Adobe, Airtable and ServiceNow’s CFOs on the Financial Value of AI | WSJ
00:00:00.000 | - So, as I mentioned in my intro there,
00:00:02.260 | we're gonna start with a bit of a discussion
00:00:03.560 | about kind of the commercial side of generative AI.
00:00:05.460 | All three of you are at companies
00:00:07.080 | where there is a very obvious kind of commercial component
00:00:09.640 | of this that is perhaps a lot more immediate
00:00:12.920 | than it is for some folks in the audience.
00:00:14.760 | So, Dan, maybe let's start with you.
00:00:16.760 | You know, Adobe moved quickly
00:00:17.880 | to launch its generative AI tool, Firefly.
00:00:20.560 | Talk to me about the scale
00:00:22.440 | of the business opportunity here for Adobe.
00:00:24.360 | And, you know, how much is this an opportunity
00:00:26.800 | for Adobe to drive additional revenue
00:00:28.640 | versus an expectation of your customers
00:00:32.200 | that there will be generative AI plugins
00:00:34.160 | to what Adobe offers?
00:00:35.560 | - Yeah, and I appreciate the question
00:00:37.080 | and the reference to Firefly.
00:00:38.400 | It's not even one year old.
00:00:40.120 | I think later this month it turns one.
00:00:43.440 | But I think the opportunity is massive.
00:00:45.080 | As you think about the three areas
00:00:46.760 | we as a company are focused on,
00:00:48.760 | unleashing creativity for all,
00:00:52.080 | accelerating document productivity,
00:00:54.640 | empowering digital businesses,
00:00:57.340 | all three are gonna see a significant,
00:00:59.320 | meaningful tailwind as a result
00:01:01.680 | of this technology inflection.
00:01:03.840 | And let me just take a step back
00:01:05.080 | and share a little bit of why
00:01:06.560 | I have so much conviction in it.
00:01:08.680 | As you look at major and technology inflections
00:01:11.080 | as they shape end markets,
00:01:12.840 | innovation goes up, economic growth goes up.
00:01:16.080 | Adobe's got a 40-year track record
00:01:18.720 | of being a key catalyst to many of the trends
00:01:21.800 | that shape the digital economy.
00:01:24.160 | And there's a difference between capitalizing on a trend
00:01:28.060 | and shaping, being a key catalyst of that trend.
00:01:32.180 | We've got a long track record of doing it.
00:01:34.340 | This inflection's gonna be no different.
00:01:36.780 | And as I think about the AI stack,
00:01:39.160 | there's data, there's models,
00:01:43.060 | and then there's the products and workflows
00:01:45.520 | that bring these technologies to life.
00:01:48.060 | The third of the three is the most relevant importance.
00:01:52.520 | That's where the magic happens with these technologies
00:01:55.440 | because it's only in the context
00:01:57.240 | of those products and workflows
00:01:59.220 | that this technology leads to real productivity enhancement.
00:02:03.800 | And so we're incredibly bullish about this technology
00:02:07.180 | and we fully intend to be a key catalyst
00:02:10.180 | in shaping our end markets
00:02:11.820 | and driving success for our customers.
00:02:13.980 | - So one of the questions for,
00:02:17.040 | I imagine for all three of you on this panel,
00:02:18.740 | but then we're gonna stay with you on this for a moment,
00:02:20.420 | is sort of like how to commercialize this work.
00:02:23.480 | If the layer that you're thinking about is like,
00:02:26.500 | what are the products that sit below the data
00:02:28.400 | in the language model,
00:02:29.640 | how do you kind of commercialize those?
00:02:32.120 | Adobe has these kind of generative AI credits
00:02:34.680 | that it's launched.
00:02:36.200 | So talk to me a little bit about how you kind of arrived
00:02:38.360 | at that as a decision.
00:02:39.880 | I think this is in the context of a lot of companies
00:02:43.760 | kind of scrutinizing how much they spend
00:02:45.360 | on software and services and so on.
00:02:47.120 | And so the commercialization piece I think is fascinating
00:02:49.640 | and pitching it at the right level.
00:02:51.340 | - Yeah, what's great about the way we're positioned
00:02:53.340 | is we get to help our customers with both top line growth
00:02:57.600 | and bottom line productivity.
00:02:59.640 | So there's a uniqueness of how we engage with our customers
00:03:02.700 | that leads to that simultaneous benefit,
00:03:05.340 | which we really like.
00:03:06.800 | Generative credits is one of a number of vectors
00:03:11.800 | that we're driving from a commercialization standpoint.
00:03:15.300 | As you think about the accessibility of our tools
00:03:18.320 | and products, making them more approachable and accessible.
00:03:22.280 | And as we think about customers engaging
00:03:25.140 | with the technology, time to success,
00:03:28.640 | how quickly they can be productive
00:03:30.640 | is an important component of long-term retention.
00:03:33.920 | So more people top of funnel
00:03:35.520 | through more accessible technology,
00:03:38.120 | deeper level of engagement leading to better retention.
00:03:41.960 | And then as we engage with custom models,
00:03:45.460 | adding more value to the relationships we already have.
00:03:49.520 | Generative credits is one of a number of vectors
00:03:52.880 | that are available to the company.
00:03:54.840 | And as we think about deep power users
00:03:57.480 | of our products and our technology,
00:03:59.920 | when you tokenize engagement,
00:04:02.560 | turn a text into an image or a video or audio,
00:04:06.960 | those power users are consuming the technology
00:04:13.080 | at a differential rate.
00:04:14.720 | This allows them, based on that enhanced productivity,
00:04:18.060 | to make sure that the value that they realize
00:04:22.100 | is commensurate with the economics.
00:04:23.940 | And so there's a usage-based consumption model
00:04:27.260 | that sits alongside the subscription
00:04:29.560 | that's tailored to the power users
00:04:31.740 | that gets tapped into as they derive value
00:04:34.460 | from the technology.
00:04:35.740 | - Got it, okay.
00:04:36.900 | So Junior, I'd love to bring you in.
00:04:38.700 | ServiceNow launched its generative AI offering,
00:04:42.340 | I think in September, is that right?
00:04:44.120 | - Yeah. - That's right.
00:04:44.940 | - So you just reported on kind of the first quarter
00:04:47.340 | since it launched.
