Bohan Liu 0:00:14.4: 

My name is Bohan Liu, and I'm the engagement lead at Keyrus. For those of you who have never heard of Keyrus, we're actually a global consulting firm who has been in the Anaplan ecosystem for over a decade. Personally, I have delivered Anaplan projects for some of the largest, most complex organizations, one of them being Meta. Today, I'm incredibly honored to be joined by Dan Marske, who will tell us more about the Anaplan presence at Meta. Dan, to get us started, do you want to tell us a little bit about yourself and your role at Meta? 

 

Dan Marske 0:00:45.8: 

Absolutely. Good afternoon, everyone. Really appreciate the amazing turnout. Love seeing a lot of familiar faces in the crowd. My name is Dan, and I am a business planning lead at Meta. What me and my group do is we work with various teams, all throughout the organization, that have some sort of a problem or a challenge, and they're looking for a software or technology solution to help them overcome that. Typically, they'll come to us, and we will vet out, first off, is there some sort of internal option we have that might solve their need, but if not, we work with companies like Anaplan to ultimately scope out, develop, implement, and deploy solutions for these teams? Really, all throughout the org, and from that point on, we help to maintain those, go forward, help them to continuously improve upon the platform, grow their roadmap over time. 

 

Bohan Liu 0:01:53.1: 

Amazing. To kick things off, Dan, could you tell us more about the Anaplan ecosystem that exists at Meta today? 

 

Dan Marske 0:02:00.6: 

I can. So I'd say the Anaplan and Meta partnership is really one of the all-time success stories. Anaplan and Meta have been working together for coming up on a decade now, which is, honestly, really incredible. So long ago, in fact, that the company had a different name. It was called Facebook at the time. You may have heard of it. Yes, what started out really as just one small use case within our sales operations org has grown and grown throughout the past ten years. To the point now where we are literally managing quarter over quarter, billions of dollars in ad revenue in Anaplan, and that starts with our territory and quota. It goes to building the book of business for our sales reps, and then ultimately, releasing these sales plans as early as we possibly can into the quarter, so they can go out and sell. So that was the original use case, and more recently, in 2022, that was our first foray into the workforce planning and contact center space, which is what we're going to talk about today. Beyond that, we're also looking at a few other areas, like budget management, OpEx, management, project management. So it's a really long and great relationship with Anaplan. 

 

Bohan Liu 0:03:29.5: 

So you mentioned workforce planning for vendor. Could you tell us more about what does vendor management look like at Meta? 

 

Dan Marske 0:03:39.4: 

It's very complicated, is what I would say. Essentially, every year, there are billions of dollars that get spent in what we call BPO, business process outsourcing, at Meta. There is a demand for resources to fulfill some sort of need that can't be met by the internal staffing capabilities of Meta. So what we do is partner with tons of vendors globally to fill those staffing needs. The group specifically that we're working with is focused on content moderation. It's focused on support ops, and it's focused on AI labeling services. I think probably the easiest way to think about what actually happens is, let's say that I have launched my new company. This is Dan LLC. It's a fast-moving company, and we're trying to get on the map. So we've signed up for some advertising on Facebook. There's a problem that happens with my ad. In Anaplan today, that's how we do our planning, to make sure that there's going to be someone available to go ahead and meet that incoming need. If it's a chat, if it's an email, anything. That's how we do that planning. 

 

Bohan Liu 0:04:59.3: 

Amazing. So before you transformed this process into Anaplan, what did the process look like for you? 

 

Dan Marske 0:05:07.8: 

Also, very complicated, but unfortunately, in a different way. So I think one of the, the real eye-opening things is when you come to a very large organization, these processes and these functions that maybe grow to an incredibly large scale, like we are today, where, again, it's billions of dollars being managed here. They started out small, probably in someone's spreadsheet, and as it grew organically over time, it stayed in the spreadsheets. That was the problem that we ran into, where it was tons and tons of disaggregated planning across business units all over the world, and we had really poor ability to measure, poor ability to forecast, and everything took way longer than it should have. 

 

Bohan Liu 0:05:53.4: 

Right. So it sounds like it's a process that incorporates a lot of different business user groups. Could you tell us more about what were the users that were involved in this vendor management process? 

