Lindsay DiPietro 0:00:10.1: 

Hi, everybody. Thank you for joining us today. Sorry we stand between you and happy hour. My name is Lindsay DiPietro. I lead our product management function for our retail applications, and I'm joined by James, who leads our go-to-market for retail applications, and ex Syrup, newly minted Anaplanner. You got it. You know, James, what's interesting is the last time that I was in the US, I watched the Canadian women's hockey team lose the gold medal - in my opinion, they lost the gold medal to the US women's team - and so being back here feels a little bittersweet for me. 

James Theuerkauf 0:00:55.4: 

Well, Lindsay, if Team Canada had you on their side, they would have fared better. If you don't know, Lindsay's an accomplished ice hockey player. I know very little about ice hockey, but here are some of her stats that you see on this slide here. 132 career assists, top three in school history in goals assists, points, power plays on shorthanded points, game winning points plus-minus, as well as 0.98 assists per contest, as well as Wayne State Athletic Hall of Fame. Exactly. 

Lindsay DiPietro 0:01:27.4: 

Thanks for taking me down memory lane. Very humbling. That does serve a purpose here, though. I'm curious. How many people have been to a live sporting event, or played high-level athletics and have watched a great team come together, and it seems like the people can read each other's minds? Anybody been to one of those games? Okay, great. I'm here to tell you that's not magic. When it happens, it's actually a bunch of very practiced systems that are working perfectly in sync together, and what that requires is for the players to have a deep sense of trust in those systems. What that trust does is, it creates this momentum or creates this rhythm for the team, and that rhythm creates what we call structural confidence, and that's really no different than how businesses operate. You have teams or different departments across different parts of the organization, and they're all potentially operating in different systems. Those systems, sometimes people trust them. Sometimes they don't. What we're really here to talk about today is how you can create that structural confidence, how you can create that rhythm, so instead of playing on your heels, you get to play on your toes. 

James Theuerkauf 0:03:01.1: 

Yes, that's right. The confidence really relies on a foundation of data, of intelligence and of workflows, and we'll get into that in quite a bit more detail in a moment. Just foreshadowing that, getting this right is what compounds advantage, as opposed to compounding risk. 

Lindsay DiPietro 0:03:26.0: 

Yes. So, how do you create that advantage? We think the advantage really comes from three things. One is sensing, so the systems being able to sense what is going on and being able to present that in a way for individuals to confidently take action or decide or make decisions, and for teams to be all aligned and executing on that strategy together. That's really what creates this continuous planning. That's what creates the rhythm and the confidence within the teams. To be honest, a lot of businesses are doing this today, but a lot of businesses are also doing those things in isolation of each other. What we're trying to do is actually bring those things together so you can have this continuous flow and information. So what we're going to talk about today is where this breaks down and really how we can help. 

James Theuerkauf 0:04:24.1: 

Exactly right. So let's talk about decision misalignment. I would pose that the vast majority of planning teams typically make good decisions. Merchandise planners make good decisions. Allocators make good decisions. Buyers make good decisions. Finance makes good decisions, but they typically make good decisions in isolation. So even when decisions are good in isolation, they might not be good in their entirety. They might be good decisions, but they're not aligned decisions. Multiple examples. We work a lot with allocation teams, and if you look at your sell-through report or sales reporting you might think, 'Oh, we definitely need to allocate more product from A to B,' but if you don't know what the cost of that is and what the opportunity cost is of not allocating it to B, you're missing out on a core component of that distribution allocation strategy. Or color trending, for example. Your planning team might see that a certain color is really picking up, like the Gen Z yellow is really picking up, but if you've already committed and locked in your OTP to the navy blue, there's maybe not that much you can do. 

James Theuerkauf 0:05:34.2: 

The navy-blue decision might have been the right decision a couple of weeks ago, but if those are not connected and they're not aligned, then ultimately for the business it's actually not the right decision. So that's what we think of as decision misalignment. If we go to the next slide here, we think of decision misalignment not necessarily being a process problem. It can be a process problem, but it's not necessarily only a process problem. What it is, it's really an architectural problem. What we mean by that is on these four components that you see on this slide here. Really these are the reasons why decisions maybe break down, or why decision alignment breaks down. On the first we have, 'The data foundations are fragmented.' We all know this. Data lives in different places. People look at different data. 'Is your cut the same as my cut?' That's a core component for decisions breaking down. Then we've got the second one of cross-functional teams lacking shared guardrails. This example is, I might be optimizing for one thing, but someone else for something else. So what are those guardrails? What are the rules of the game that we're actually optimizing against? 

