Jeremy Diamond 0:00:12.4:
Good afternoon. Wow. The volume is up. You can hear me, yes? All right, everybody. Thank you for coming. How about those keynotes? Was everybody in? All right. So you are in the finance track. If you thought you were in something else, you're in the finance track. So welcome. We'll have four sessions. The first one is around finance as strategist, trust AI and the journey at Nasdaq. I'm Jeremy Diamond. I look after our financial services practice. Thank you for being here. It's clear, based on the keynotes this morning, that finance leaders are quickly moving along the AI maturity curve, being asked to have more precise handles on their business almost in real time. At Anaplan, we're thoughtfully bringing AI capabilities to the market, aligning the needs of the modern finance leader. Nasdaq is leading from the front of the house and the back of the House. They're really innovating on AI strategies internally, as well as into their products that they serve to their customers, really leading technology advancement and adopting innovation that delivers value. So we'll hear from Nasdaq in a minute. First, I would like to introduce to the stage Neil Thomas.
Neil Thomas 0:01:32.4:
Thank you, Jeremy. Good afternoon, everybody. How are we all doing? Everybody good? All right. This is a proper accent on stage at last. Solid English accent. We either play Bond villains, evil geniuses, or we do Anaplan presentations. So let's get started. All right, so we have an agenda. I'm going to talk for just a few minutes to get us in the mood to talk to Katie - she is obviously the most important person on stage - and we're going to talk a little bit about AI as a theme in the marketplace and how Anaplan thinks about that, addresses that, and then apply it to Katie's experiences of leading the charge with Anaplan AI. So let's do a little bit of AI exploration. This one always makes me laugh in the sense of, AI inquisitive. We're all AI-curious. I'm not quite sure what that means, but how's it going to help our business? How are we going to make practical use out of this amazing technology that's arriving extremely quickly in our business day-to-day lives? So there are a whole bunch of folks that are just exploring. That's the vast majority of the market. That's what we see, just embarking on their journey, thinking about embarking on their journey, and going through that maturity curve that Jeremy hinted at.
Neil Thomas 0:02:46.5:
The interesting thing from a facts and figures perspective is, around about 60 per cent of organizations - this is from a Gartner study very recently - have actually started that journey. So there's still a whole bunch of folks that are too nervous, not ready to go. Maybe some of you are in this room, and that's where Katie comes in to inspire us all to lean in and take this challenge. What's very concerning, of course, and this is truly where Katie comes in, is of those 58 per cent, 83 per cent are struggling with, 'Where's the value? Where is the value? How do I get a return out of this? How do I explain the money and time that we've invested is actually generating value?' Obviously, we'll get Katie's insights into how she thinks about that. From an Anaplan perspective, we've looked at the survey. We grouped that into four distinct areas. So you can see there security. Of course, it's got to be security compliant. That will hold us back if it wasn't. The process itself. We chose the wrong types of processes. Maybe we got too ambitious. Culture. Did our folks, the people, actually lean into this and say, 'Hey, I'm ready for AI. I'm ready to learn. I'm ready to absorb and see how this applies in my day-to-day job'?
Neil Thomas 0:03:57.0:
Of course, you can't do anything without the data. We all know that data is the key to everything. We've got to have our data pretty well organized, in context, in one place for AI and our users to work their magic with it. So how does Anaplan help you with that, as you start thinking about this exploration journey? There are three areas. One is obviously predictive, so working with numbers. That should be easy for finance folks. 'How do I apply predictive AI to my forecasts, act as a guardrail? How does it fit in? Is it more accurate than my human experience?' We heard a little bit about that in the keynote, how it became more and more accurate over time, and the humans started to get things wrong or were out, and we had to change our behavior and apply their value somewhere else. Secondly, are my users ready? Are they ready to ask questions of the system? Are they going to be comfortable in this environment? Last but not least, my favorite AI application we have is CoModeler, because that is contained within the center of excellence or my Anaplan administrators, where I can be much more productive. It's much more controlled. Doesn't necessarily affect my end users, other than they see more return and more applications appearing on their desktops.