 

Dan Marske 0:06:06.0: 

Definitely, so it has grown today to really a true end-to-end process. It starts out where we've got teams that are forecasting what the demand is that we're going to need. We've got folks that then manage the contracts, that allocate that out to the different folks delivering the work. We've got folks on the billing team that are managing the invoices, making sure that people get paid. Then, finally, we actually have representatives from each of our vendors that are literally in the same Anaplan model as our internal planners. So there's a back and forth on invoices that happens every single month. 

 

Bohan Liu 0:06:45.5: 

Right. So it sounds like back then, you needed a process to consolidate all these different groups, hundreds of internal users, and also hundreds of vendor users. What made Anaplan the right choice? 

 

Dan Marske 0:06:58.9: 

I mean, first and foremost, Anaplan, we have had the relationship with them for so long, it really was a no-brainer. They were already set up and knew all of our systems were set up to integrate with a lot of our internal tools, which is, we've got a lot of them. Secondarily, the time to develop and deploy a solution in Anaplan, compared to what it would have taken to build something custom to meet these needs, was just like night and day. As an example, one of the most recent units that we onboarded, and it was a pretty custom job. We were able to get them into Anaplan end-to-end in five months, which is amazing. Then, not only that, once it's up and running, comparing this to an internally developed and managed solution, we don't have to make the expenditure of having multiple software engineers who are maintaining this year over year, making the updates, managing the integrations. It's all in Anaplan, and we're able to be managed by our current ecosystem. 

 

 

Bohan Liu 0:08:07.0: 

You may have touched this earlier, before, but what does the process look like now in Anaplan? 

 

Dan Marske 0:08:14.0: 

Also still complicated, but a lot easier than it used to be. So yes, I'll walk through a little bit of what that end-to-end process actually looks like. So it starts with the actual, what we call case volumes. So when folks are reaching out, and that, again, comes in via a chat or an email, maybe by phone for some people. We bring that in, and then we feed it through Anaplan Forecaster, which is able to run through the nine different algorithms and project for some groups as far out as one year, what we think the incoming demand is going to be. That's across all of our different business units globally, across languages, a ton of different metrics that we look at. From that point, we layer in different assumptions, non-productive time that's going to be spent, shrinkage, all sorts of things, to eventually convert that from an hours requirement to a headcount requirement. How many FTEs do we need to actually meet that demand? From there -I told you this was complicated - we actually look at the supply available across our different vendors, and then allocate that out intelligently for the most cost-effective outcome, right? If this could be met by five different vendors, who's got the best rate? Who does it make sense for this work to go to? Then once that work's actually delivered, we bring it in. We're able to measure it and compile the invoice directly in Anaplan. 

 

Bohan Liu 0:09:44.8: 

Wow. Just hearing you describe it, it sounds like a complicated process indeed. A transformation like this must not be easy. Could you tell us more about the role that your implementation partner, Keyrus, played throughout this engagement? 

 

Dan Marske 0:09:58.7: 

Yes, I mean, much like Anaplan really was a no-brainer for this solution, and Keyrus is one of the leaders in the workforce planning space. I'm sure you all may have seen in the keynote earlier today, they've recently, in partnership with Anaplan, are rolling out the contact center planning application. So very deep knowledge and expertise. Were able to help advise us, not only on the nuts and bolts of the Anaplan implementation, but also, hey, how can we maybe improve our processes as we are maturing them, not just from a technology standpoint, but bringing it up to the 21st century? Wales 

 

Bohan Liu 0:10:40.7: 

So it sounds like you have achieved some remarkable outcomes, including saving OpEx and saving many hours, and also streamlined the whole end-to-end process. Could you tell us more about that? Like, what are some of the realized outcomes after this implementation? 

 

Dan Marske 0:10:57.7: 

Yes, I think we've got a couple of them up there on the screen. As I mentioned, this was a very, very disaggregated process. Tons of spreadsheets. Everyone had their own process. We measured the before and after impact of this, and happy, but also a little bit astounded to find out that we were saving over 1500 hours a month on running these processes. That is just like direct to the bottom line of internal resources that are having to execute these very inefficient processes time and time again. Secondly, one of the other downsides of managing everything in spreadsheets, is it really limited how accurate we could be when measuring the performance of our vendors? I mean, we quite literally were not able to accurately measure the contracts that they were executing against. We could only look at the month granularity in spreadsheets. Now, by going to Anaplan, with a single source of truth, we're able to calculate down to the week, down to the day, and in some cases, down to the hour of the work that's actually being performed. So much more accurate, correct costs on our payments. 