James Theuerkauf 0:06:45.5: 

Then we've got intelligence and workflows operating independently, on the top right that you see there. This is the core thing, of 'Oh, my workflow is this, but then I take it out, I put it in Excel, I make my decision elsewhere, and then I put it back in there,' and you have this breakdown of intelligence and workflows, and then the last one is insights and action being separated in time. The insight comes at one moment. The time it takes from the insight to the action is - what we're going to be calling in this report that we're going to be releasing - the latency tax. What's the cost of you actually being slow? We've done this vast survey of 500 leaders in CPG and retail in planning. We're actually going to be releasing the full results in April, but a sneak peek, the core insight is that for every dollar of revenue, every dollar of revenue carries a five-cent disconnected decision tax or latency tax, which is massive if you think about that, for the scale of global retail organizations. So the root cause of being, as you were saying, Lindsay, on your heels as opposed to on your toes is uncertainty. 

James Theuerkauf 0:08:02.4: 

So the core question that Anaplan will help you answer, and that we're going to be exploring over the next couple of minutes with you all, is, 'What foundation allows planning to work as a system and individuals to act with certainty in a world that is inherently and ever more uncertain?' So the solution is a foundation for structural decision confidence. This is a complex slide. I'm going to spend some time on it and break down the individual components. I'm going to speak about the two components, the data layer and the intelligence layer, and then Lindsay's going to speak about the workflow layer, effectively. The crux of all of this is that we can make decisions together that are connected and that speak to one another, and so let's speak about the data layer first. You see that, obviously, within the Anaplan world. That's data ontology, which means, 'What does our data actually mean? What's the framework? What's the canonical data model that explains what our data means?' and that's extremely important. 

James Theuerkauf 0:09:15.0: 

As you know, for those of you that work in planning, how much time you spend on aligning numbers. 'Can you please recut this with that? What if you cut it this way?' and then the hours and hours that are spent in endless meetings on this alignment of numbers, that's not what you're really good at, and that's really not what the business necessarily needs. What the business needs is acting upon these insights. That's why we need like one version of the truth, or one version of what our company's data really means. That's one view of true demand, one view of true inventory, one view of the financial truth. It requires four things. It's standardized data structures, i.e. like a data ontology or a data model. Harmonized hierarchies across teams. Different hierarchies - style, fabric, fit, color, et cetera - mean the same thing. Consistent product and location logic and continuously refreshed signals. Even if we all have the same data, but it's not actually up-to-date data, the data is not that useful. So those four components really allow the foundational elements of what it means to shift from 'What's happening? What's the insight?' to, 'What should we do? What's the action?' So that's part one. That's the data layer. 

James Theuerkauf 0:10:37.1: 

Then we've got the intelligence layer. That's the core thing that the company that I founded, Syrup, that we worked on, so it's near and dear to my heart, the intelligence layer. Really what that means is going from reporting and making sense of the past to understanding uncertainty in the future and what you should do about it. There's four requirements that we've got on the intelligence layer as well. It's insights at every hierarchy, i.e. 'It's great that a certain product is performing well nationally, but what does it actually mean regionally?' There's massive variance across that. It also requires continuous updates that data. Demand is always updating, and more so now than ever before, so a sensing engine needs to be able to pick up those trends and those microtrends, and be able to do that across different time horizons. Easter last year is different to Easter this year. A certain color trend last year is not the same as this year. So how are we picking those up and how are we sensing those? 

James Theuerkauf 0:11:40.7: 

Then there's the signal responsiveness. We don't want to be overly responsive if an ice hockey team walks into a store and buys a bunch of stuff, and then we're overreacting to that. We need to be able to exclude those components, but we need to be able to act on things as they pick up, otherwise, we've missed the trend entirely. Then the last component is being able to actually act on what the intelligence is doing. We saw that in this keynote this morning. Dashboards are great, but you can't just act on dashboards. What you need is recommendations or actions that the intelligence is providing to you, and that's where we are combining forecasting and optimization. Forecasting is the component that will tell you 'The demand for this product is 50,000 units over the next X period of time, but you only have 30,000,' so how do you translate a future-looking unconstrained view to actual implementation? 

Lindsay DiPietro 0:12:44.1: 

Thanks for that. I get to talk about the fun part, the workflows. Any time I get to engage with a customer, I actually spend a lot of time talking about the data layer, more so than anything else, because I think without that it's really hard to be successful in the next layer. So when we think about workflow, or we think about our applications that sit on top of all of your data and all of the intelligence that we bring, what we're then able to do is have these applications have intelligence embedded in the workflow already. So think about from MFP to assortment planning, I can actively see my budget. I know how my assortment plan is rolling up to that budget. Oh wait, I'm sensing what's happening in-season as well, so I know how much I have in my open to buy, to spend, to chase into, or maybe I need to take some markdowns because I'm way over budget. That also happens. So it's important, like James said, not only to just provide a dashboard. Cool. So what? People are making decisions and they need to make decisions every single day. 