Neil Thomas 0:05:06.9:
So that's how I think about it, Anaplan thinks about it, and there are a bunch of areas that you can start in and, of course, start to address some of those four key areas of concern. CoModeler is the most straightforward. If I've got my data organized, I can really go off and extend my models, document my models, as Joe showed this morning. If I'm further along and I've got the right kind of culture and mindset, I can start working with my analyst and start allowing my users to start interacting with AI. Of course, I need this to be compliant. I need to explain the security to my senior executives. So we can go on a journey. I look after our applications area, particularly with a focus on finance, which is why I'm in the finance track, and what we see with our integrated financial planning application is a hub, a really good starting point for getting that data organized and then extending out into the organization. So a journey may look like this, if you're embarking in a finance project, which is, I'll start with an application. That brings me structure. It gets my data nicely organized, and then I'll extend perhaps into predictive because I'm a numbers person. Let's get my forecast accuracy, my forecast process optimized with predictive, and then through the journey perhaps into analyst and CoModeler.
Neil Thomas 0:06:18.1:
So how do we unlock the value? Where are the benefits? Obviously, the benefits for each group are slightly different. We're talking about productivity, efficiency, doing more work with our model builders, taking all their expertise and just making them that much more productive. With our analysts and our planners who are working in the system, obviously, we're looking at driving consistency, standards of behavior, making it very easy for them to interact with that data, all those lovely grids and screens. How do I get there more quickly to the insights that I need? Again, being extremely compliant with my overall process from a security standpoint, a data quality perspective. Finance is at the core of all of this. If you're thinking about embarking on that journey, again, back to the four things to really think about. Is my data ready? From the perspective of picking the right process, don't boil the ocean. Maybe start a little small, get some practice, and then grow from there. Obviously, have I got the right culture, whether within my Anaplan master builders - are they ready? Are they willing? Are they going to take the bumps and bruises of going through this learning curve? - or, of course, with my end users. Are they ready? Are they leaning in and helping me optimize this? And then, of course, ticking that security standard with the right kind of compliance help from us at Anaplan.
Neil Thomas 0:07:33.9: So that's a little bit about the journey and how perhaps to apply some of the products that you heard about today, and I'm delighted that what we'll do is we'll talk to Katie, who you heard from this morning, in a little bit more detail about her project and her journey. So please. Katie, come on up to the stage. Katie Bird from Nasdaq. Good afternoon. Did I say good morning earlier?
Katie Bird 0:07:53.1:
Yes, we have the post-lunch slot.
Neil Thomas 0:07:56.1:
We do, we do.
Katie Bird 0:07:58.3:
We've got to keep them awake.
Neil Thomas 0:07:59.2:
We're going to try. We're going to try. So I'm going to chat to Katie for a little bit. We're going to make sure, or try to make sure, in case we get too gossipy, that we leave time for questions from the audience. So, Katie, thank you for the repeat act, as I mentioned. I hope we're paying you overtime, or extra big steak dinner tonight or something like that. Thank you very much. Just for the folks that may have missed it this morning, could you explain your role and where you fit in at Nasdaq? Let's start there again.
Katie Bird 0:08:28.1:
Yes. So I lead our corporate financial planning and analysis team within Nasdaq, and we have a very federated finance structure. So we have the corporate organization. We are responsible for bringing everything together into that holistic Nasdaq view, and then we have divisional finance teams that are embedded within each of our three divisions, and they're working very closely with each of our business teams, each of our product teams on that bottoms-up input into that process. So we have one overall process that we run, but we have lots of mini-processes that are feeding into that across each of our businesses.
Neil Thomas 0:09:09.5:
Excellent. Then, everybody says they're complex. Everybody I have a half on stage. 'We're really complex.' So obviously, Nasdaq is complex as well. Can you give the audience a flavor of maybe one or two of the most complex planning problems that you guys have to deal with?