00:04:48.180 | So let's just start with kind of what is it
00:04:50.540 | and what has the uptake been since it launched?
00:04:52.460 | - Yeah, so we've been actually,
00:04:54.320 | ServiceNow's been investing in AI for years and years.
00:04:56.660 | We actually had our first AI SKUs launched back in 2018.
00:05:00.940 | That was more predictive AI.
00:05:02.900 | And we saw pretty great uptick,
00:05:05.460 | both on pricing as well as penetration in our user base.
00:05:09.040 | So fast forward to September, we launched NowAssist,
00:05:12.420 | which is our generative AI SKUs.
00:05:14.700 | We launched them on September 30th,
00:05:17.540 | so one day left of the quarter.
00:05:19.260 | And we actually had several multi-million dollar deals
00:05:22.060 | just in that one day.
00:05:24.140 | Q4 was the first full quarter of the launch.
00:05:27.460 | And it was by far our best performing
00:05:29.940 | new product launch ever to date.
00:05:32.380 | And so the opportunity is vast.
00:05:36.080 | And it's all about really driving value for customers.
00:05:39.560 | What we've done is we've infused
00:05:41.460 | human-like conversational interfaces throughout the platform
00:05:45.460 | and it's about getting the users,
00:05:47.140 | whether they're employees or customers,
00:05:50.160 | the information they need much faster
00:05:52.920 | and more reliably and more accurate.
00:05:55.180 | And then it's about getting developers more productive.
00:05:58.140 | And so we're seeing an uptick across the board,
00:06:01.080 | across the platform portfolio.
00:06:04.460 | And I think Gartner estimates that
00:06:08.140 | three trillion is gonna be spent on AI in the next,
00:06:11.800 | until 2027, with a third of that going to generative AI.
00:06:16.280 | So the opportunity is here and it's now.
00:06:19.120 | And we're seeing lots of interest
00:06:21.540 | from customers across the board.
00:06:23.640 | And it's always about value add, right?
00:06:26.080 | And so how are customers gonna get to value more quickly?
00:06:29.920 | When we think about that,
00:06:31.400 | we try to leverage 10% of the value internally in pricing
00:06:36.440 | and give 90% of the value to customers.
00:06:38.720 | And so those conversations are extremely powerful
00:06:42.540 | and there's not one CFO or CEO that I talk to
00:06:47.540 | that is not really leaning in
00:06:50.900 | to what gen AI can do for their business
00:06:53.040 | because they believe, and I completely agree with them,
00:06:56.060 | that it's not only a productivity win
00:06:57.900 | and a cost savings win for them,
00:06:59.500 | but it's business model innovation, it's top-line growth.
00:07:02.860 | And so we're really excited
00:07:05.380 | about the use cases for ServiceNow.
00:07:10.380 | - How do you differentiate what ServiceNow offers?
00:07:13.820 | All of your peers are coming out with generative AI offering
00:07:17.980 | and I'm sure everyone in this room gets pitched
00:07:20.380 | something generative AI related, frankly, all the time,
00:07:23.500 | probably twice on Sundays.
00:07:24.820 | So how do you differentiate it and say,
00:07:28.180 | ours is the real deal?
00:07:30.220 | - Well, our strategy right now I think is unique
00:07:33.940 | in that we are focused on domain-specific
00:07:36.740 | large language models
00:07:37.780 | within the ServiceNow platform already.
00:07:40.580 | So it's customers or new customers
00:07:43.100 | that have already worked with the platform,
00:07:45.820 | understand the platform.
00:07:46.980 | And so if you're working on ServiceNow
00:07:48.740 | to have that incremental capability
00:07:51.060 | and that incremental value add,
00:07:53.180 | if you think about it, it's like what CFO is gonna say,
00:07:56.140 | I don't want AI, I just want that standard SKU.
00:07:59.260 | No, they want the power
00:08:00.660 | and it's within the ServiceNow platform.
00:08:02.660 | So all the great workflows,
00:08:03.980 | all the great abilities to scale your operations,
00:08:07.420 | it's already there and it's already in the business.
00:08:09.740 | And so other companies are doing the same
00:08:11.980 | within their platforms to supercharge.
00:08:15.140 | Back to your question initially,
00:08:16.780 | how much is us leaning in and believing in the top line
00:08:21.780 | or customers pulling for it?
00:08:24.940 | It's both, like customers are going to,
00:08:27.740 | if they're not asking for it already,
00:08:29.980 | they will be very, very quickly.
00:08:31.820 | So those companies and those platforms
00:08:35.100 | that are building it in
00:08:36.300 | have got to really pick up pace and do it faster.
00:08:39.020 | Okay, so Anne-Marie, I'd like to bring you in.
00:08:41.820 | Tell me a little bit about how Airtable
00:08:43.420 | is thinking about this from a commercial standpoint
00:08:45.060 | and then we'll move the conversation on
00:08:46.300 | to sort of generative AI for the internal finance function.
00:08:49.420 | Yeah, absolutely.
00:08:50.420 | So we're really excited about how to leverage GenAI
00:08:54.620 | within our existing platforms.
00:08:56.460 | We're a low-code platform.
00:08:58.220 | We already work with 80% of Fortune 100 companies
00:09:01.780 | to enable their knowledge workers, citizen developers,
00:09:04.860 | to create apps for their day-to-day work
00:09:06.980 | and their workflows.
00:09:08.060 | It's a really natural point to insert generative AI,
00:09:12.140 | which is very use case specific,
00:09:14.020 | integrated into the data and the workflow
00:09:16.540 | that already exists.
00:09:18.020 | And so it's basically making things faster,
00:09:20.100 | making things easier,
00:09:21.580 | allowing GenAI to use the context of all the data
00:09:25.660 | more so than a human can.
00:09:27.940 | And we're thinking of it as,
00:09:29.780 | the big kind of functional use cases for us
00:09:31.780 | are marketing, product, sales, operations.
00:09:35.780 | And so we think of Airtable as sort of the co-pilot
00:09:38.380 | for all those operations,
00:09:39.740 | just like our engineering teams are using GitHub co-pilot
00:09:43.180 | and developers are probably the first
00:09:45.740 | in really getting efficiencies on GenAI,
00:09:49.140 | but we see that happening across these other functions.
00:09:51.660 | And I can give you sort of like a internal
00:09:54.220 | and customer example of how that comes to life
00:09:56.340 | and our digital product team, as you can imagine,
00:09:58.940 | is most excited about this.
