Plan It Forward! A Live Finance Conversation with McAfee & Activision.
Jeff Brobst, VP FP&A at McAfee, shares how he and his team escaped the “one size fits all” fate of financial planning by tailoring their Hyperion implementation with Anaplan. Patrick O’Brien, Senior Manager of Finance, explains how Anaplan has contributed to increased communication and 50% gains in productivity. Now, their teams both have a single source of truth in the cloud – one that doesn’t require IT maintenance or consuming updates – so they can rapidly build, change, and share models at the speed of thought. Learn how you can too!
VP FP&A at McAfee
Senior Manager of Finance, Activision
CMO at Anaplan
October 22nd 2013 Video Webinar
Mark Sarbiewski: Good morning, good afternoon, or good evening, depending on where you are in the world, and welcome to today’s web webinar Plan it Forward: A Live Finance Conversation with McAfee and Activision, brought to you by Anaplan. I am Mark Sarbiewski and I am your moderator today. We have just a few announcements before we begin. This webinar is designed to be interactive between you and the presenters. The webcasting console can be resized to fit your entire browser window. All the components and widgets on the console can be resized as well. They are moveable, stackable, resizable and minimizable. The Menu doc at the bottom of the screen shows icons for all the widgets that are part of the console. The Menu doc allows you to open or minimize a widget. This webcast is highly interactive and we encourage your participation during today’s event. Please use the Q&A widget to submit your questions to today’s presenters. Simply enter your question into the text area provided and click ‘Submit’ to send your questions.
So with that, I want to welcome everyone to this webinar, Patrick and Jeff here as well, for this Live Finance Conversation with McAfee and Activision. Before we jump into some background from both Patrick and Jeff, I want to just give you a word or two about Anaplan. We are a business modeling and planning solution but relatively new in the market. We have been out for about three years commercially and we are a venture-backed company, with some great backers, and we have more than 100 customers. We are really the fastest growing business modeling and planning solution in the market right now. And we are not going to do a demo today or give you lots of screen shots; this is a live conversation. But I wanted to give you just a sense of what the Anaplan solution is like, to give you some context for the conversations that we are going to have.
The first thing to picture in your mind is a spreadsheet like interface. But instead of the C2-F4 it’s really based on intuitive business rules, so it will be revenue minus cost of goods equals margin, etc. And really you have that spreadsheet-like immediacy, so that business analysts and model builders are using that interface to build essentially whatever model they want. So that’s #1; the very intuitive interface. #2 is the In-Memory Calculation Engine. We combined the best, it’s a patented engine, it is a hugely powerful capability that’s built from scratch to solve these very complicated business modeling and planning problems. It combines in-memory database technology, spreadsheet, cell-oriented technology, cube, the best of cube and even what I would call graph database, so relationships between entities. The net effect is you can really build many incredibly powerful multidimensional models with up to tens of billions of cells. So as we will hear both from Patrick and Jeff, you have to almost rethink what’s possible in terms of modeling with that engine.
And then the last piece to keep in mind is it’s all available as a service. So it’s delivered through the Cloud and the Cloud isn’t what’s special about it, it’s really the engine and the interface, but the Cloud does enable everyone across an enterprise to engage on that same model: anywhere, anytime access, portable devices as well as obviously computers/laptops, and to interact with the same set of data. So we see, and you will hear today, how hundreds and sometimes thousands of people are interacting with a plan with actual data coming into that plan. It’s really a forward-looking view where you have collaboration happening across your enterprise and that’s what’s kind of cool about the SaaS nature of our solution. We will send out after the webinar links to videos; you can see the products, you can do all of that. We didn’t want to take the time today; we want to get into the finance, future and how it’s being used. So you will see that, you obviously can get it at www.anaplan.com. But keep those things in mind: that spreadsheet interface, the calculation engine, all delivered through the Cloud.