 

Bohan Liu 0:12:12.4: 

So, looking back at this end-to-end transformation journey, what were some of the one or two key lessons that you would like to share with the audience today, for some of those who might be thinking about going through a similar transformation? 

 

Dan Marske 0:12:26.5: 

Yes, I mean, I would say the chief, most important thing that we are thinking about with our current application, as well as everything else that we're rolling about, is flexibility. In this new AI world, which is transforming literally everything, especially in the workforce planning space, where we've got AI agents and ways to remediate these calls without getting a human in the loop, how can we make our software as flexible as the business process itself? So that when the business needs to change, we can change in Anaplan to keep up with that demand. As part of all the new applications that we're rolling out, we have designed every key component of our hierarchies, our core hierarchies, to be flexible, to meet, let's say, changes with the contract. So changes that might have previously taken months or months plus, can be knocked out in a matter of weeks. So much more flexibility and adaptability with the business. 

 

Bohan Liu 0:13:27.2: 

Amazing. Looking forward, what are one or the two things that you're looking forward to, and that's coming up in the horizon? 

 

Dan Marske 0:13:36.5: 

So I mentioned a little bit at the beginning about the budget and OpEx management. So really, this workforce planning component is actually only itself a piece of a much bigger puzzle. Ultimately, the work that gets delivered is the result of someone, somewhere internally, has got a project that needs to get kicked off the ground. So we're actually bringing that next level up into Anaplan, so we can have a complete view of the project from initiation. We're tracking the budget. We're doing approvals. We've got leadership in there. So they can have a complete end-to-end view of how their budget is performing, and how the work is being delivered, and yes, really excited about rolling that out in the next couple of months. 

 

Bohan Liu 0:14:22.1: 

Yes. All right. Thank you so much, Dan. Yes. Thank you so much for sharing this journey. We're now open for questions, if there are any questions from the audience. [long pause] 

 

Audience 0:14:40.2: 

Hi, I'm Jen. I am at Databricks right now, and we are also encountering different challenges around workforce planning. So it looks like you have a very massive data set, and also multiple different dimensions. Are you currently on Polaris to have that flexibility and possibly hyper-model? That's my first question. Second question. You talked about the flexibility around ever-changing dimensions and hierarchies. How do you actually technically address that? 

 

Dan Marske 0:15:12.1: 

So the first one is, we, on our original implementations, built those on classic, because that's when I think Polaris had just come out. We hadn't yet gone through that process, but all of the new ones were evaluating Polaris, and especially, like you said, the very massive data sets. If we do see sparsity, that's where we're going to look to deploy that. As we have more of these teams coming on board, and we have more of these reporting needs that are going to be, by the nature of how many different ways they might want to view these reports, looking to Polaris for that. Let's say that downstream components. Then, as far as the flexibility of the hierarchy. So, within Anaplan, I'd say these hierarchies are a reflection of the contract. So it says, this is the language that the work needs to be delivered in. This is the location, this is the region, this is the skill set, all these different components. So we've built it so that you can actually select those, and then with an effective dating, you can say, hey, we've switched our contracts over as of this day or this week, so forth, and so it's all timestamped, and it's built to flow and be managed by an admin to reflect those changes. 

 

Audience 0:16:35.0: 

One more question. Is this like an operational system? Oh, sorry, thanks. Is this more like for conducting operations? Like, or is it also a budgeting forecasting tool somehow? 

 

Dan Marske 0:16:45.5: 

Yes, there is a huge budgeting and forecasting component as well. Thank you for reminding me, because I would be remiss to those planners if I didn't bring it in. So it's not just for managing, let's say, the contracts and projects that are live, but we do have the ability for the finance team to actually forecast what we call internally - these are yet to launch contracts - and do the, let's say, analysis of, okay, what if we go with, let's say, this vendor, or this location, or this site, and so forth? 

 

Audience 0:17:15.0: 

We love to learn how you achieve. 

 

Audience 0:17:20.4: 

I heard you mentioned there are lots of vendor information that goes into Anaplan and internal user. Maybe some other inputs you are using. How are these various data getting integrated into Anaplan? Is it an automated connection to an external system, or is it manual? 