00:10:00.220 | And they've been ingesting all of our customer feedback
00:10:04.500 | using GenAI, which means reams of data from Salesforce,
00:10:08.300 | from Gong, from web.
00:10:10.340 | GenAI then categorizes it by topic,
00:10:13.860 | sentiment ranks it sort of minus five to plus five,
00:10:17.660 | puts it in prioritization of what will make
00:10:20.220 | the biggest impact for customers,
00:10:22.620 | literally sort of responds with its own ideas
00:10:25.780 | of what product could we create to answer that feedback.
00:10:30.460 | How would we describe it?
00:10:31.540 | What would we name it?
00:10:32.740 | Now, as you can imagine, it's really important
00:10:34.540 | to have humans in the loop
00:10:36.340 | on this sort of workflow automation.
00:10:38.700 | And so we believe strongly in like human creativity
00:10:41.540 | and judgment, it has to be there.
00:10:43.180 | And then GenAI can be there for enhancing scale
00:10:46.580 | and just sort of machine scale context gathering.
00:10:49.940 | And then our product team literally then attaches that
00:10:51.980 | into the marketing workflow where they're generating
00:10:54.340 | messaging and ad copy emails that get better open rates
00:10:59.340 | and conversion.
00:11:00.540 | So that's our internal enthusiasm on our dogfooding.
00:11:04.060 | Externally, we have one of our really large cloud customers.
00:11:07.620 | They already use us for campaign management
00:11:09.980 | on the marketing side.
00:11:11.300 | They're now using our GenAI pilot to generate ad copy,
00:11:16.300 | depending on what the creative brief is,
00:11:17.940 | what past performance is.
00:11:19.580 | They're finding existing marketing assets
00:11:22.260 | that otherwise wouldn't have been tagged
00:11:23.980 | to this specific initiative.
00:11:25.820 | So it reduces content waste.
00:11:27.940 | And so I sort of have been thinking of it
00:11:29.340 | as this really exciting, like win, win, win, win.
00:11:32.220 | You save time, you save cost, you improve revenues
00:11:35.180 | by faster speed to market, better conversion.
00:11:38.700 | And then really powerfully,
00:11:40.420 | our employees just seem happier
00:11:42.060 | 'cause you're taking away the tedious tasks
00:11:44.140 | and the manual tasks, giving them more powerful content.
00:11:47.380 | And then their creativity and judgment
00:11:49.020 | is still the primary focus.
00:11:51.260 | And I think it improves retention and satisfaction.
00:11:54.020 | And for engineering teams, that's a really big deal
00:11:55.900 | as a CFO to see engineers happy as a software company.
00:11:59.300 | - Sure. - It's been really exciting.
00:12:00.780 | - Sure, so we'll come back to the human in the loop piece
00:12:02.860 | 'cause I do want to kind of discuss that
00:12:04.700 | in a bit more detail.
00:12:05.620 | But first, I'd like to focus a little bit
00:12:07.300 | on sort of generative AI
00:12:08.340 | for the finance function specifically.
00:12:10.060 | And so Gina, I'd like to start with you.
00:12:12.620 | You told me on our prep call last week
00:12:14.740 | that you're especially excited about using generative AI
00:12:17.580 | for sort of revenue recognition.
00:12:19.340 | That was one of the examples that we discussed.
00:12:21.140 | So talk to me a little bit about that.
00:12:22.900 | Is that something that you're currently piloting?
00:12:24.980 | And if so, what have the results been so far?
00:12:27.380 | - Yeah, so I think there's numerous use cases within finance
00:12:31.380 | and we're piloting a whole bunch of them.
00:12:32.820 | We had a hackathon and we picked like the top 10 areas
00:12:37.260 | that we want to focus and we're going with them first,
00:12:39.540 | but we have a list of use cases.
00:12:43.180 | P&A is a big area, automated forecasting.
00:12:47.180 | How do you forecast like a top line
00:12:51.460 | or a bottoms up and what's better?
00:12:53.060 | So we're doing a ton of that.
00:12:54.380 | But the revenue recognition piece is super interesting.
00:12:57.340 | If you think about it,
00:12:58.740 | it's all based on complex contract terms, right?
00:13:01.980 | And so the sales organizations ping finance
00:13:05.180 | between 3,000 and five times a quarter
00:13:07.900 | on just trying to understand the best way
00:13:10.020 | to formulate the contract to get the right rev rec.
00:13:13.100 | And how do we use Gen AI to automatically answer?
00:13:18.180 | So many of them are similar questions
00:13:20.540 | and how do we automate that?
00:13:22.100 | And by the way, not only is it automation time
00:13:24.620 | from a finance perspective,
00:13:25.820 | but then my sales teams are spending less time
00:13:28.740 | figuring out how to draft the contract
00:13:32.340 | and spending much more time with customers.
00:13:34.380 | So I think it's not only a productivity perspective,
00:13:36.980 | but also how do I get my sales people
00:13:39.340 | spending more time with customers?
00:13:40.700 | It's gonna be a much better customer engagement.
00:13:44.260 | And so twofold is an area.
00:13:46.060 | We've been piloting it and it's going extremely well.
00:13:49.180 | The finance teams love it because again,
00:13:51.580 | to Amberine's point,
00:13:52.940 | it's about automating the rotes and the routine
00:13:56.740 | and really giving time back
00:13:58.260 | for much more complex, interesting work.
00:14:00.980 | And we're having lots of engagement across the board
00:14:04.220 | from our finance teams and really leaning in
00:14:06.740 | to how Gen AI is gonna simplify what they do every day.
00:14:10.740 | And they're so excited about it.
00:14:13.180 | - And are there any concerns about sort of the accuracy,
00:14:17.460 | the accuracy of a use case like that?
00:14:20.260 | Let's say whatever model you're using is 95% accurate.
00:14:24.300 | We talked about having a human in the loop,
00:14:25.860 | but what has the pilot revealed in that regard?
00:14:29.420 | And what kind of safeguards have you put in place
00:14:32.500 | to make sure you're capturing that 5% of concern?
00:14:35.940 | - Well, I think the point is that the human interface
00:14:38.620 | is super important, right?
00:14:39.940 | There's always gonna be a check on the backend
00:14:42.660 | as to does this contract term meet what it needs?
