, two, three thing

01:11:07.380 | has caused people to think

01:11:08.780 | there's this linear extrapolation.

01:11:11.100 | I do think the text has kind of been solved

01:11:15.580 | and everything's kind of been scanned.

01:11:17.740 | And I think the next steps not linear are super linear.

01:11:22.740 | I think it's gonna be a little bit of a find.

01:11:25.940 | - Well, I think that, I mean, the growth of cloud

01:11:29.380 | as just a shape of the curve is somewhat instructive

01:11:32.060 | because it at least closely approximates

01:11:35.540 | like the change management dynamic

01:11:37.620 | that is required to change your infrastructure architecture.

01:11:42.340 | So obviously not a perfect analogy,

01:11:44.020 | but like anybody in '06, '07, when we saw AWS,

01:11:48.900 | we're like, oh shit, like this is clearly the future.

01:11:51.220 | I would literally, I would never wanna go

01:11:53.020 | and manage servers again.

01:11:54.620 | But now here we are, it's 2024

01:11:57.220 | and you have the biggest growth rates

01:11:59.940 | from a dollar standpoint still happening.

01:12:02.140 | And so why is it taking 18 years for that still to occur?

01:12:06.820 | It's because there's somebody in a data center

01:12:09.020 | in Minneapolis that still has to just move the data.

01:12:13.180 | They have to move the services into the cloud.

01:12:15.820 | And so the equivalent of this is we'll talk to a customer

01:12:18.700 | and they say, we'll show them the demo

01:12:21.620 | of AI can now read that invoice

01:12:24.980 | and do what the human was doing instantly

01:12:27.380 | and much more cost-effectively.

01:12:28.860 | But it's still gonna take a year or two

01:12:30.660 | to do the change management of the workflow itself

01:12:33.380 | to swap in AI for where the human was.

01:12:35.940 | And so now multiply that-

01:12:36.780 | - Technically, it should have been an XML file

01:12:39.060 | and not an invoice.

01:12:40.140 | - Yeah, exactly.

01:12:40.980 | Well, I believe me, I'm extremely,

01:12:43.380 | I'm glad that it's actually a document still.

01:12:46.780 | And we're still in business

01:12:49.020 | 'cause most of these things still are documents

01:12:50.780 | and not XML files.

01:12:51.620 | - Fair enough.

01:12:52.460 | - I still think the press, they watch the movie "Her",

01:12:56.820 | they expect chat GPT-7 to be this all-knowing,

01:13:01.820 | all-loving personality that they can talk to.

01:13:05.820 | And this does spin to one thing

01:13:08.380 | I would be hyper-optimistic about.

01:13:10.900 | I do think this memory issue is a big, big deal,

01:13:15.100 | at least on the consumer side.

01:13:16.900 | So none of the major contenders today

01:13:21.620 | can remember who you are

01:13:23.300 | because it would require them,

01:13:25.780 | and I mean, remember over five years, 10 years

01:13:28.260 | to really become a personal assistant kind of thing

01:13:31.260 | that "Her" represented.

01:13:32.660 | And it's because you'd have to redo the model

01:13:35.060 | for each human.

01:13:35.900 | And of course that makes no sense economically.

01:13:38.700 | So it's actually a huge structural problem.

01:13:41.180 | There are complaints in Reddit

01:13:43.780 | about character AI on this front.

01:13:45.820 | I think all of the people sitting atop these companies

01:13:49.500 | know that this is a huge breakthrough opportunity.

01:13:53.780 | And the first one that gets to it,

01:13:55.460 | I think could outrun whoever the contender is

01:13:58.620 | at that moment in time,

01:14:00.180 | because the things you could do,

01:14:02.300 | if it could remember everyone in your contact database,

01:14:06.180 | all your emails that you've done,

01:14:07.620 | like the thing, the personal productivity

01:14:10.300 | an individual could have,

01:14:12.020 | if this thing could be that way.

01:14:14.340 | And I think most of the press thinks

01:14:15.860 | it'll be that way tomorrow.

01:14:17.340 | And no one's really solved this problem yet.

01:14:19.420 | I think it's a huge opportunity.

01:14:21.460 | - Yeah, 100% right.

01:14:23.340 | - I'm gonna move us on to the fourth topic.

01:14:25.060 | I mean, we could, we have to do a whole show

01:14:26.980 | just on that one that we're just on,

01:14:29.140 | because it's so good.

01:14:30.700 | And I'm quite certain, Bill,

01:14:32.980 | that what you just talked about,

01:14:35.340 | a fiduciary, an agent, an assistant

01:14:37.740 | that knows me longitudinally,

01:14:39.460 | that knows everything about me,

01:14:40.620 | like my personal assistant,

01:14:42.420 | five years from now, this is cracked, right?