Katie Bird 0:09:22.5:
Yes. So I think the first is just the diversity of revenue streams that we have within the organization. We have trading businesses both in the US and in Europe that both operate very differently. We have an index business that has its own unique revenue profile, and then we have a whole bunch of SaaS businesses, but they all have a little bit of their own unique flavor as well. So what we've really tried to do is figure out, 'How do we have a process that allows for that very discrete use case within the business, but is still feeding into the overall corporate picture?' So that's one part of the complexity. I think another is just all the different technologies that we use, and making sure that each new piece we bring into the picture can work, can play well with the others. So that's another piece of the complexity. Then I would say, lastly, it's really wanting to get to more of that very close to real-time ability to understand what's going on in our businesses. We talked about it a little bit this morning, but there's a lot of things that happen in the world every day that impact some part of our business, and sometimes many parts of our business.
Katie Bird 0:10:51.2:
Of course, our CFO wants to know what is the impact of that, and that takes a lot of people running around updating Excel files to figure out. That's what we're really trying to solve for, is how do we get much faster time to value on those insights? Because we do really use that to drive the decision-making process within the organization.
Neil Thomas 0:11:12.5:
Within your COE and the federated finance, what are the types of characters in there? How did you decide to design your COE from a skills and ability perspective?
Katie Bird 0:11:23.4:
Yes. So first of all, have a COE. I highly recommend that if you just let everybody go off and build, there's a lot of flexibility in the platform and you're going to end up with a nightmare trying to bring it all back together. So we have a very small COE. We brought in a solution architect, which I also highly recommend. He's fabulous, and he can… I have my big picture of where I want us to get in my head. He can translate that into, 'How are we going to structure all of this?' We've actually built things very agile and very iteratively as we've gone along. So last year… We just started a year ago. We actually signed our contract the day before Connect last year. Jeremy threatened to not let me come to dinner if we hadn't signed it. We started with two of our more complex businesses for different reasons, one of our SaaS businesses that, first of all, really wanted to go first, and it's really great to have a team that's excited to change and excited to be part of that process. But they also wanted to be able to forecast at a customer level, have it integrated with our CRM and all of our pipeline data, and really have this kind of end-to-end integrated process.
Katie Bird 0:12:46.1:
So that was one use case, and we felt like if we could build that well, and it worked well, we could copy-paste that across the rest of our SaaS businesses with just little tweaks where we needed to make those. So that was one, and then the other was our index business, and we talked about this a little bit this morning too, but very, very different type of model. They wanted to be able to go to a level of detail they could never get to in Excel. The reality is, all of these models are in Excel today, and we're moving from Excel into Anaplan. So they were really able to reimagine their whole planning process because they could now get to this very, very low level of granularity, and so we started there. Each of those teams co-built the models with our COE. We also have an implementation partner, Accordion. They're here. They're excellent. Highly recommend. We did those two. They demoed what they built back to the other divisional finance teams, and then everybody wanted to go next. That's a great problem to have, because when you're trying to change how people do their jobs, it's a heck of a lot easier if they want to change. So then that became the challenge of, 'Okay, how do we prioritize the rest of these models?'
Katie Bird 0:14:13.3:
The point about data is critically important. It took us a lot longer to get the data ready than it did to actually build the model. So we built the models very quickly, but as we tried to integrate all the pieces in, that's what took the longest amount of time. When you do things in Excel, you can do whatever you want. You can overwrite data, you can change data, and we did not want to replicate that in our shiny new tool. So that took a lot of time. What we've actually done as we move forward is, the data is the homework. You've got to get your data ready before you can get in line to get your model built, and so that allows us to move faster because we're building the models that are ready. They've done the hard work on the data, and it gives the other teams who need to work on their data a bit more time to work through that so that they're ready to go when their time comes up in our process.
Neil Thomas 0:15:14.2:
Just taking you back to the COE for a second, so you have a solution architect. Do you have anybody very data-technical or AI-qualified?
Katie Bird 0:15:20.9:
We have a whole team that is part of our corporate technology team that manages data, so they manage all the integrations of the data. We have a master data management tool that we've put in place that manages all of our master hierarchies, and then we actually leverage Databricks as our corporate data lake house, and so we have a whole technical team that supports that. So we obviously partner very closely with them, but the models themselves we are building with Accordion, with the hope of becoming self-sufficient as we get through these. So that's really how we've got it structured today.