00:14:46.500 | And so there's going to be that review,
00:14:48.860 | but it's that interaction,
00:14:50.180 | that back and forth email exchange.
00:14:52.740 | Do you understand getting on the phone?
00:14:54.180 | Do you understand that?
00:14:55.300 | That's able to really be automated.
00:14:56.900 | So you're 100% right.
00:14:58.260 | Accuracy is really important,
00:15:00.620 | but that human interface and interaction
00:15:03.620 | is extremely important.
00:15:05.140 | And by the way, CFOs,
00:15:06.740 | I don't know if you've gotten the rep
00:15:08.780 | in your rep letters yet.
00:15:09.860 | A lot of the big four accounting firms are repping,
00:15:13.820 | asking you to rep to make sure that if you're using AI,
00:15:16.940 | that there's that backend kind of closed loop around it.
00:15:20.100 | So really important, it's a great point,
00:15:22.820 | but it's not to say that we shouldn't be afraid
00:15:25.140 | of using AI because of that.
00:15:27.140 | You just need to manage around it,
00:15:28.420 | just like we've managed around other technology evolutions
00:15:34.220 | in the years that we've been CFOs.
00:15:36.620 | - Sure.
00:15:37.780 | Dan, similarly, you held a kind of a finance team hackathon,
00:15:41.060 | which I think you said you generated about 100 ideas
00:15:43.340 | of sort of specific use cases.
00:15:45.660 | As I understand it,
00:15:46.500 | you're moving forward with five pilots, is that right?
00:15:49.820 | So I'd like to focus on one today,
00:15:51.020 | which was, you've called it kind of the forecasting engine.
00:15:54.660 | So I think this touches a little bit on Gina's point
00:15:56.500 | about sort of FP&A and so on.
00:15:57.860 | So tell me again, how will this work?
00:16:00.020 | And ultimately, what is the sort of end goal
00:16:02.020 | that you think is attainable
00:16:04.220 | and how far away are you from reaching that?
00:16:06.220 | - Yeah, and before I go into the detail,
00:16:08.300 | just about the hackathon, Gina was talking about it,
00:16:10.700 | I was talking about it.
00:16:11.580 | We were commenting backstage on the enthusiasm
00:16:14.660 | that we see in our organizations
00:16:16.900 | around this technology inflection.
00:16:19.420 | Part of transformation is about change management
00:16:23.820 | and they're inextricably linked.
00:16:25.860 | How do you build that groundswell of enthusiasm
00:16:28.860 | within the organizations
00:16:30.380 | to lean into this technology inflection?
00:16:33.420 | You bring them into the process
00:16:35.100 | and you help them architect the solution
00:16:38.580 | and the way of working.
00:16:39.900 | So a hackathon is a really great way
00:16:41.820 | to build that groundswell.
00:16:43.700 | Yeah, we had over 100 submissions.
00:16:47.020 | We boiled it down to 20.
00:16:49.300 | Five of them were prioritized,
00:16:51.500 | but now we have a living, breathing pipeline
00:16:53.900 | of future opportunities
00:16:55.980 | that we're gonna continue to focus on.
00:16:58.020 | So one of them is a forecasting engine.
00:16:59.900 | So like most companies,
00:17:01.540 | the annual planning process is a heavy lift
00:17:04.220 | inside of the company
00:17:05.460 | and we are in the process of pivoting
00:17:07.540 | to a more fluid environment,
00:17:09.460 | a rolling forecast environment.
00:17:11.860 | But the way we communicate with investors in the street,
00:17:14.940 | that annual process is gonna be a thing.
00:17:17.460 | Now the question is,
00:17:18.460 | is how do you pivot to a rolling forecast
00:17:20.940 | and still maintain discipline from an annual standpoint?
00:17:24.940 | You communicate with investors.
00:17:26.860 | When you synthesize all of the data
00:17:28.860 | that sits inside of the company,
00:17:30.980 | you weigh it appropriately,
00:17:32.860 | you backtest it to make sure that the model
00:17:36.060 | and the algorithm that you tune
00:17:37.860 | accurately reflects the performance of the company,
00:17:41.100 | you can get to a rough cut annual plan
00:17:45.300 | very, very quickly.
00:17:46.980 | I mean, this is a month and a half process
00:17:49.460 | that you can start to access
00:17:51.060 | a first rough cut approximation within minutes.
00:17:55.700 | We're early in the process.
00:17:57.740 | We did a, let's see what this spits out
00:18:01.140 | from a prior year
00:18:03.940 | and it was probably 85, 90%
00:18:08.580 | of what the teams ground through
00:18:11.300 | and it was done within minutes.
00:18:13.420 | So it's a really powerful tool
00:18:15.180 | to sit alongside the planning process
00:18:17.980 | and the value in my view of a financial organization
00:18:22.220 | isn't about rolling up numbers.
00:18:24.660 | That's obviously important.
00:18:26.180 | Integrity of the numbers is super important
00:18:28.940 | but when do you see a signal
00:18:31.820 | and what's the insight from that signal?
00:18:34.180 | When do you take action based on that insight
00:18:39.740 | and how long does it take you to drive impact?
00:18:42.660 | There is a speed of execution
00:18:44.900 | that comes with this technology
00:18:47.220 | because the time to signal and time to insight
00:18:50.340 | is gonna get sped up dramatically
00:18:52.380 | and then we take the great people
00:18:54.660 | that call our company home
00:18:56.540 | and turn them on to an action-oriented,
00:18:59.180 | solution-oriented mindset to solve business problems
00:19:03.540 | and having that forecasting engine
00:19:06.380 | sit alongside an annual planning process
00:19:09.340 | is an important aspect of focusing on
00:19:13.260 | if we do nothing, this is the likely outcome.
00:19:16.740 | Now, what are the do different leads
00:19:18.420 | to get a different outcome?
00:19:20.100 | You transition to solution space very quickly
00:19:22.820 | and it's not just an annual planning process.
00:19:25.540 | We've got these predictive engines
00:19:27.220 | running alongside intra-quarter performance
00:19:30.180 | and it's fluid.