 

Dan Marske 0:17:37.9: 

Yes. Actually, the majority of our partners are conducting this work inside our own internal tool. So these chats are coming in, and it's because they're doing the work in this tool, it's able to track when did it come in the exact time and date stamps of everything. Then, yes, that is through an automation anonymized. It's put into a data warehouse, and then we bring it in every day. 

 

Audience 0:18:05.3: 

Sorry, if you can help me visualize. So the vendors use your internally developed tool and putting the information manually there, but does this internal tool send the data directly to Anaplan or through a database? 

 

Dan Marske 0:18:18.2: 

Into a data warehouse, and then it's into… Again, to clarify as well, that's the actual vendors who are performing the work are in, let's say, that tool. Then we also have vendors' representatives who are in Anaplan, and they're the ones that are reconciling invoices and making adjustments in Anaplan itself. 

 

Audience 0:18:37.5: 

So we set up licenses for those vendors as well. 

 

Dan Marske 0:18:40.9: 

Correct. 

 

Audience 0:18:40.9: 

Can I ask…? 

 

Dan Marske 0:18:42.0: 

There are licenses for those vendors. 

 

Audience 0:18:42.4: 

Okay. Can I ask the number of licenses that we have for the model? 

 

Dan Marske 0:18:47.2: 

I will say across the board at Meta, we're up to over 1000 users, I think, and 15 to 20 applications. 

 

Audience 0:18:57.9: 

Okay. 

 

Dan Marske 0:18:58.6: 

Then, specifically, I think, with the most recent one, we're looking at 300 vendors, approximately. 

 

Audience 0:19:07.0: 

And they will have like edit access. It's not just read-only. Is that correct? 

 

Dan Marske 0:19:12.5: 

Correct, yes. Another good point to bring up is, obviously, something that we're very careful about is, we have, in all honesty, competing teams that are all working in the same shared space. So a lot of effort has had to go into the security and access, and to make sure that each vendor can only see their own company's information. 

 

Audience 0:19:35.0: 

Okay, thank you. 

 

[long pause] 

 

Audience 0:19:44.5: 

Curious on that security side. Are you managing that with roles, or in a different way? 

 

Dan Marske 0:19:51.8: 

Pretty much like, if you were to go to Anapedia and look up security, each and every thing we're using in one way or another. I will say that a lot of the really core components are rolling up to a vendor level, and so there is that like, list level access that each vendor only has access to their own. That's the ultimate fail-safe, let's say. 

 

Audience 0:20:17.3: 

Then a bit of a pivot. How are you going about analyzing what should be built in-house at Meta on Anaplan versus using key partners like Keyrus? 

 

Dan Marske 0:20:29.3: 

I think, first off, having a partner is a really great way to augment your own capabilities. I am not a classically trained workforce planning expert, but Bohan and the folks at Keyrus are. So really, whenever possible, we do like to rely on the expertise of our trusted partners to help us improve our own processes. It really depends on the scale of the project. Given infinite time and infinite resources, we could build anything. It's about when do we actually need this tool implemented? What's the budget that we have? There's some of the economics that go into that, and then we do is, I think we talked about it earlier, like a build versus buy analysis, where we'll crunch the numbers of what it would take to implement, and, yes, go through that. 

 

Audience 0:21:24.9: 

Hi, I have a question. You mentioned you have such huge models, just wanted to know, which integration platforms are you using across those models? The one which we are using is not that great, so just wanted to see! 

 

Dan Marske 0:21:41.2: 

Okay, I hope we're not using the same one! 

 

Audience 0:21:44.3: 

Hope not. 

 

Dan Marske 0:21:46.7: 

We are using Cloudworks for a lot of our integration, and we've started to roll out the Anaplan data orchestrator as well. We also have a big, very robust, very talented internal team of data engineers that run via custom API, a lot of the integrations as well. Essentially, you could think of it as like, anything external is probably an API, and anything moving data around within Anaplan is probably going to be Cloudworks or EDM. What are you using? Can I ask? 

 

Audience 0:22:22.2: 

Ricardo. 

 

Dan Marske 0:22:23.4: 

Ricardo, okay. We'll talk after! 