Neil Thomas 0:16:03.1:
Yes. When you and I first chatted, the culture… So clearly, Nasdaq is easy on AI from a cultural perspective because you guys are leaning in for your customers. It's just natural to do it internally. I just want to talk about that, maybe, just a little bit more of maybe how easy it was, or not, and then some advice for other folks that are trying to get ready for this.
Katie Bird 0:16:24.3:
Yes. It's obviously very easy when you have a top-down drive, from our CEO all the way through our leadership team, that's very focused on embracing AI and really figuring out how we do this well. Our CFO is the same way. She's very tech-forward, very data-driven. Today I am her AI agent. She wants to know something, she sends me a note. I'm very excited to give her an actual AI agent to do that with. So that makes it a lot easier, for sure. What's hard is the rest of us who actually have to figure out how to use this and how to bring it into our day-to-day, and I think you have people who have really embraced it and they're really excited about it, and they're using Copilot to summarize all their emails for them, and they're using it in their personal life, and so this is just fine for them. They're very excited about it, and then you have other folks that are a little bit skeptical or a little bit nervous, so that's actually where we spend more time, I think, is the actual day-to-day users of these tools. I think as people start to use it and they realize the human doesn't go away… We know our business in a way that AI is never really going to understand our business. It can answer questions. It can highlight things that don't make sense, but what that means and what we should do about it, that's still the intuition part of our jobs, and it doesn't replace that, at least not yet. Hopefully not for a while.
Katie Bird 0:18:10.2:
So I think once people start to realize, 'Okay, this is really just doing the things I don't really enjoy spending my time on, and I actually have time to…' It's telling me what all my variances are, and so then I can actually think about, 'Okay, well, what do I do about it?' instead of, 'I spent 80 per cent of my time figuring out what all my variances were.' I think that's where you kind of start to see the bend in the adoption curve in the right way like, 'This is actually going to really help me be more strategic in my job.'
Neil Thomas 0:18:43.2:
All right. So a year ago we signed the contract. Just give us a flavor of progress. I know you talked a little bit about the two initial pilot areas, but progress to date and some lessons learned, perhaps, along the way.
Katie Bird 0:18:58.1:
Yes. So I mentioned we started with a couple of POCs last year, really to pressure test what we believed to be true, but we wanted to know it was true before we went all-in across the organization. Then we expanded that partnership towards the end of last year, and we're now moving across all of our different revenue streams. So that's one track. We're also going back and layering Forecaster into some of those models that we built last year, and we're really thinking of it initially as like a challenger model to the model that the team is doing. So it's another perspective on what the forecast could be, and I think if we do that for a while, it builds trust in the outputs of Forecaster. It'd be fun to see who's more accurate. So that's one track. The second piece we're doing is workforce planning. So for us, 60 per cent of our cost base is people, so it's by far our largest expense, and we don't have a great integrated process today between how we forecast people and how we actually hire people. So it becomes this big reconciliation exercise every month of, 'What did we say we were going to do? What did we actually do? How do we need to change what we say we're going to do now?'
Katie Bird 0:20:31.4:
So workforce planning builds that bridge, and it all becomes very seamlessly interconnected. It also gives our people team a lot more insight into what we're planning to do than they have today. They don't really have a view into our forecast today, who we're planning to hire, where and at what levels. So they end up having to be very reactive as recs come their way to hire, and this gives them the ability to see that, but then to be able to plan themselves for how they're going to meet the needs of that hiring plan. So that's the second track.
Neil Thomas 0:21:12.5:
I have to do an ad. So, you are using our operational workforce planning, and then driving the team crazy with extensions.
Katie Bird 0:21:22.0:
We give a lot of feedback.
Neil Thomas 0:21:24.4:
Yes, maybe not!
Katie Bird 0:21:26.5:
I also think it's a newer app, and the product team has been phenomenal about sitting down and understanding what we're trying to do. I think we're definitely pushing the limits of the app in some ways, but a lot of it is stuff that they're like, 'This is great. We want to just build this into the app and not have it be an extension.' As I was saying this morning, finding a platform that also was very much a partner was really important to us, because we're probably going to keep doing that. Sorry, guys. We've been working with them on kind of expanding what we can actually do with that tool. Then the last piece, which we've not started yet, but it's next, is the integrated financial planning. So we're doing these pieces, and then we're going to bring it all together with the IFP app, and then also layer in the expense side of everything. So you'll have a huge chunk of your expenses coming from the workforce planning implementation, but then being able to build out the rest of that expense forecasting, bring the revenue models into it, and now you start to get that holistic picture.