And what we want to talk about today is really what’s changed in finance in the last few years with some specific context from Activision and McAfee. How can those connected operational plans be even more strategic to the business; how does this fit with IT? We invited folks that are Hyperion customers today, as both McAfee and Activision are, to address: how does this fit with Hyperion? Hundreds, thousands of companies in the Fortune 1000 have Hyperion, so we want to explore that, and then discuss what some predictions for the future are. So I want to start the agenda here today with Jeff Brobst, who is the Vice President of Financial Planning and Analysis at McAfee. We will move on to Patrick so they will each give some context and background, the problems and the challenges, the opportunities they saw, what led them to look for a different approach potentially, and then we are going to open it up after those sessions to a roundtable discussion – and we really, really want your questions. You have access to two great minds here, lots of good experience, so please do send in those questions and we will have a very nice and fun live discussion.
So with that, I am going to pass it over to Jeff here and have him give us some background. I think you all know McAfee, but maybe a word or two on that for those who might not be familiar with its position in the market?
Jeff Brobst: McAfee is the leading dedicated security software provider in the world, with 7,000 plus employees and near 3 billion in revenue.
Mark Sarbiewski: Outstanding. So what were you facing way back when? Just as a little bit of background: McAfee has been working with Anaplan as one of our earliest customers. You have a couple of years under your belt, whereas Activision is relatively new. So we are going to get both of those perspectives. But I am curious Jeff, the problem that you were facing that even led you to go looking for something different.
Jeff Brobst: Well if you look at our journey, we essentially had a bunch of processes that were burdened by spreadsheets. I mean massive amounts of data, a planning system that wasn’t flexible enough to handle really the complex calculations that a spreadsheet can handle, and then processes that take three months or longer when it should take much less time. So that’s what we were looking at when we were looking at Anaplan, essentially to streamline our processes first across sales finance and modeling and target assignments and so forth. And once we got that going and really became comfortable with Anaplan’s capabilities, we shifted to building commissions application. And that’s probably a crown jewel of all of our applications because it’s really state-of-the art and we have worked with a lot of different applications in our past and we think it’s better than all of them bar none.
So then we looked at the corporate FP&A side: bookings, forecasting, revenue planning, workforce planning and so forth. We got to the point where we were asking, ‘really what do we need Hyperion planning for?’ because these systems essentially became subledgers Hyperion planning. Hyperion planning essentially has become more of a corporate compliance tool more than anything else.
Mark Sarbiewski: Right, interesting. I think you touched on one of the problems that actually attracted me to Anaplan. My background is in products and I have been spending a long time with sales forces over the years trying to get them to perform at a different level. And I think one of the things that you touched on there was there is an aspect of that that’s an analytics problem, right, which is how do I get the right reps, the right targets, the right products, the right quota, all of that and bring in history because you have a lot of good information. That’s kind of one of the areas where you guys have explored.
Jeff Brobst: And it’s fundamentally an iterative process and you can’t do that in spreadsheets. Essentially you have these massive spreadsheets and probably have to segregate them first by feeder and then by team and then you have to get it back and when you get it back, you don’t like the answer so you do it again, I mean that’s part of the three month process, right. So with Anaplan, essentially you have a spreadsheet that’s subscribed to all the data sets you want, I mean millions and millions of rows; you don’t have to aggregate themselves. I mean that was one of the biggest part was we were spending 80% of our time gathering and transforming data and not really doing the analysis. So we have kind of flipped that. I mean the data is all available, the model is subscribed to corporate certified datasets, you use what you want to use to get what you need to do and you are all working at the same time.
So at the end of the day, you have a lot more agility. The agility is not only in how the application works but the fact that in, again, for us Hyperion planning, it is essentially administered by IT types and that’s their job. But you wouldn’t expect them to be experts of revenue recognition or how to pay commissions or something like that. If you think about the lifecycle of building a model, it’s a business requirements documentation – ‘let me explain’ and ‘no that’s not what I wanted,’ ‘hey can I just do it myself,’ ‘that’s what I do in spreadsheets in the first place,’ right. So you are getting to that model. Plus there is constraint with upgrades and so forth, and for us, we end up having just a huge backlog of requirements. And so that agility is basically people who are closest to the data are doing the models. Basically the people who were managing the spreadsheets in the first place, you give them Anaplan. And then you look at the performance side of it: There are no calc scripts; it’s an in-memory calculation. So you hit “Enter” and basically your whole model – and we are talking billions of cells – are automatically updated. There are no calc scripts, you know exactly what you have, and you don’t have to go through this process of sending data around. It’s basically always available to you. So you have very transparent processes that are owned by the business.