00:19:32.780 | I get input from the sales team
00:19:34.660 | on what they think they're gonna do
00:19:36.420 | and I've got my predictive engine
00:19:38.980 | and it's like, well, we disagree on that
00:19:41.180 | and we disagree on that
00:19:42.380 | and what actions are we gonna take
00:19:44.580 | to shape the outcome,
00:19:45.740 | to get where we need to be for the quarter
00:19:47.900 | and then we take a look at it at the end of the quarter
00:19:49.980 | and the accuracy embedded in this,
00:19:51.780 | if it's done well, is pretty impressive.
00:19:55.260 | - How does that change your,
00:19:56.780 | the skillset required of you as the CFO?
00:20:00.540 | If you have this information,
00:20:03.420 | how does that change the way that you then
00:20:05.740 | pass that information out to the rest of the C-suite,
00:20:08.020 | the rest of the company,
00:20:09.420 | use that to potentially influence decision-making
00:20:11.580 | and so on?
00:20:12.420 | Essentially, is this a kind of a powerful
00:20:14.940 | new decision-making tool that you have,
00:20:17.340 | you're the kind of C-suite owner of
00:20:19.060 | and so how do you cascade that out?
00:20:21.100 | I guess that's a question for all three of you, honestly,
00:20:22.660 | but let's start with you, Dan.
00:20:24.100 | - Yeah, so from my point of view,
00:20:25.420 | everybody's gonna have their own philosophy
00:20:28.020 | on how to do this role and do it well.
00:20:30.900 | From my perspective,
00:20:32.500 | I think the role of finance CFO,
00:20:35.020 | sort through complexity,
00:20:36.260 | get the core underlying root cause drivers,
00:20:38.780 | frame the debate and dialogue,
00:20:40.940 | frame the decision-making process,
00:20:43.900 | sharpen business decision-making
00:20:46.020 | and once decisions are made,
00:20:47.260 | go drive execution and drive impact inside of the company.
00:20:51.180 | And so this is a key tool that gets us, again,
00:20:53.860 | to that signal and insights faster.
00:20:56.140 | But there's a democratization of data.
00:20:58.260 | I don't think this sits in a finance silo.
00:21:02.260 | It's how you permeate this information
00:21:04.900 | to help shape that discussion,
00:21:06.820 | real-time business discussion inside of the company
00:21:09.980 | to get to better business decisions
00:21:12.620 | and better business outcomes.
00:21:14.860 | It's pervasive.
00:21:15.860 | - I would just add that the role of the CFO
00:21:18.340 | has been evolving for years now, right?
00:21:20.260 | And this is just one area in addition.
00:21:23.700 | So the point that Dan was making,
00:21:25.540 | it's not about just having a forecast number quicker,
00:21:29.580 | which is super powerful.
00:21:30.980 | It's what you then do with the time that you have,
00:21:33.500 | that information earlier.
00:21:34.980 | You know, I talk about culture of finance.
00:21:37.580 | I want the finance organization
00:21:39.540 | to be the first call of the business.
00:21:41.820 | It's not just about compiling the data
00:21:44.220 | and reporting out on it.
00:21:45.580 | It's about really driving impact and driving,
00:21:49.220 | okay, here's that signal.
00:21:51.420 | This is what we need to be doing.
00:21:52.740 | And by the way, we've been doing this for a while now, right?
00:21:55.020 | So we have, Dan, early insights into a quarter.
00:21:57.980 | So if things in certain areas
00:21:59.820 | look like that they're not shaping up the way we thought,
00:22:03.460 | get in there early.
00:22:04.660 | And it's not just about telling the salesperson
00:22:07.700 | or telling the marketing person.
00:22:09.180 | It's about working with them jointly to go solve it,
00:22:12.980 | thinking really throughly,
00:22:14.060 | thinking strategically about
00:22:15.980 | what this information is able to drive.
00:22:18.900 | And so I think the role is completely evolving.
00:22:21.460 | And we are probably the only person in the C-suite,
00:22:25.100 | besides the CEO,
00:22:26.580 | who has that bird's eye vantage across the enterprise
00:22:30.580 | that can really help drive impact in any area.
00:22:35.580 | And so, you know, oftentimes back in the day,
00:22:38.820 | I'd get my hands slapped to say,
00:22:40.620 | that's not finance, right?
00:22:42.220 | That's not finance.
00:22:43.060 | This is mine.
00:22:44.220 | Now, they're like, Gina, bring it, right?
00:22:47.380 | They're like, they want you involved
00:22:48.620 | 'cause they know that you have the information
00:22:49.940 | that's gonna help them solve their problems
00:22:52.140 | and get to the better results.
00:22:53.380 | And so that role has been evolving for a while.
00:22:55.820 | And Gen AI is only gonna continue that evolution.
00:22:59.540 | - Sure, and Marie?
00:23:00.900 | - Yeah, exactly what Gina said.
00:23:02.420 | I mean, we are in such a privileged position
00:23:05.020 | where we see absolutely everything.
00:23:07.980 | And I think that gives a lot of power,
00:23:09.860 | also a lot of responsibility.
00:23:11.660 | What I also think is super helpful
00:23:13.260 | is every one of my colleagues knows
00:23:16.620 | that I have relatively the most pure intentions
00:23:19.940 | of doing the best thing for the business
00:23:21.580 | and always thinking about the business and the customers.
00:23:24.460 | And so while different functional groups,
00:23:26.620 | depending on the quarter, depending on the year,
00:23:28.140 | have different priorities
00:23:29.180 | where they're very focused on one thing, as finance,
00:23:32.180 | we're focused on the kind of long-term success
00:23:34.940 | of the business as a whole.
00:23:36.500 | And I think that just adds to the power of the voice,
00:23:38.660 | to the direction we give.
00:23:41.100 | And then in terms of your earlier question
00:23:43.380 | of use cases within finance,
00:23:45.540 | we have a lot of parallel experiences,
00:23:48.500 | revenue recognition, forecasting.
00:23:50.540 | I'd say one of my teams,
00:23:52.260 | which has become a bit of a hero and usually isn't,
00:23:55.140 | is our accounts receivable team
00:23:56.740 | because we've been able to use Gen AI
00:23:58.900 | in a lot of different ways for them.
00:24:00.420 | So content generation in terms of emails and follow-ups,
00:24:04.540 | chat in terms of interfaces with salespeople and customers,
00:24:08.660 | as well as just forecasting on cash collection times,
00:24:11.540 | prioritizing the risk of customers.