 

[long pause] 

 

Audience 0:22:39.9: 

So you said a lot of these initiatives used to live offline in [unclear words 0:22:45.4:] cells, right, in various formats and forms, and different shapes, I believe. 

 

Dan Marske 0:22:50.5: 

Shapes and sizes. 

 

Audience 0:22:52.9: 

Did you have to go through the groundwork of trying to standardize them and come to some kind of alignment first, or did you go ahead and start building them in Anaplan first, and go through iteration over iteration, to get to a standardized process in the end? 

 

Dan Marske 0:23:09.6: 

I mean, either foolishly or not, we decided to really go all the way in, with the initial implementation. So we knew that even if we were to fix the process, just by nature of keeping it in spreadsheets, right, it's still not going to be the most efficient thing. It's not going to be able to change and flex in the way that we needed it to be. So yes, from the beginning, the design was always, let's get this process in there, and we'll refine it as we go along. Typically, the way that we'll go about these implementations is we'll go through a lengthy, let's say, several-week process to fully define exactly what are the requirements, and that's where we go through that refining of that end-to-end process, and that's the blueprint that we end up building from during implementation. 

 

Audience 0:24:01.1: 

Does your team have internal modelers or it's all done by Keyrus? 

 

Dan Marske 0:24:06.9: 

Yes, we have a relatively small - in fact, they're all in this room right now - team of internal model builders. It is me and front row, Jackie, and beyond that, we do rely on external partners, both Keyrus and a few others, for those outsourced, let's say, build. If that's run, support, if that's net new implementations, it really depends. We'll work with a lot of different folks. 

 

Audience 0:24:36.5: 

Okay, so Keyrus is not just supporting from an initial architect perspective. It's also like BAU support. 

 

Dan Marske 0:24:41.7: 

That's correct. 

 

Audience 0:24:42.7: 

Okay, thank you. 

 

[long pause] 

 

Unknown Speaker 0:24:53.0: 

Any other questions? You've got Keyrus here. You've got Meta, one of the big companies around. A lot of things you could ask. 

 

Audience 0:25:03.4: 

Curious if you could talk to us a little bit about the interconnectivity of your 15 to 20 apps. Are they all their own thing? Are they talking to each other? 

 

Dan Marske 0:25:11.5: 

Yes, they are. I would say the easiest example to draw is, we chatted about briefly at the beginning, which is in the sales operations space. So if anyone knows, like SPM, this is an incredibly interconnected process, where it starts out with setting the territory and quota, and all the way down to the sales reps getting paid. The way that we operate is on a quarterly cycle, where actions happen in one model that gets passed off to the next, and the outputs of that model are the outputs of that one. So, yes, and then there's, let's say, different various windows where they're open for users to plan and make their decisions, and it's a complicated dance, but yes, they all talk together. 

 

Audience 0:25:59.6: 

So I had a question. How many of your organizations work with different vendors from contractor perspective? What would be the typical, I think, vendor experience when they come into Anaplan? What do they see? What are the different things that they can take action on, and how do they communicate with you if there is something that they see is not correct? 

 

Dan Marske 0:26:25.5: 

Yes, that's a great question. Also, by nature of the security, we do ensure that each of our vendors has a role that is assigned, and what dashboard pages they're able to see is limited. So they're only able to see exactly what information it is they need to execute their job. Then, for the notifications, we've got Anaplan Workflow. If you all have access to that, which is able to, once something is done, we're able to, on a trigger basis, let's just say, hey, something is pending for me. It'll send out a notification, or they can press a button and type up a message, and that will then get routed to whoever the correct internal resource is. So yes, they're all in Anaplan, and it's the same dashboarding experience that the rest of our org uses. It's just their own little slice of the pie. 

 

Audience 0:27:24.8: 

Great presentation. With all the vendors, do you drive any invoicing out of Anaplan? 

 

Dan Marske 0:27:32.0: 

We do. Yes. As a matter of fact, each vendor, after going through a process to, let's just say, if there's any adjustments that need to be made, maybe someone from Meta said that we needed to spend 100 extra hours doing X, Y, Z activity. We'll agree upon the final approved invoice, and then that literally is exported from our Anaplan model to our internal procurement system. So that's incredibly connected, and it's, yes, Anaplan is the source of truth. 