01:14:44.660 | The cost curve is gonna come down,

01:14:46.220 | the security curve, the confidence curve,

01:14:48.140 | all of those things are gonna intersect.

01:14:49.700 | And we're going to give a personal assistant

01:14:51.700 | in people's pocket to billions of people.

01:14:54.620 | And when we think about the broad implications

01:14:56.980 | to Apple, to Google, to Meta, et cetera,

01:14:59.380 | of this change, right?

01:15:01.100 | We haven't seen this scale of change in 25 years,

01:15:03.420 | but we'll come back to it.

01:15:04.300 | Because Aaron, you had a tweet

01:15:06.660 | that really caught Bill and I's imagination.

01:15:09.540 | You know, maybe we can pull it up on immigration.

01:15:12.780 | And you know, like the world is, you know, right now,

01:15:15.860 | I think appropriately objecting

01:15:17.820 | to the insanity of illegal immigration

01:15:20.260 | and what's going on at the Southern border

01:15:21.980 | and lots of fights and finger pointing

01:15:25.180 | as to why that's going on.

01:15:26.780 | But what was interesting about this

01:15:28.140 | is you took on the subject of legal immigration.

01:15:31.580 | And, you know, why don't you explain to us a little bit

01:15:34.820 | about what the tweet was about and about this chart?

01:15:37.300 | I saw it went a little bit parabolic

01:15:39.180 | in terms of Twitter.

01:15:42.140 | - Well, I mean, unfortunately,

01:15:43.740 | Elon wrapped it into the illegal immigration as well.

01:15:46.780 | So that was the only reason for the parabolic nature.

01:15:49.900 | But my point was-

01:15:52.380 | - 22 million views, yes.

01:15:54.780 | - Yeah, everybody really liked H1B immigration this day.

01:15:58.100 | Actually, the amount of racist responses I got

01:16:00.460 | was pretty impressive.

01:16:02.140 | So, I mean, it's so simple.

01:16:05.100 | It is just the biggest cell phone in economic history.

01:16:09.220 | We have the best environment

01:16:11.540 | for the smartest people on earth to contribute value.

01:16:16.540 | And we are literally doing everything we can

01:16:19.540 | to make sure they cannot work here, pay taxes here,

01:16:22.140 | and start their next company here.

01:16:23.820 | And so it's just the most offensively illogical thing

01:16:28.820 | that I think we have in the country.

01:16:31.220 | I put it above all other things that we do

01:16:34.460 | that is totally inane.

01:16:36.580 | And we are, you know, by the grace of God,

01:16:39.420 | Apple was created here, Google was created here,

01:16:42.020 | Microsoft was created here, Meadow was created here.

01:16:44.180 | It is not a given that in 20 years from now,

01:16:45.900 | those companies will be created here.

01:16:47.740 | So why not make sure, why not just like seal that

01:16:50.540 | and lock that in as a monopoly

01:16:52.580 | and get all of that talent here?

01:16:53.940 | And the thing that is crazy to me

01:16:55.980 | is that U.S. companies are currently employing

01:17:00.820 | all of the people we're talking about moving here.

01:17:02.860 | They're just employing them in a different country.

01:17:04.980 | And we have expanded dramatically in Poland.

01:17:08.460 | Facebook obviously expanded dramatically in India and UK.

01:17:12.100 | Like we are employing all of these people.

01:17:14.580 | It's just, we're not gonna get the local community benefits.

01:17:18.380 | We're not gonna get the local tax dollar benefits.

01:17:20.180 | And then we're certainly, when they go off

01:17:21.780 | and leave one of our companies,

01:17:23.340 | they're going to build a company

01:17:25.220 | that is in and native to their country.

01:17:28.300 | And then it's over.

01:17:29.460 | And we just have to pray

01:17:30.660 | that that is not the next Apple or Google or Microsoft.

01:17:33.260 | And otherwise you just flip the entire

01:17:35.460 | kind of economic environment.

01:17:36.580 | And so it's just very frustrating.

01:17:38.440 | This just happens to be a chart of H-1B demand

01:17:41.420 | and applications per year.

01:17:43.220 | It's actually not even that interesting of a graph

01:17:45.380 | because if you actually looked at the true demand,

01:17:47.620 | if you could measure how many people actually wanted

01:17:50.140 | and were capable and qualified to work in the U.S.,

01:17:52.620 | the number would be five times larger.

01:17:54.780 | And again, we are currently employing all of these people.

01:17:57.340 | This is not going to reduce U.S. jobs.

01:18:00.260 | They are employed.