Mark Sarbiewski: Right, right. And you mentioned something to me last night – I think I see it on the slide here too – is that there are nuances across these different units and there are different user types that you are trying to bring into the process here and that’s something that you found compelling as well.
Jeff Brobst: You wouldn’t expect R&D to plan the same way that sales plans or the same way that facilities plan. I mean it’s inherently different but these tools, the ERP based tools essentially try to force you to one model.
Mark Sarbiewski: To a particular way.
Jeff Brobst: And so what we found is again 90% of the people in kind of Op-Ex planning were essentially doing it on spreadsheets and then uploading into Hyperion. So it’s really not any added value planning. I mean essentially that’s what it gets you at the end of the day.
Mark Sarbiewski: Interesting.
Jeff Brobst: So it’s been phenomenally successful for us.
Mark Sarbiewski: And how so? Give me some concrete stuff.
Jeff Brobst: Well, so I mean our quota planning has gone from three months – and that’s a stretch, sometimes it would take even longer than that – to three to four weeks because essentially you are iterating all the time. Our revenue forecast accuracy this last quarter was the best at Intel, was like 0.1% variance. So across of all of Intel, a huge company, we have the best forecast accuracy because we are using these tools and we are iterating and we have access to the lowest level of information to form calculated, smart decisions. So for us, it’s really about driving added value as finance people, going from gathering data and being reporting people to actually driving behavioral changes through this. I didn’t mention that we have also been doing operational planning in the forms of deal desk, basically being able to optimize discount because we have the entire history of our transactions. They can really look at, hey what makes sense from a budget standpoint, the margins we are trying to target, competitive pressures. While salesperson is on the phone, they literally can plug in that information. So it’s really streamlined us. At the end of the day, we are producing more with less people.
Mark Sarbiewski: So it’s interesting, I am curious to know how it went from one use to another. It got started with one thing and I think you did a good job explaining that and then how did people then say wait a minute, I can use it for this or get picked up by other use cases?
Jeff Brobst: It doesn’t take long when these people see it to have that ‘aha’ moment, that light bulb goes off and say, ‘what, I don’t have to run those reports out of SAP to get this information?’ No, it’s in a hub. You have a model, which essentially looks like a spreadsheet but you can subscribe to any data that you want, and then wait, I can use this for X, I can use this for Y. And so it’s really proliferated.
Mark Sarbiewski: Your data point is an interesting one which is when I was looking at the company several months back I loved the fact that it was kind of an action-oriented thing. You are bringing in actual data from the existing systems you have and using that to sort of see where you are essentially in real time, like track your way forward as opposed to just keep score what happened at the end.
Jeff Brobst: Exactly. It’s driving behavior; our data looks much better. I can’t probably disclose how much but let me tell you we are doing well.
Mark Sarbiewski: Right, that’s fantastic, excellent. So that’s context for McAfee and what you were working on. I am going to pass it over to Patrick here, for a little bit of background from you on Activision. I think a lot of people will know all your game titles maybe more than the company.
Patrick O’Brien: Yeah, absolutely, thanks Mark. At Activision, we are a global video game publisher, with over four and a half billion in revenues last year. Some of the franchises that folks might be familiar with are Call of Duty and World of Warcraft and the Skylanders. So we have been doing very well and as we as a business transition from just the retail primarily business to more and more digital, every dollar counts and our data and our accuracy has to be that much better for us to deliver better ROI. So that’s a real focus for us. And in contrast to McAfee, we are very new to Anaplan. We just went through an implementation process; we are about a month in and I can say so far so good.
Mark Sarbiewski: You’ve been live for a month, right?
Patrick O’Brien: Exactly. And so far so good, and we still have a lot to learn. But before I talk more about what we are doing with Anaplan, let me comment on how we got there. Back in 2012, we had an issue with our current trade promotion management system. The integrity of its data wasn’t very strong and at a critical time it failed us and it was a huge challenge to overcome at quarter close. So at that point, we made the decision to look at other vendors that maybe might have a better system that we could use and allow us to do more things. And so we engaged a consultant that did a global search for someone in the TPM space. Ultimately, they landed on Anaplan which we were really attracted to because #1 it’s very flexible and agile, and #2 if you are familiar with Excel it’s pretty easy to transition to Anaplan, and that’s the great thing. So everyone works in spreadsheets across corporate cultures and that’s just part of what it is.