 

Audience 0:28:07.9: 

Back to that vendor experience. Do you have the vendors ask, could they create their own Anaplan models, so that they could use that to forecast their resources that need to be deployed? 

 

Dan Marske 0:28:18.5: 

We haven't been asked yet. I think the answer would probably be no, but we do, in all seriousness, take our vendor feedback very seriously, because it is a partnership. A lot of times also, they will have… These folks are using the Anaplan application, sometimes more than people internally, and so we do take and listen to their feedback. A lot of, I would say, really great improvements and changes have come from suggestions from our vendor partners. 

 

Audience 0:28:53.5: 

Can you talk about any change management challenges and how you overcome them? 

 

Dan Marske 0:29:00.4: 

Yes. So it sounds like you have some experience with this. Definitely, as we have grown, and more and more teams are operating within the same tool, and we have more and more vendor partners that we're working with, and this huge ecosystem, this huge volume of users, every single change needs to have a really high level of scrutiny. A small mistake, from an Anaplan perspective, like one wrong formula, could be truly devastating in terms of security. So we do have all of our models are up on ALM, and we do have a, let's say, code review process, scheduled deployments, and really, if anything is flagged, or if there's anything that's touching, really, one of our core reporting lists or modules, that is picked over with a fine-tooth comb. Yes. 

 

Audience 0:29:59.2: 

So, speaking from the change management perspective, do you ever have users saying, no, I don't want to use the tool? I still just want to use my Excel. 

 

Dan Marske 0:30:07.2: 

Sometimes, yes. At the end of the day, I think there is always a bit of a friction to do what's new and to change your process. If I've been doing something the same way for a long period of time, that's always a challenge. We spent a lot of time on the training, materials, documentation. We have open office hours all the time, where we're spending hours of the week with new folks to the tool, older folks in the tool, and, yes, I mean, it's a challenge. Ultimately, I think training and working with folks is a remedy. 

 

Audience 0:30:49.7: 

Just a follow-up question. Since you mentioned that Anaplan is involving [sic] in invoicing, does it mean that Anaplan needs to follow Sox compliance and is subject to audit? 

 

Dan Marske 0:31:01.7: 

Yes. 

 

Audience 0:31:02.7: 

So that means you have lots of documentation you have to do. 

 

Dan Marske 0:31:06.1: 

That's correct, and then we talked about workflow and approvals in some way. That is, yes, one of the really key components is that there is effectively someone internally who is giving that final stamp of approval before anything is allowed, that the vendor has been working on, to then go into a financial system. 

 

[long pause] 

 

Audience 0:31:36.7: 

I'm curious a little bit. Are all of your solutions more demand-driven from your users and internal customers, or are you and your team going out and finding the use cases, finding the process gaps that can be delivered on Anaplan, and building the roadmap that way? 

 

Dan Marske 0:31:56.8: 

So I'd say it's a bit of both. Meta is such a large organization that it's difficult to know what challenges all of our different teams and groups are having. When we hear about something, and when we vet it out, it's really common that this isn't just a problem that this one group has, right? This is like the probably 50 to 100 plus, who knows, teams are experiencing the same challenges and difficulties. So, that's what we'd like to do, is build up a core solution that's proven. It's been adapted internally, and then we do look out for those other groups and try to bring them those proven solutions, and once it's been built, the time to deploy and bring them on board is a lot quicker. 

 

Audience 0:32:50.3: 

I'm curious, the vendors you're referring to here are only the BPO vendors, or all the different vendors? 

 

Dan Marske 0:32:56.8: 

Just the BPO vendors. Essentially, within Anaplan, it varies from based on the size of the vendor, but I'd say between one and ten at the most vendors per organization, that is in the tool, and those are the folks that are managing that process. 

 

Audience 0:33:16.6: 

Would you say there's plans to incorporate different type of vendors or contingent workers later on? 

 

Dan Marske 0:33:25.4: 

I mean, I think so in just the same way that we don't know all the different groups that might benefit from this solution, as can we scale different parts of the business. I think it's the same use case. It's a contact center workforce planning thing, but the actual work that's being delivered can be, like, wildly different based on the team, like we mentioned, right? We've got support ops, we've got content moderation, and even AI labeling services. So yes, types of vendor. Absolutely. 