01:18:01.100 | There's no way, the kind of responses I always get

01:18:04.120 | is somebody will randomly, some troll will be like,

01:18:06.820 | you're just trying to take U.S. jobs.

01:18:08.320 | It's like, no, no, no.

01:18:09.160 | Like no matter what this person on the other end

01:18:11.340 | is going to be employed by our company,

01:18:13.420 | they're just not going to be living in the U.S.

01:18:15.780 | and improving our economy.

01:18:17.740 | And so I think it's crazy.

01:18:19.300 | And I think that, I'm a part of a kind of a tech lobbying

01:18:23.300 | kind of group or whatever.

01:18:24.140 | We lobby about all the things all the time.

01:18:27.340 | And so we'll meet with Congress and talk to the senators

01:18:30.340 | and Congress people about this.

01:18:33.820 | And what's really sad is that on a one-on-one basis,

01:18:38.820 | everybody, right, left, everywhere in between

01:18:42.240 | agrees with the issue.

01:18:43.460 | And there's just always one thing that's holding it up.

01:18:46.960 | It's no, we have to solve the whole immigration package

01:18:49.920 | or we can't give this one,

01:18:52.720 | we can't give into this particular deal

01:18:54.720 | that we're working on.

01:18:55.840 | And so this one little number in a document,

01:18:58.680 | i.e. the quota that we have,

01:19:01.640 | you just change that one number

01:19:03.020 | and our whole problem goes away

01:19:04.740 | and nobody's willing to do it.

01:19:05.980 | Nobody's willing to just like sign up for that project

01:19:08.500 | in Congress to say, I'm going to do it.

01:19:10.700 | - Brad, do you have the chart of the actual grants

01:19:13.620 | that happens each year?

01:19:14.900 | It's flat, it's fixed.

01:19:17.580 | - It's basically 85,000 a year

01:19:19.060 | and it's been that way for 20 years.

01:19:20.780 | - 65K, that's H1B and then there's another form.

01:19:24.380 | And the process is remarkably bureaucratic.

01:19:27.020 | So that's another thing that probably would make

01:19:29.140 | the demand higher if it were,

01:19:31.560 | as we've proven on the Southern border,

01:19:33.360 | if you ease the pathway, you create more demand.

01:19:37.540 | Yeah, it's tough.

01:19:40.720 | I think the biggest thing that we probably lose

01:19:45.720 | by not increasing that is this future founder thing

01:19:49.760 | that you're talking about.

01:19:50.940 | And on one hand, it's kind of impossible to numerate

01:19:55.240 | like why that could be so huge.

01:19:57.820 | But when you look at the legacy of immigrant founded

01:20:00.760 | companies in the Bay Area and the head count

01:20:03.280 | that's now under those companies,

01:20:05.200 | it's almost impossible not to believe.

01:20:07.480 | And why wouldn't you want that aperture bigger?

01:20:11.200 | - Well, there's also the argument of okay,

01:20:15.920 | this is just a Silicon Valley thing.

01:20:17.120 | And I mean, clearly it's not like,

01:20:18.400 | so we all know it's not.

01:20:19.780 | But like, even if you wanted to prove that it's not,

01:20:21.680 | Silicon Valley, well, we're gonna literally run out

01:20:23.520 | of real estate, okay?

01:20:24.400 | So like, and room, we know that.

01:20:26.920 | So guess what?

01:20:27.740 | If I'm a freaking town in anywhere in the country,

01:20:30.520 | and I could just be like, what?

01:20:31.800 | I can create like 10,000 tech jobs if I just become,

01:20:35.200 | if I'm able to get this amazing talent

01:20:37.480 | and they're all gonna move here.

01:20:38.800 | Like, this is totally a country wide opportunity for us.

01:20:42.460 | And we're just sitting on this and not doing anything about

01:20:45.880 | and I think it's insane.

01:20:46.800 | So yeah.

01:20:48.080 | - You know, this really comes down to the question

01:20:50.520 | of whether we have the will, right?

01:20:52.960 | To increase the quota of legal immigrants.

01:20:56.920 | Again, we're talking about legal immigrants

01:20:58.840 | who companies want to hire in order to drive our economy.

01:21:02.800 | To your point, Aaron, we're effectively,

01:21:04.960 | you're shipping your GDP to Poland.

01:21:08.280 | That would be US GDP.

01:21:09.880 | Instead, it's the GDP of Poland

01:21:12.340 | because it's workers in Poland who are doing the work

01:21:14.940 | that you can't hire to be done here.

01:21:17.480 | And Bill, you make the point about,

01:21:19.400 | it's hard to enumerate what we won't have started,

01:21:23.400 | the counterfactual of the companies that won't be started.