00:24:13.820 | And they've been really excited
00:24:15.220 | and it's just been nice to see a team
00:24:16.820 | that usually doesn't see the limelight
00:24:18.820 | pop up through this technology.
00:24:21.100 | - Couldn't agree more.
00:24:21.940 | It's nice to hear that our use cases are very aligned.
00:24:24.780 | - Yeah, I would say, just on your point, Amber,
00:24:28.700 | we had a CIO network summit a couple of weeks ago
00:24:31.060 | and the CIO of Cisco was on one of the panels
00:24:33.580 | and he shared a similar anecdote to yours,
00:24:35.580 | but from an HR context where he said
00:24:37.940 | the response to recruitment emails that they send out
00:24:40.900 | is significantly higher when they're generated
00:24:44.380 | by a generative AI tool as opposed to by a human
00:24:47.100 | because it sounds counterintuitive,
00:24:49.780 | but it's more personalized using the generative AI tool.
00:24:52.340 | - It's not cut and paste, right?
00:24:53.740 | - Yeah, yeah.
00:24:55.140 | But I'm curious, for this panel,
00:24:56.740 | with some of these use cases that we're talking about
00:24:58.300 | with finance specifically,
00:24:59.580 | right now, generative AI is quite expensive,
00:25:04.740 | whether it's the commercialization of the models,
00:25:08.020 | the licensing, enterprise license for the model itself,
00:25:11.180 | whether it's the data that is required to be fed into that,
00:25:13.660 | whether it's the human time required on the front end
00:25:17.020 | to get these things up and running.
00:25:18.740 | So how are you thinking about the cost
00:25:21.220 | of some of these use cases that you're thinking about
00:25:23.100 | internally and how long they'll take to pay off?
00:25:26.660 | The value of what you've just described
00:25:28.340 | is clearly transformational in theory.
00:25:31.660 | Are you there yet?
00:25:32.700 | And how do you weigh cost versus return?
00:25:35.700 | - I'll start there because I think
00:25:37.180 | it's a really important conversation.
00:25:39.940 | A lot of CFOs are all about ROI
00:25:43.660 | and it's super important, right?
00:25:45.060 | And it's about how quickly am I gonna get the productivity
00:25:47.900 | and what does it mean for cost savings?
00:25:51.500 | I think this is a whole different ballgame.
00:25:53.580 | And I think you've got to invest upfront
00:25:56.900 | because if you don't, you're gonna be left behind.
00:25:59.980 | And how I think about it,
00:26:02.140 | and it's really hard to kind of pinpoint attrition,
00:26:04.820 | for example, but we go back
00:26:06.340 | to the employee engagement perspective,
00:26:08.860 | and this is not just in finance.
00:26:10.180 | This is across the board in the organization.
00:26:12.900 | But when employees are doing more creative,
00:26:16.180 | more impactful work, they're much more engaged,
00:26:18.700 | which means attrition is lower and you get to,
00:26:21.660 | which is expensive from a cost perspective.
00:26:24.060 | But even more importantly than the cost,
00:26:26.180 | it's keeping that brain trust internal and inside.
00:26:29.940 | That is invaluable.
00:26:31.860 | So you've gotta be able to take the numbers,
00:26:34.420 | but also take the opportunity cost of not doing it
00:26:38.060 | into account when you're really thinking ROI.
00:26:40.540 | And that's across the board, whether it's a finance
00:26:43.220 | or whether I'm selling to a customer
00:26:45.100 | when I'm talking to customers.
00:26:46.340 | It's about the value add and the productivity,
00:26:48.580 | but it's also about what happens if I don't invest today?
00:26:52.660 | One of the first acquisitions I greenlit
00:26:54.780 | when I joined in 2020 was a company called Element AI.
00:26:58.740 | And it was about hiring talent in the AI space
00:27:02.700 | that didn't really have great ROI in the initial years.
00:27:07.700 | But at the end of the day, infusing AI into the platform
00:27:12.020 | was gonna be more and more important as we went on.
00:27:14.900 | And so we greenlit it, and it's one of the reasons
00:27:17.660 | why we were able to be first to market
00:27:19.260 | with actual Gen AI product in market.
00:27:22.020 | And so as CFOs, you need to be a little bit more open today
00:27:27.020 | I think than maybe in times past
00:27:28.980 | of really investing in front of the curve.
00:27:31.540 | You need to be smart.
00:27:33.460 | It's about taking balance risks
00:27:35.500 | and saying yes to a few things potentially,
00:27:39.620 | but really making sure that you're not just hamstrung
00:27:43.140 | by the numbers only.
00:27:45.500 | Because I think in this day and age,
00:27:47.540 | not investing is really gonna put you in trouble
00:27:51.180 | for the mid and long-term.
00:27:53.420 | - Anne-Marie?
00:27:54.340 | - Yeah, so you mentioned this earlier, Ben.
00:27:56.460 | Over the course of the last few years,
00:27:58.060 | we had a lot of tool sprawl
00:27:59.700 | and way too much technology spend.
00:28:02.140 | So we have a tech council, and they see their role
00:28:05.020 | as primarily figuring out how to consolidate tools
00:28:08.380 | and reduce spend.
00:28:09.620 | We didn't wanna put AI into that council
00:28:11.860 | because the mandate of that council
00:28:13.260 | is very focused on cost savings.
00:28:15.820 | And so we created a separate AI council,
00:28:18.500 | which is kind of mandated to go find
00:28:20.460 | the most interesting value-added use cases
00:28:23.260 | and sort of on the four wins philosophy.
00:28:25.860 | We want use cases where we get all four of those things,
00:28:27.980 | like time-saving, cost-saving, revenue improvement,
00:28:30.940 | morale, inspiration for the employees.
00:28:33.260 | And that AI council is looking at the world differently
00:28:36.700 | and surfacing different use cases.
00:28:38.780 | And I review those every kind of week or two weeks.
00:28:42.020 | And I'll often come in and be like,
00:28:43.540 | "Why did we say no to this?"
00:28:44.860 | And the answer will be like,
00:28:45.700 | "We weren't sure about the cost versus value."
00:28:47.900 | And then we'll talk through,
00:28:49.460 | "Are we underestimating the value
00:28:51.020 | "to make sure that we push on that use case?"
00:28:54.260 | I think when it comes to commercialization
00:28:56.140 | as buyers of software, we're seeing,
00:28:58.820 | there's a lot of evolution.