Neil Thomas 0:22:37.0:
So you've gone revenue, workforce, and then full financials to finish. That's your journey. I meant to ask you this. You said about a good partner. How did you decide Anaplan was going to be a good partner, and what does that mean? Actually, one thing I would ask also is, because you're very demanding - I've heard this - so recommendations to the audience here. When I meet a lot of customers, they're not as demanding as you, but you're one of our favorite customers because you are demanding and helping us get better. So maybe some thoughts on, 'Anaplan is a good partner. I realize this because A, B, C,' but also some of those interactions with our apps team, how they could leverage their Anaplan relationship even more and get more value. That would be super.
Katie Bird 0:23:26.1:
So you can tell when you sit down with someone if they're going to be a good partner or not. You can tell. You have a conversation, and either they just keep giving you their pitch and showing you their demo that's totally not aligned with what your business does at all, or they listen to what you're saying and they tailor their demos to exactly what you're trying to do. We sat down and said, 'Look, this is a journey for us, and we're going to have to walk the organization through this, and you guys are going to have to walk with us. We're going to start small. We're not going to sign the contract for all the things day one. We know we're going to do that eventually, but we've got to prove this to the organization. We've got to make sure we feel really good about it,' and I think that's actually a testament to your team and your confidence in your product. We basically were saying, 'We want to try it before we're willing to sign a big contract with you,' and you said, 'Great, because we know you're going to love it, and you're still going to sign the big contract with us.' So I think the flexibility and being willing to work with us on that was… That's the partnership, but also the confidence you all had in, 'They're going to get there.'
Neil Thomas 0:24:48.0:
Excellent. As you look at the technology, initial evaluation and today, the key sort of features and capabilities that you were, 'It's got to have, got to have…' Any thoughts on that checklist?
Katie Bird 0:25:05.0:
Yes. So we've just started playing with CoModeler. We're really excited about it. I think that is a huge enabler for the teams, and it's funny because each new thing we build… We're doing this in a very agile way on purpose because you can deliver value very quickly, right? We've already made two teams' lives better by building and deploying their models, even though the rest of the models aren't done yet. So that has been really key, but I think CoModeler helping us build things faster or giving recommendations on how we can make some of the models we've built better, and I think the teams who have their models, inherently, the more you use something, the more you realize, 'Oh, I want to do this instead,' or dynamics change in the business and you want to do something differently. So my hope is that that will be an enabler to those divisional teams to update pieces of their models themselves, and not have it all get bottlenecked in the COE. So that's one thing we're really excited about. I love the scenario capability. We do a lot of scenarios. Some of our businesses have a lot of beta impacts on them, and so being able to look at, 'Here's our forecast, but what if this happens, or what if this happens, and how is that going to change the picture?' we can't do that today because everything's sitting in a separate Excel model.
Katie Bird 0:26:37.0:
So that capability I'm very excited about, because not only can we do it for a business, but as we get everything brought together with IFP, then we can do it across the whole business. So that's going to be a big time saver for us. Then I'm really excited about Forecaster and figuring out, 'Can we get better at forecasting by bringing more of a machine learning-based process into it?'
Neil Thomas 0:27:05.5:
Got you. Okay, fantastic. Now, you're one of the few people that, when you were talking about Anaplan, talked about MCP, which is great. Obviously, back to the culture of Nasdaq and just where you are. Give us a flavor of maybe… You may not have started this yet, but some of the, 'How I'm going to leverage that open architecture that Joe talked about with other parts of the business data, that bigger vision.' That would be exciting.