But the great thing about Anaplan is that it takes advantage of that knowledge base in the models that it builds. And so that was a huge starting point for us and what really attracted us to Anaplan. Really we had four disparate user groups who used Anaplan, the same data but in different ways. We had the sales field force who obviously needed to execute promotions at retail on a real time basis, we had retail managers at headquarters locally who needed to manage their budgets and communicate plans to the field, you had finance who needed to analyze the data and perform ROI analysis on them, and then of course, you had accounting on the back-end that needed to reconcile all the activity that we had done. And the great thing with Anaplan is that allowed us to serve all these disparate groups’ needs while also being very easy to administer, to setup, and to train people on. And part of that is its Excel based kind of roots.
Mark Sarbiewski: Yeah, it’s Excel with an ability to tailor how people interact with it. So they see a view that makes sense to them, their dashboard, the data that they care about and you control that. You as the model, the people that are putting the model and plans together can help work with these different groups, right?
Patrick O’Brien: Right exactly. And so as I have mentioned earlier, we are only a month into our journey, and we are really excited about what we can potentially do with Anaplan and what models we can potentially build out, not just for the finance team but also for the retail marketing team, for the sales team and perhaps, there might be other applications across the company. We are not there yet. But in the meantime, we are enjoying a nice clean user interface. I think I have called it about a 50% gain in productivity on the execution to what we use Anaplan for and tracking our activities at retail. It’s been a big success for us so far. It’s early days yet but I know the users are very happy with it. I think the biggest asset to us is that it allows us to be flexible, and if we don’t like something, we can change it real time, we don’t need to wait on IT or wait even on Anaplan to do something for us; we can do it ourselves. So it really empowers each group and we are off to a good start.
Mark Sarbiewski: Yeah. And I think your quote at the top of this gets at one of the points that I would actually like both of you to comment on which is that I know there are a lot of folks in the audience that have a lot of familiarity with Hyperion and other solutions that are out there. And so I have heard it over and over from people, you almost have to think a little bit differently with Anaplan and of what’s possible, and I am curious if you could give us some examples. Let’s start with Patrick and then maybe Jeff you can think about that.
Patrick O’Brien: Well yeah I think the less time that you have to spend on finding the data and aggregating the data… I mean the great thing about Anaplan is that it’s all there real time, and that allows you to spend time on other things like looking at the data differently and exploring different models. So that’s been a big plus for us.
Mark Sarbiewski: I think too I heard maybe it was you Jeff that said it’s something you would never do in potentially other systems. Run a bunch of what-ifs, build some models and throw them away if they are no good, use them.
Jefrf Brobst: This is your spreadsheet. Save as, I have got a new model, don’t like it, delete it. I mean you can’t do that in a regular ERP application about throwing everything off I mean. So just again, part of that agility is you can test on-the-fly essentially without disrupting anything or breaking anything because you have that capability with spreadsheets but you are not having to gather the data and you have available millions of rows of records to you at the same time.
Mark Sarbiewski: And both of you talked, another thing, you are bringing in users that aren’t planning experts; this isn’t where they live. Now there is obviously that familiarity with spreadsheets. I think that’s a pretty good knowledge base across companies, so there is some translation there. But how did that go, right? What was the thought process of like ‘okay, I really want to get them involved;’ how did you make that happen with these plans you put together?
Patrick O’Brien: Yeah. So when we began the implementation process, we took the extra time upfront to solicit feedback from all the user groups to understand what their real needs were. And again, the great thing with Anaplan was that we are all out, we are able to create different dashboards and different views depending on the perspective of the user. And so each user group, and within each user group, each user can have their own customized data or dashboard. And doing that upfront was a good move for us, it made the implementation a little bit longer than we were expecting but I think at the end of the day we got a result we really liked. And the great thing is that if we want to change things in the future, we know how to do it and we can do it.
Jeff Brobst: Right, right. We have role-based views so a sales rep sees something totally different than the VP, different than the finance analyst. They are all different skill levels. But it’s all in the same model; we don’t have to have separate models for each type of skill level. They essentially log in and they get what they really need.