00:29:00.020 | Like we're seeing some vendors price at zero,
00:29:03.180 | others price at like 80% of their base product.
00:29:06.380 | And I think everyone's trying to figure out
00:29:08.180 | exactly where the value equation shakes out.
00:29:11.300 | What will the take rate be at different pricing?
00:29:13.580 | So I think there's a lot to be learned
00:29:16.220 | about the evolution of what will this cost us over time.
00:29:19.780 | - Got it.
00:29:20.620 | - It's an interesting point though,
00:29:21.460 | what you bring up.
00:29:22.500 | And we're hearing this a lot because we get the question,
00:29:25.420 | I'm sure you do Jen as well,
00:29:27.140 | of how are IT budgets being allocated
00:29:32.140 | to Gen AI versus the other pieces?
00:29:34.780 | And while you're right,
00:29:36.260 | they're definitely focused on cost takeout
00:29:39.820 | because of the application proliferation that we've seen.
00:29:43.700 | It's one of the reasons why you're seeing
00:29:45.300 | a lot more platform consolidation.
00:29:47.300 | Like instead of having best point solutions
00:29:49.540 | in different areas, which by the way is very costly
00:29:52.420 | and from a security perspective, a nightmare,
00:29:55.100 | to platform consolidation,
00:29:56.860 | which really enables, I think,
00:29:58.860 | an interesting way of how CIOs are thinking
00:30:04.860 | about their budgets and spend.
00:30:06.300 | It's like you're being mindful of costs here,
00:30:09.900 | but taking some of that savings
00:30:11.620 | and allocating it to the Gen AI
00:30:13.660 | and the modernization and the IT of the future.
00:30:17.100 | And I think that's what we're seeing
00:30:18.740 | when we're talking to customers.
00:30:20.340 | I don't know, Dan, if you're seeing the same.
00:30:22.300 | - Yeah, 100%.
00:30:23.460 | I agree with everything Gina and Amberine just said,
00:30:26.980 | and I'll augment it, build on it with two comments.
00:30:30.980 | Change is coming.
00:30:33.020 | We have a choice.
00:30:34.980 | We can lead or we can follow.
00:30:39.260 | We can lead or it can be imposed upon us.
00:30:42.780 | I know where I am on that equation,
00:30:44.900 | but change is coming.
00:30:46.580 | Second thing, not everything that can be counted counts
00:30:51.580 | and not everything that counts can be counted.
00:30:55.860 | And so this is one of those periods of time
00:30:59.380 | where we're gonna have to put on our comfort hat
00:31:02.300 | with ambiguity and be smart about the choices we make.
00:31:08.460 | Prioritize, focus on business impact, super important.
00:31:13.460 | - And be able to pivot, and be able to pivot.
00:31:19.860 | If something you made a bet on is not coming to fruition,
00:31:23.380 | pivot fast and go to the next one.
00:31:26.060 | - We're going to open up to audience questions.
00:31:27.500 | I think we'll have time for one audience question.
00:31:29.700 | If anyone has one, get your thinking caps on.
00:31:32.460 | But I guess my follow-up question
00:31:34.780 | from what you're just describing is like,
00:31:36.020 | how do you make sure that you don't over-commit?
00:31:37.620 | How do you make sure that in a year, two years time,
00:31:39.500 | you don't turn around and go,
00:31:40.900 | that really was quite expensive.
00:31:44.900 | Yeah, any thoughts on that?
00:31:47.740 | You're all very enthusiastic about this,
00:31:51.060 | but there must be a ceiling.
00:31:52.700 | - Yeah, so and I mean it in all sincerity.
00:31:55.460 | We see it on our product teams.
00:31:57.500 | Hyper-focused from a prioritization standpoint
00:32:01.260 | on what really is going to move the needle.
00:32:03.300 | That's the external facing aspects of our roadmap.
00:32:07.580 | The stuff is going to be expensive to invest in.
00:32:11.540 | There's going to be a dynamic where that line is drawn.
00:32:16.500 | What is below the line today
00:32:18.700 | may have been above the line yesterday.
00:32:20.820 | Be focused and prioritize.
00:32:24.260 | We can't do 100 things well.
00:32:26.060 | We're doing five, we're taking a journey.
00:32:28.580 | We're going to learn a lot in this process.
00:32:30.980 | The five that we're initially doing
00:32:33.460 | are solving real business impacts.
00:32:36.140 | You see the pull through from the business,
00:32:38.460 | our business partners,
00:32:40.020 | and we're not getting out over our skis.
00:32:41.780 | We are moving fast and we're being aggressive,
00:32:44.420 | but we're not getting out over our skis.
00:32:46.820 | And eventually you're going to develop this flywheel
00:32:49.540 | inside of the company where you can do more and more
00:32:52.020 | and develop momentum,
00:32:53.220 | but it's going to be based on proof points
00:32:55.020 | and real experience
00:32:56.100 | as opposed to discussions and conversations.
00:32:59.540 | - The only thing I'd add to that just very quickly
00:33:01.460 | is prioritization 100%.
00:33:03.740 | As CFOs, I think we have a really important seat
00:33:07.260 | at the table to help the company prioritize.
00:33:09.860 | Because again, we see cross-enterprise.
00:33:12.380 | The things that are going to be most impactful
00:33:15.420 | are on the list first,
00:33:16.780 | and you just keep plowing down the list.
00:33:18.380 | But we have that responsibility
00:33:20.860 | to help the company prioritize.
00:33:22.260 | Otherwise, because everyone's pulling.
00:33:24.060 | Everyone wants AI, marketing, finance, legal.
00:33:28.380 | How do we really focus on what's the most meaningful?
00:33:31.700 | I think CFOs have a real role to play there.
00:33:34.540 | - Yeah, and to build on that point,
00:33:36.060 | we have internally our North Star is
00:33:38.860 | anything that helps our own product development on AI
00:33:42.020 | gets priority.
00:33:43.620 | And so we haven't introduced our officials Q yet,
00:33:46.740 | but we already had sales in Q4
00:33:48.420 | because customers were asking for it.
00:33:50.340 | So any investment we make into our teams
00:33:52.500 | and developing our own product helps our customers,
00:33:54.940 | helps the business.