 

[long pause] 

 

Audience 0:34:10.0: 

Thank you for sharing. This is a great story. I'm curious, in the improved billing accuracy, what percentage have you improved, if you could share specifics? 

 

Dan Marske 0:34:22.4: 

I think we're looking at a bullet point. The billing accuracy. 

 

Audience 0:34:26.9: 

Yes. Billing accuracy. 

 

Dan Marske 0:34:28.5: 

Yes. So I don't know if I can give you a percentage point, but it's more in the actual granularity that we were able to enable by moving to Anaplan. Previously, because of the disaggregated nature of just being in all these different spreadsheets, everyone has their own process, it truly was only possible for us to measure performance to a month level by going into Anaplan, and we have these actual work performance coming in. We can tell, hey, this was delivered on Saturday, and it was nighttime, or it was daytime, or it was a holiday. All these different metrics, which have different rates attached to them, and we can much, much more accurately and confidently calculate what's owed for the work that's been delivered. 

 

Audience 0:35:23.5: 

Thank you. 

 

Bohan Liu 0:35:26.7: 

So, in other words, like, previously, you know how much they worked throughout a month, but to effectively support someone that's coming through a chat… We'll have ten people calling at 9:00 am on the Super Bowl day. You have to have enough people to staff for that particular time period. So if that compliance was not met, then you actually can't, shouldn't build in a way… That's the sense of getting to the granular, interval level. It's really ensuring that, okay, vendor, dealing with the work at the time and at interval level that you expect them to deliver. 

 

Dan Marske 0:36:05.0: 

Yes, and that's actually another great call out too, just about something like the Super Bowl. By coming in into Anaplan, we're also able to much more accurately consider and plan around, not just holidays, but what we term as atypical events, is the internal term. You can imagine that, yes, the Super Bowl, there might be more folks that need TV advertisements, or if it's Black Friday, that could be an influx in the US of expected demand. So, yes, much more confident, accurate forecasts and measurement of the work. 

 

Bohan Liu 0:36:49.4: 

Okay, maybe last one. 

 

Unknown Speaker 0:36:50.2: 

Couple more minutes, if there's any more. Any more for any more. Last chance. A lot of great questions. We're a couple of minutes early, but that's okay, because that gives us time for a coffee. Oh, we have another question. 

 

Audience 0:37:05.9: 

We have more time, so what else are you looking to optimize next? 

 

Dan Marske 0:37:11.9: 

The upstream piece that we had mentioned, which is the starting point for a lot of the work that's being delivered. That is a net new use case for us, at least in Anaplan, and hopefully going to be going live in the next couple of months. Our objective is by getting that starting point, let's say, of this work, into Anaplan, we'll hopefully unlock, I guess, better visibility, really, into, like, the end-to-end pipeline, what the demand is going to be across the board, and, yes, just really excited about that. 

 

Audience 0:37:56.9: 

Have you been using Keyrus all the time? 

Dan Marske 0:37:59.8:  

For the workforce planning component.  

 

Audience 0:38:03.5: 

Or anything else?  

 

Dan Marske 0:38:04.5: 

Not everything Anaplan. So I haven't actually been here ten years, but I think there's been a few different vendor partners. Yes, right now, Keyrus is the provider for most of our BAU and run support for our live applications. 

 

Audience 0:38:23.7: 

Is there anything about them that stands out for you? 

 

Dan Marske 0:38:26.6: 

About Keyrus? Wow.  

 

Bohan Liu 0:38:32.1: 

I promise [over speaking 0:38:32.1:]. 

 

Dan Marske 0:38:32.1: 

We only have a few minutes left. I mean, in all honesty, up here, we've got Bohan, who's been really an incredible asset to our team, from the Keyrus side of things. Was one of the key - been very humble - but key solution architects and minds behind this application as it rolled out and helped come up with the more creative solutions, let's say on the ways that we went about things. Yes, I think it's also been a huge reason why we're continuing to iterate and improve upon these models and processes. So yes, it's just been a very good partnership. 

 

Bohan Liu 0:39:14.8: 

Thank you, Dan. I appreciate that. 

 

Dan Marske 0:39:16.9: 

You're welcome. 

 

Unknown Speaker 0:39:19.8: 

Great way to wrap it up. Thank you, both. Thank you for all your great questions. We're right on time. Thank you very much, Dan. Thank you.