01:21:26.360 | But I look at just the acceleration generally,

01:21:29.160 | the moment we live in, we just talked about,

01:21:31.680 | the acceleration around AI.

01:21:33.660 | This is not just about economic advantage.

01:21:36.560 | This is about national security.

01:21:38.780 | This is about national strategic advantage, right?

01:21:42.160 | That we have these things invented here, right?

01:21:45.400 | Again, we're lucky that Chad GPT was invented here,

01:21:49.400 | the open AI is here, that Anthropic is here,

01:21:52.580 | but that's not a given.

01:21:53.960 | And the best way, and this has been a source

01:21:56.540 | of our national strategic advantage

01:21:59.080 | for hundreds of years, right?

01:22:01.020 | And Bill pinged me this morning and he said,

01:22:04.240 | it reminds me of this speech that Reagan gave,

01:22:07.840 | his last speech in office.

01:22:09.660 | - To be fair, I found it on the Twitters, but keep going.

01:22:16.100 | - Maybe we'll play it here, play it here,

01:22:19.040 | and then we'll end with all three of us

01:22:21.560 | by just rapping in reaction to the video.

01:22:24.700 | - It is that lady who gives us our great

01:22:29.200 | and special place in the world.

01:22:31.520 | For it's the great life force of each generation

01:22:33.880 | of new Americans that guarantees that America's triumph

01:22:38.120 | shall continue unsurpassed into the next century and beyond.

01:22:41.700 | Other countries may seek to compete with us,

01:22:45.420 | but in one vital area as a beacon of freedom

01:22:48.320 | and opportunity that draws the people of the world,

01:22:51.800 | no country on earth comes close.

01:22:54.580 | This I believe is one of the most important sources

01:22:58.720 | of America's greatness.

01:23:00.720 | We lead the world because unique among nations,

01:23:04.720 | we draw our people, our strength from every country

01:23:08.440 | and every corner of the world.

01:23:10.640 | - You know, you can't say it better than that.

01:23:13.880 | And you know, you look at what it's given us, right?

01:23:18.380 | You know, I came of age in 1978

01:23:21.680 | when the Japanese were devastating our auto industry,

01:23:25.680 | where the view in the country was that our best days

01:23:28.800 | were behind us.

01:23:29.640 | I grew up in a place that people referred to

01:23:31.640 | as the Rust Belt.

01:23:32.860 | We didn't have really venture capital.

01:23:34.760 | We, you know, the innovations around technology

01:23:37.240 | were just starting.

01:23:38.640 | And, you know, and so I think about,

01:23:42.440 | we now have such great global advantage in technology.

01:23:46.880 | The only way, right, in which we give way on that

01:23:51.160 | is to shoot the golden goose.

01:23:53.200 | And part of the equation, you know,

01:23:55.600 | the most important part is the human talent

01:23:59.160 | and desire to start the next thing,

01:24:01.560 | to invent the next thing.

01:24:03.520 | And I think you pointed out something

01:24:05.680 | incredibly important, Aaron, which is yes,

01:24:09.080 | we need to solve the illegal immigrant crisis

01:24:11.720 | on the Southern border.

01:24:13.120 | But, you know, we really do need to focus

01:24:15.320 | on how we continue to be the place

01:24:17.640 | that all the best entrepreneurs on the planet,

01:24:19.600 | the best technologists on the planet

01:24:21.600 | wanna come start the next company.

01:24:24.360 | Well, Aaron, thanks for being here.

01:24:26.360 | You know, one of the reasons Bill and I wanted to,

01:24:29.040 | you know, to do a pod wasn't just to hear each other,

01:24:31.600 | but was really to tap into the people we admire,

01:24:34.560 | the big thinkers in Silicon Valley, you know,

01:24:37.160 | kicking around with them on occasion.

01:24:39.800 | So really appreciate you taking the time

01:24:41.480 | and doing this today.

01:24:43.560 | - I look forward to when those people join the pod,

01:24:45.560 | so I'll tune in for that one.

01:24:49.040 | - Thank you.

01:24:49.960 | - But thanks for having me.

01:24:51.000 | I appreciate it.

01:24:51.840 | And good luck on the charts.

01:24:54.640 | (upbeat music)

01:24:57.240 | ♪ VG ♪

01:25:02.240 | ♪ VG ♪

01:25:03.600 | ♪ VG ♪

01:25:04.440 | ♪ VG ♪

01:25:05.480 | ♪ Squared ♪

BG2 with Bill Gurley, Brad Gerstner & Aaron Levie | Software Valuations, Earnings, Immigration | E02

Chapters