Mark Sarbiewski: Yeah. One of the things that I hear a lot is that there’s faith in the data, the confidence because you subscribe to it, it’s all the same data, it’s a different view of the model but it’s that same data.
Jeff Brobst: I mean spreadsheet errors are legendary, right?
Mark Sarbiewski: Yes, yes.
Jeff Brobst: So you don’t have that.
Mark Sarbiewski: Absolutely.
Jeff Brobst: You put your logic into Anaplan and it’s done; you don’t have to worry about cell references or anything like that.
Mark Sarbiewski: Absolutely. So I think we are going to move on – we have already transitioned into the roundtable part of the discussion here, and there are some questions coming in. I know there are some that people will ask. We get them all the time, and one is on security. So who better to answer that than McAfee? You guys must have looked at this hard, how did you look at that whole situation?
Jeff Brobst: We take security very seriously. And our IT, our information security group is just topnotch. I mean they really tried to break it and they couldn’t; they passed with flying colors. So if you think about it, a web-based tool is going to be inherently more secure than spreadsheets because spreadsheets can fly around, you don’t know who has them, right, so they are inherently insecure. So this is much better for us.
Mark Sarbiewski: Interesting. And we always hear about okay implementation. So you touched on it a little bit Patrick. I think as someone who has just gone through that first one, you talked a little bit about taking some extra time because you had these different users but what surprised you; how did that go for you?
Patrick O’Brien: It went well. I mean I mentioned it took little bit longer than we initially expected, but we thought it would be three or four weeks and it took three months. I think in this world that’s pretty good.
Mark Sarbiewski: It’s lightening speed, right?
Patrick O’Brien: Yeah exactly. So I think the more time you can spend upfront getting the right structure and hierarchy and the back-end right so that the models can all work with each other, I think the better off you will be. And it’s been a very iterative process and I don’t think we are ever done; we are always going to tweak and adjust on the fly as we need to.
Mark Sarbiewski: Yeah. Again it kind of goes with ‘what’s a little different about Anaplan?’ You brought up two points that I will hit, which is that we have more and more templates that accelerate that process, and we do the implementations in an agile fashion. Our goal for some of those first models is a couple of weeks. But the more sophisticated you get, and I liked your points about thinking through the different roles and how this is all going to fit together, you take a little bit of extra time, make it two or three months. And you are building a model that’s yours, that’s very tailored exactly to what you want that you then own beyond that. And that actually leads me to the question of, ‘okay, what’s the role of IT in this?’ You are both obviously pulling in data from other systems and working with IT; what part of it is really under your control and where does IT fit?
Jeff Brobst: ETL and security really is where IT plays in this and that’s where they want to play, right. They want to live in the data architecture world, the data that we are gathering. They don’t want to do reporting, but quite frankly it’s the same as that other scenario. They don’t understand financial reporting and they are not meant to be; you wouldn’t expect them t. So no, you put the models in the hands of the actual business users; IT is kind of free to doing what their core competency is.
Patrick O’Brien: Yeah and I like all those sentiments. I mean the ability to make the changes that we need for our business purposes on the fly without having to rely on IT is a great strength of Anaplan. Obviously IT is still important in relationship with the system because they need Anaplan to work with other systems like for us, it’s Hyperion, etc.
Jeff Brobst: And everyone loves not having to go through upgrade cycles. We get notices and all of a sudden poof, we have four or five new formulas that we requested. I mean Anaplan is very responsive to us, and poof, we have two more charts we can use without us having to do anything really. So it’s constantly improving. Over our three year timeframe, it’s evolved a lot but we haven’t had to do any upgrades; we just keep moving.
Mark Sarbiewski: And so that’s one of the extra beauties of a SaaS implementation that works great for us as well. We are able to make sure you guys are always having the latest; we get real time feedback from you. If you don’t like something or something is not working, we know in seconds, whereas historically sometimes it’s been months before you really figure out if something is working for someone. I want to take a question from the audience here, and the question is how does Anaplan incorporate non-financial/strategic trends and strategic agendas in multiyear planning? So did that question make sense here, Jeff? It’s like the non financial and strategic trends and strategic agendas into multiyear planning.