Lindsay DiPietro 0:13:58.5: 

So it's about having that intelligence from step to step, the guardrails of your entire process already. Already set within those applications so it can guardrail how it's making decisions or what it's recommending. What you end up doing is being able to surface the risks before they happen. You're able to surface recommendations on what steps you should take to remediate something, which in the past, you would have planners just spending so much time just digging around to find how these things are connected, or throwing something over the fence like 'Not my problem. Hey, I did my budget. You should stay within budget.' Well, that's not how the world works, right? Things change all the time, and it's changing so fast, and so you need this constant sensing so that you can take the actions that you need to take to move the business forward. So our applications out of the box come with these best-practice workflows, with the intelligence embedded in them, so that you're getting all of that out the gate and that you can see what the trade-offs are. 'If I take this decision that's going to have this impact or this consequence from taking it.' 

Lindsay DiPietro 0:15:25.4: 

I really wish I had that when I was younger. It was probably my parents, but it didn't stick too well. So that's really the workflow layer. It can go across. We're talking about retail planning, but obviously here it can go across more than just that. Retail connects into supply chain. Retail also connects up into finance. Finance connects to workforce, and you can just imagine your whole business starting to get mapped out through these workflows. So that's really the next layer, and then on top of that, imagine being able to start asking questions about your business, about things that are happening, not just in your area of the business, but in other areas of the business. 'Hey, what would happen if I took this sweater to markdown? Should I mark it down or should I do a bundle?' 'Oh, maybe you should do a bundle because you have incremental revenue and your margin upside is better.' Imagine having that feedback that you can get from, let's call them, agents interacting with all of the workflow and the models underneath. 

Lindsay DiPietro 0:16:34.2: 

So that's the next evolution. Once you have achieved the data layer, which is extremely important - data is a currency, that's what makes this all possible - you have the intelligence embedded in the workflow, and then we layer these AI tools on top that are able to answer some questions and then eventually provide recommendations, and then eventually take actions in those workflows or kick off those workflows for you. So that's the direction that we're moving. It's a pretty exciting journey, and I honestly wish I had tools like this back in the day, but I just had to click a bunch of buttons. 

James Theuerkauf 0:17:17.3: 

Nice. Very good. So I'll take us to the end of the structured part of the presentation, and this is really forward-looking a little bit. What happens once decisions are connected? What happens once, as you were saying, Lindsay, you move from this idea of a system of record to a system of cognition, or a system of intelligence? It's really three things. One, planning can be continuous. It's no longer a linear process that starts with a long-range plan, and then an assortment plan, and then an MFP, and an MFP assortment and then allocation. Actually, it turns into loops, continuous cycles. As we've talked about, there's extreme power and financial prowess in that. The second is being able to focus on exception management. What advanced analytics and AI tools are really, really good at is taking care of the long tail, of taking care of business as usual and taking care of the core components that maybe should be fine for system to handle, right? The brilliance of the human judgment is being able to focus on the exceptions, on the not-yet-repeatable. Let the machines do the repeatable. Let the humans focus on the not-yet-repeatable on the creative and strategic. 

James Theuerkauf 0:18:38.1: 

Thirdly, as you also mentioned, Lindsay, the trade-off analysis. There is not just one answer. It's all continuous. It's all a distribution. The future's uncertain. Deterministic planning systems, as we've had them for decades and decades, are not really suitable for a world where everything carries risk. There's a distribution curve around everything, and so that allows us to do the what-if work. 'What if we did this? What if we did that? What's the actual trade-offs there?' That's the world that we're going towards. It's a world of technology handling synchronization, and then humans being able to focus on judgment. It's a world that is maybe in the future, but Anaplan is building that today in the present. 

Lindsay DiPietro 0:19:27.5: 

So, we're here - I think if you were at the keynote this morning, you heard a little bit about this - specifically for retail, because, why not? We're here. We're talking about retail. I'm really proud to announce that we GA'd two products to join our MFP application. So we just GA'd assortment planning and allocation and replenishment, and thanks to the team that's here that worked so hard to achieve that. There's many people that came to contribute to that, so shout-out to them. That completes our mini three-legged stool. You have the three major components - MFP, assortment planning and allocation replenishment - and we've built that intelligence into those applications. We'll continue to evolve that as time goes on, and as Anaplan releases new features and functions. To us, this marks a really important part in our journey. We talked about that continuous planning and making that loop. We're now capable of doing that within these three applications, and for teams to have that trust in the systems that they use and to develop that rhythm, that's what winning looks like. 