00:33:55.820 | So we'll prioritize that
00:33:57.500 | over some of the internal efficiency maybe,
00:33:59.980 | and then iterate and learn and make sure,
00:34:02.180 | like we're not gonna wait for three years
00:34:04.060 | to see what happened.
00:34:04.900 | Like every quarter we'll be iterating and pivoting.
00:34:07.500 | - Got it, okay.
00:34:08.340 | Do we have any questions from the audience?
00:34:11.100 | If so, raise your hand.
00:34:12.380 | Okay, thank you.
00:34:17.740 | Oh, yes, we have one here.
00:34:19.500 | Microphone is just making its way towards you.
00:34:21.780 | If you could just state your name and company, please.
00:34:25.300 | - Great insights.
00:34:26.140 | Thank you for the discussion.
00:34:27.140 | My name is Guru Ramamurthy, I'm from Bayer.
00:34:30.060 | Some of you mentioned earlier on in the conversation
00:34:32.420 | about not just this being a cost and productivity game,
00:34:36.500 | but also revenue.
00:34:38.100 | When you look at your clients
00:34:39.700 | and the products you're offering,
00:34:41.020 | how does it actually translate into revenue?
00:34:42.740 | Because from where I sit,
00:34:44.380 | this feels and sits more like a cost
00:34:46.380 | and productivity savings initiative,
00:34:48.580 | albeit in a supercharged environment.
00:34:52.620 | Can you articulate a little bit on the revenue side, please?
00:34:55.140 | - I'll give one.
00:34:56.180 | So we announced a partnership with Visa just last quarter.
00:34:59.820 | And what Visa is partnering with us on, with Gen AI,
00:35:03.140 | is all about building a solution to automate disputes
00:35:07.780 | for all the banks that issue Visa cards, for example.
00:35:11.740 | And so the ability for them to monetize that solution
00:35:15.340 | is real, and not only will that generate revenue for them,
00:35:19.060 | from a disputes perspective,
00:35:21.140 | the savings on just being able to resolve
00:35:24.580 | or identify issues much earlier in the process is huge.
00:35:28.260 | Billions is what they're talking about.
00:35:30.180 | And so that's just one example of how Gen AI
00:35:32.820 | is going to help our customers,
00:35:34.020 | not only from the bottom line, but also from the top line.
00:35:37.980 | - Yeah, and on the marketing side,
00:35:40.980 | we've already seen improvement
00:35:43.780 | in open rates on marketing messages,
00:35:46.060 | conversion rates on top of funnel.
00:35:48.580 | So that's one avenue.
00:35:50.740 | Closer to your use case lately,
00:35:52.940 | we have a life sciences customer who's using Gen AI
00:35:56.740 | to collate research, comb through research faster,
00:36:00.820 | and they believe it'll get them to market faster
00:36:03.460 | in order to be able to do both faster drug discovery,
00:36:06.020 | as well as get drugs on the market faster.
00:36:08.460 | So I really do think we'll see
00:36:10.300 | true revenue evolution from this.
00:36:13.380 | - There's a lot of interesting use cases for health.
00:36:19.020 | - Yeah, I'll throw out another example.
00:36:21.180 | IBM just came public with this statistic
00:36:23.660 | in the last 24, 48 hours.
00:36:26.780 | They've standardized on our technology
00:36:28.660 | from a content production and distribution.
00:36:31.500 | They're seeing 10x increase in productivity.
00:36:38.080 | They're serving that content up to their customers 60% faster
00:36:42.820 | and the content that their customers get,
00:36:44.700 | they're engaging 26x.
00:36:47.980 | A lot of it rests on those digital insights,
00:36:51.140 | being able to sort through the complexity
00:36:55.620 | of that customer data,
00:36:58.340 | productivity on the front end
00:36:59.780 | from a content creation standpoint,
00:37:02.260 | and activation and delivery,
00:37:04.680 | the insights that are fueling those types of statistics
00:37:09.680 | for a very large, complex company.
00:37:13.180 | This stuff is real for those that are leaning into it
00:37:16.380 | and leaning in in the right way.
00:37:18.020 | - We're over slightly by a minute or two,
00:37:21.620 | but Dan, I had a final closing question for you.
00:37:23.940 | It's slightly unrelated to the topic of this session,
00:37:26.220 | but one of the themes that we're exploring at this event
00:37:27.940 | is the outlook for M&A.
00:37:29.420 | It's obviously been a quiet period for M&A broadly.
00:37:32.540 | Some of that has been due to concerns
00:37:34.060 | about antitrust activity.
00:37:36.060 | Adobe recently sort of walked away
00:37:37.860 | from its planned acquisition of Figma.
00:37:40.740 | What did you learn from the process
00:37:42.500 | and given the sort of antitrust environment,
00:37:45.940 | is Adobe still in a position to be opportunistic
00:37:48.900 | or are you planning on sort of
00:37:50.460 | wait and see mode for a little bit?
00:37:52.420 | - Well, the great thing is for all the reasons
00:37:54.380 | we've been up on stage talking about,
00:37:56.020 | we've got an enormous set of opportunities in front of us
00:37:58.900 | and we couldn't be more excited about what's in front of us.
00:38:02.700 | When I take a step back philosophically
00:38:05.140 | about what it means to sit in seats like this,
00:38:07.780 | the CEO seat, we plan a decade out.
00:38:12.620 | We think about what it's gonna take to win
00:38:15.140 | and win big in our core markets,
00:38:17.020 | to be a leader, to inflect those markets,
00:38:21.180 | to be pervasive in the way in which
00:38:25.620 | our customers get work done.
00:38:27.620 | Regulatory environments ebb and flow.
00:38:31.900 | Strategic decision-making of enterprises doesn't.
00:38:36.460 | And so nothing changes as a result of an ebb and flow
00:38:40.860 | in the regulatory environment.
00:38:42.340 | We're gonna do what we've always done,
00:38:44.220 | which is drive a strong, organic engine of innovation.
00:38:48.780 | We'll complement it from time to time
00:38:50.620 | with inorganic activity,
00:38:52.380 | but we're gonna be thinking long-term
00:38:54.020 | what it takes our company to win.
00:38:55.820 | We find an opportunity, we're gonna go action it
00:38:58.700 | and we're not gonna try to be prognosticators
00:39:01.940 | of the ebb and flow of a regulatory environment.
00:39:04.460 | [BLANK_AUDIO]