Katie Bird 0:27:29.1:
So we use a lot of different technologies more broadly across finance. We have a GL system, we have a consolidation system, we have our whole procurement PO system. They're all different technologies. So it's very important to us that… Nobody really does their job in one place. So as we think about agentic AI in the future, there's really likely going to be the need that you're going to have to have agents that can move across technologies. So these very closed-ecosystem platforms are just not interesting to us, because we don't live in one platform in finance. We operate across a lot of things, and so we're not using it right now, but the ability to do that as we get to that point in our journey was really, really important to us. The other piece is, we have our own gen AI platform that we've built within the organization. We actually can't access ChatGPT because they don't want people putting proprietary code in ChatGPT. So we've built our own that works in the same way, but it's got lots of different LLMs you can choose from because different ones are good at different things. So being able to connect Anaplan to that is going to be something that'll be really helpful as well, because at the end of the day, there's a lot we can do with what you've already built in the tool, but then there's the next level up of how it can participate in a broader ecosystem of things we're building.
Neil Thomas 0:29:10.4:
Your planning cycle is a rolling 12 months. What sort of horizon do you guys look at?
Katie Bird 0:29:16.3:
So what we do today, which we're hoping we change, is we update our current-year forecast every month. So every month, towards the end of the month, we update what we think the rest of the year is going to look like, and we use that very actively in how we manage the business. We use it as part of how we prepare for earnings. It's a very important part of our process. In between those forecasts, you don't have a view. That's the blind spot. You can figure out pieces within each of the teams, but you don't have the whole big-picture view. So that's really what we want to solve for, and then we do a long-range plan, which is the current year plus two, about once a quarter. We really use that for that more strategic planning and new markets we might want to enter, new investments we might want to make and how that's going to play out over a longer-term time horizon.
Neil Thomas 0:30:17.2:
You talked a little bit about the value so far. Hard to do an ROI on all this sort of stuff. So any thoughts on a reminder of where you are now and, 'We were able to prove A, B, C and original business case, but then as we unlock more, we can see more ROI areas for that.'
Katie Bird 0:30:37.4:
Yes. So we actually have a pretty rigorous investment case process at Nasdaq. Our current CFO brought it to the organization about two years ago when she joined. So we actually do have a five-year view of how we think this will play out from an ROI perspective. Obviously, there's less in the early years because you've got to do the work to get everything in place, but then we expect there to be real savings on the other side of it. It's about scale. It's, how do you take the organization and do more and support a growing business without continuously adding more people to the team? So it's not reduce. It's, 'How do we bend that growth curve?' That's really how we're thinking about it. How do you make these teams able to support more and more as the organization grows? We're growing at double digits from a revenue perspective, and so that's how we're thinking about it. We do have an actual business case, and we require it of any investment, any organic investment we're going to make within the company.
Neil Thomas 0:31:49.3:
The primary buckets of that are doing more with the same amount of people and saving…
Katie Bird 0:31:55.0:
Some of it is other technology we think we can deprecate, which is real hard cost savings, which is great, and then some of it is the avoidance of… If you look at a trend line or you look at your unit economics around the size of the team relative to the number of things that they do, if you carried that trend line out, what would that look like? If you can bend that curve, then what is that saving you in future costs that you don't have to incur?
Neil Thomas 0:32:28.2:
That makes perfect sense. Okay, we've got a bit of time. We've got some time before questions. What have we missed? Have we missed anything we should talk about before we go to questions?
Katie Bird 0:32:38.4:
I don't think so.
Neil Thomas 0:32:39.0:
I don't think so either, no. So has anybody got a question for Katie that they'd like to desperately ask? If not, we'll get on to Katie's favorite band, best place to visit in New York, restaurant recommendations for this evening. What else can we do? Has anybody got a nice question for us over there? It's very quiet… It's because you've answered every question. Oh, sorry. Hello.
Audience 0:33:12.3:
Yes. So I was wondering if you could share what primary use case Forecaster brings to Nasdaq.