Lindsay DiPietro 0:20:52.4: 

Just to go back to hockey, this is what winning looks like. So we're here to help you win. So we probably have a lot of time for Q&A, if anybody has any questions. 

Audience 0:21:06.1: 

Hi. I have a question on data ontology. When you spoke about that layer in that visual that you had, were you talking about data ontology in the out-of-the-box Anaplan applications, or were you talking about applying it outside, then bringing it… Just anything to clarify, where were you speaking about data ontology and is it a part of the applications, or is it an extra step? 

Lindsay DiPietro 0:21:36.1: 

So we have a unified data model that exists for our retail applications, and so out of the box we have a point of view on the data that you need. Obviously, every customer is different, and so what we don't know is what you call a style, or we don't know what you call a color, but what we do know across our applications, specifically for retail - and this is true in the other LOBs - is what data we need to make these models work. So that actually lives in our Anaplan data orchestrator layer. All of that is pre-configured, and then it's just about working with your implementation partner to discuss 'What does that actually mean? What levels of the hierarchy do you have? How does it map to our existing data ontology that we need?' Now, there's also cases where customers do that outside of Anaplan. Some of our larger customers actually have their own snowflake instances or have their own data warehousing, and then they bring it in and, again, map it to the specific ontology for the applications so that they work seamlessly. So does that answer your question? It almost does. Where else can we… 

Audience 0:23:02.5: 

You said about ADO leveraging data ontology part, where it helps in mapping and maybe providing insights. I didn't really follow that part. 

Lindsay DiPietro 0:23:32.5: 

The insights happen in the… I mean, there are going to be insights within ADO itself, but for now the insights happen in the applications themselves. So the data model, the unified data model, exists in ADO for the applications. So the customer will bring their data into ADO. It will map to that specific structure, and that is what feeds the applications, and then the intelligence exists in the applications. 

Audience 0:24:00.4: 

Okay. Thank you. 

Audience 0:24:06.3: 

Hi, Alison from the retail team at Anaplan. So talking to a lot of customers. They love the three-legged stool, 'We want to do connected end-to-end retail planning in Anaplan,' but a question we get time and time again is, 'Where should we start?' 

Lindsay DiPietro 0:24:21.4: 

Great question. I'll take it. I have also been on the other side of the fence trying to figure out where you should start. I don't always think there's a clear-cut answer. I would say probably a couple of things. One, if they're really concerned about achieving value quickly, or demonstrating value quickly, then a great place to start is allocation replenishment because you… It's one of the closest things to achieving results. You're immediately sending product to stores. You can immediately measure your impact. So we've had a lot of customers start their journey there, because what they're trying to prove is a larger business case as to why they should also use assortment planning, and why they should also use MFP. So that's one approach, is you take the 'How do I get the most value quickly?' Allocation replenishment can be a very good place to start. If they're looking for a more transformational journey and they already know they want to tackle all of these end-to-end processes, then my opinion is that MFP is a great place to start. It's more aggregated in terms of the data. Usually data ends up being a pretty big sticking point in a lot of implementations. 

Lindsay DiPietro 0:25:43.7: 

MFP is a really good, easy step into the entire planning process. It's the start, which is also really nice, and it flows nicely then to assortment planning and then to allocation replenishment, and so as people are looking at their processes, it makes sense to connect them one to the next to the next. So when you start with allocation replenishment, the downside to that is then you're starting to go more upstream into the process, back, and then you start thinking about the decisions you made here, and maybe you would have changed them because you would have done something different upstream. So I don't think there's one clear-cut answer. Again, if you're trying to drive value quickly, allocation replenishment is a great way to start. If you're looking at a whole transformational journey, I would suggest starting with MFP and working your way through. It's harder to prove the value immediately with some of those applications, other than time saving and making people's lives a hell of a lot easier. 

James Theuerkauf 0:26:46.3: 

I think that's right. I think the question, the question back is, 'What do you want to prove?' Exactly as you were saying, Lindsay, if it's, 'I want to show value quickly, and I need this to be ROI-positive because of a business case in month three,' allocation replenishment. The moment we send out the first allocation, you're making money. If you need to show to the office of the CFO that this is a cross-company endeavor, MFP is fantastic. If you're like, 'Hey, we're embarking on a broader transformation of our merch and planning organization,' assortment planning is an epic place to start. You can also do all three at the same time. Some of our customers, as you all know, are doing. I think it's really what you are looking to prove, and that then dictates what the sequencing is of our three-legged stool. Very good. Well, thank you all very much. This was great. 

Lindsay DiPietro 0:27:44.1: 

Thank you.