Katie Bird 0:33:20.1:
So we're looking at it in two different… In our current models that we've already built, we're going back and layering it in. So for index, that's actually a very interesting use case for us because we have a lot of historical data on flows and market performance, and how that aligns to other macro data and things that are going on from a volatility perspective or whatnot. Can we then use that to create some scenarios of how the future could look based on the same macro data? So that's one place we're looking at it. In the SaaS business that we did, we have a longer time to value or time to revenue on the sales there because there's long implementation. It's our anti-financial crime business, and there's longer implementation cycles, but they're also relative to the size of the bank. So being able to bring Forecaster in to look at, 'Are we making the right assumptions on that time to value based on what we've actually experienced in the past? Can we get more granular with it instead of the big buckets that we use today?' So those are the two initial use cases we're looking at.
Audience 0:34:38.1:
Okay. Is it more of a dollar-based output or like a volume output, you would say?
Katie Bird 0:34:43.3:
It's more of, can you use it to help inform how your drivers might be different, which then translates into how your forecast might be different?
Audience 0:34:54.2:
Perfect. Thanks.
Neil Thomas 0:34:57.5:
Drivers being volume, price or anything.
Katie Bird 0:34:58.7
Drivers being whatever is the driver for that particular business.
Neil Thomas 0:35:04.4:
Pretty much any variable.
Audience 0:35:08.8:
Thank you. It's a very interesting discussion. You were mentioning that you were converting data from Excel to Anaplan. Are you replacing Excel so you now won't use Excel? Then, did you use AI to clean the data such that it could be updated?
Katie Bird 0:35:31.1:
You can never completely replace Excel. Not in finance, I don't think. We are moving the models out of Excel and into Anaplan. There's still a lot of analysis the teams are used to doing in Excel. Anaplan has a great plugin that you can use to enable that, but we're also trying to get away from that as much as possible because you can't put AI on top of Excel. It doesn't work. So having everything actually in the platform enables us to layer the AI capabilities on top of it in a way we can't do in Excel. So that's been a big piece of it. We've not used AI to clean the data, necessarily. What we've really had to do is figure out, go through the process of saying, 'Where is the source data for this? Does it have a system, and can we integrate it in directly? Is it in our corporate lake house already?' which is great, because then there's connectors where we can bring it into the data orchestrator and have it available for the models, and then, 'Where are we manually manipulating that data, and can we stop doing that?' So if, for example, we have it integrated with our pipeline, and if the finance team is changing assumptions on when deals are going to close, can we actually change it in Salesforce and have it flow through instead of have this overwrite process at the end?
Katie Bird 0:37:06.5:
Now you've got conflict between what your sales team is saying and what your forecast is saying. It also highlights those things where the finance team doesn't believe you, and so then you have to actually sit down and have a conversation about it. So that's really where the data cleanliness is really important, but also trying to remove as much of that manual manipulation as possible so that you can really… That's how you get to that faster, almost real-time process, is when everything is just flowing, and somebody changes a close date in Salesforce and it's flowing all the way through to your revenue.
Neil Thomas 0:37:53.2:
So basically, you would say once-off type of analysis, Excel… We do have a very nice Excel add in it. It's actually fantastic, yes. Probably the best I've ever seen. Repeated obviously has to be in Anaplan for all the other advantages.
Katie Bird 0:38:07.5:
Yes, and there's a lot we do in Excel. There's lots of little models that live in Excel today. There's the big revenue models, but there's lots of other things that we do in Excel today, and it's really cool to watch the teams say, 'Okay, we could do this in Anaplan,' or you think about M&A. We've done a lot of M&A in the past, and so can we leverage Anaplan to overlay a potential acquisition model and see how that rolls up into our financial picture? So we have a very long list of ideas and things that we want to tackle. We have a very structured plan for this year. We have the idea list of things that the users have come up with of, 'I could probably do this a lot easier in Anaplan.' You start to connect more and more of the pieces by doing that, and then it's not, 'I've made this change. Make sure you update the forecast for this.' It's just, 'I made this change and it updates the forecast.'
Audience 0:39:14.4:
Hi. So you had talked a little bit about your process around measuring investments, the ROI on investments and how you've thought about that with the Anaplan use cases. Have you identified use cases to use Anaplan to further that analysis and process across the company?
Katie Bird 0:39:33.3:
So we actually do use Anaplan for this. It was the third thing that we built last year as part of our POC, because we have this great process, but operationally, it was kind of a nightmare. You had thousands of business cases in Excel files, and you can't manage that. So one of the things we built last year is, we rebuilt that whole process in Anaplan, so the business models actually all live in the application. It's a little bit separate today. It kind of sits over here, and it's not connected yet. As we get IFP in place, then the idea is that we actually interconnect those two things, and so you've got the real-time updates back into the models of how you're progressing from an actual perspective. Just building it in Anaplan, and then we put a great dashboard on top of it, we gave our senior leadership team access to it, there's a ton of reporting we can now do on, 'Where are we investing? What products?' We do this on our client-facing products too. Everything has a business case, and we put them in different horizons based on their return profiles. So we can actually understand where we are spending our capital organically across the organization, and which buckets of return profiles is it in?
Katie Bird 0:40:59.1:
Obviously, you want to put more in the higher-return ones, and how do you bring down the must-do, low-return things. We are using it for that, and it's been tremendously helpful.
Audience 0:41:11.5:
Thank you.
Neil Thomas 0:41:14.4:
So you justify buying more Anaplan in Anaplan?
Katie Bird 0:41:19.2:
Exactly.
Neil Thomas 0:41:19.3:
We should have an AI agent that promotes us to the top or something.
Katie Bird 0:41:24.2:
Luckily, my team runs that process too.
Neil Thomas 0:41:28.4:
So any other questions from the audience?
Audience 0:41:32.2:
Did you start with Classic, or start with Polaris?
Katie Bird 0:41:35.2:
So the question is, did we start with Classic, or did we start with Polaris? We started with Classic, because again, we were just pressure testing it, and so the first three models we built are in Classic. As we've now taken this big step into bringing the two apps in and doing a lot more, we've brought Polaris in as well, and so we're building some of the new models there. We'll probably eventually come back and move the early ones, but we don't have to. We can still connect them in, and I heard this morning that I've got until 2035 at least to get that done. So yes, that's' how we went about it.
Audience 0:42:17.0:
[Unclear phrase 0:42:17.0] biggest difference as before?
Katie Bird 0:42:20.5:
I don't actually know because I'm not as close to the actual model building as my team is, but I can grab one of them for you afterwards and have them answer it.
Neil Thomas 0:42:30.2:
Just to be clear, all applications that you heard about today, they're all in Polaris as a default. Eventually, you'll probably switch over at the right time when it's worthwhile, possibly to integrate more data more easily, that kind of thing. No rush at all. All right. Looks like we've exhausted everybody, it seems. Thank you, Danielle, for… Oh, we have a hand up. Sorry, Danielle. You've got more walking to do.
Katie Bird 0:42:54.1:
You're getting your steps in today.
Neil Thomas 0:42:55.4:
Get your steps in. That's right.
Audience 0:43:02.2:
Has the implementation improved the forecast accuracy, and if so, by how much?
Katie Bird 0:43:11.2:
So I don't actually know, because we're not running it in parallel. So we've done pretty hard cut-overs to the Anaplan model, and we're not continuing to keep the manual process in place. As we layer Forecaster in I will be able to measure that, because then we'll almost have two models. We'll have the model that the team is putting in, we'll have the model that Forecaster is driving, and we can actually compare them to each other. For me, it's less about the accuracy. It's more about how much deeper we can go in the forecast, and how much faster we can turn the forecast around, and that has improved significantly.
Neil Thomas 0:43:59.4:
You said, depth, speed and frequency, right? No more monthly, get away from that. Yes. You'd need a parallel universe for the comparisons.
Katie Bird 0:44:09.3:
Right.
Neil Thomas 0:44:11.3:
All right. Well, we will stop there. So Katie, thank you very much indeed, I appreciate that. Wonderful. We'll let you leave as I wrap up. Oh, I had another slide with beautiful QR codes and things. Oh, there we go. So thank you again, Katie. Thank you so much. We've talked about these three products primarily today, so feel free to scan QR codes if you're into that kind of thing. It's always nice to see when somebody scans a QR code. Thank you for doing that. Makes me feel like this was worthwhile. Great. Just thank you, everybody for attending, and hopefully you'll hang out with Lincoln Financial in 15 minutes. We'll be back on with Natalie. Thanks, everybody.