Jeff Brobst: In our multiyear planning, when we are benchmarking against our competitors and our own five year plans, and so we are tracking it and because it is agile, you can change things on the fly. One of the nice things is you can make several assumption changes in your strategic planning and have it roll through a five year model like that. You couldn’t do that before. Producing five year financials actually in software was unheard of but now that we are under Intel, we have that discipline now. We didn’t know how to handle it before Anaplan really.
Mark Sarbiewski: I don’t know if you have any thoughts on that.
Patrick O’Brien: Yeah. I mean we don’t use Anaplan as a planning tool right now and that’s something we are looking to get to. But I am confident that we would be able to incorporate the strategic plans within our model building. And certainly on the retail and the sales execution there is an element of strategy that goes into that but that’s something, kind of a next step for us.
Mark Sarbiewski: So I am curious, you are bringing something different, you are bringing a different level of information and the result was awesome that you said, the accuracy of this. Has it changed how finance acts and connects to the other functions in the company in your opinion? And I know it wasn’t the greatest question in the world, but really what is it?
Jeff Brobst: There is a lot more transparency. If you go back to the sales, finance, and ops model, there is no voodoo in how we came up with the numbers. You can draw all the way down essentially and see exactly how it was built. So there is a lot more transparency. Inherently in spreadsheets, you are going to get an answer and then people are going to question the answer and then there is going to be a back and forth and explaining and proving the answer. In Anaplan, you don’t have to do that; you can draw down yourself; the rep can draw down themselves; our regional directors can draw down themselves, right. So there is a lot more transparency so it has really helped the communications between the groups.
Mark Sarbiewski: Yeah. I think that’s actually a fascinating point. I’ve heard that before. When you have the plans with all the data, everybody is looking at the same pictures of data. You can drill down and see down to the most granular level what goes into that. It drives different behavior and I am curious if you have seen some of that.
Patrick O’Brien: Yeah. I mean it’s been a big improvement so far, with the transparency and the integrity of the data. It gives us confidence and I think everyone, all these different groups using the data, have confidence that we are all on the same page. So that are going to be good results but I think we are off to a good start there.
Mark Sarbiewski: Yeah. So following on, on that point, if you have this base level, this foundation sort of transparency, this idea that we can extend and build models that really suit a particular business, where do you think you can go with this? Not just Anaplan, but where can finance go in the next few years that probably you hadn’t really thought you would get to before?
Patrick O’Brien: Well I think corporations are being asked to do more with less. Finance needs to get leaner and more efficient, and I think a robust system like Anaplan allows us to spend less time looking up data and reconciling data, and more time giving actual advice and strategy to our business partners, to the sales teams and to the retail teams. So less time on systems, and more time giving advice.
Jeff Brobst: We are evolving to essentially more analytics and reporting visuals. You have nailed planning. I am not sure how much farther planning is going to go in Anaplan; it’s already the best. For us, the next step is essentially the visualization, analytics, and reporting side that we are kind of pushing the envelope.
Mark Sarbiewski: Yeah. I think that’s one of the questions here from the audience: what is it that Anaplan needs to do more of? I think you are starting to touch on some of that. What are some of those things to bring the capability to you that you wish was there?
Jeff Brobst: Well so first it’s tough to tell people they shouldn’t use Excel and so we have tried to tell them, hey Anaplan, you can do the pivoting and the reporting in Anaplan but some people inherently want to have that kind of tight integration, and to an extent, there is some truth behind that. I mean if you are doing a board deck, you are not doing a web cast, you are probably doing PowerPoint. So how do you get that data analysis into PowerPoint? So there are cases where it’s not going to do everything. We have requested essentially tighter integration with Microsoft Office. And then you have the back-end engine to do what the other in-memory tools do, so just beefing up kind of that front end to get to the same level.
Respondent: Yeah. I think for us, and we are betting on Anaplan for the long term so that when we need updates or maybe if it doesn’t scroll right or we need a better printer interface or we need more integration with PowerPoint because we have a board deck to do, that Anaplan will be there and they will be updating their systems and they will be partners with us. That’s a big bet that we are making. And again, we are a month in and so I can’t ask that question whether we are there yet. But that’s our bet and we are confident that Anaplan is tech savvy, wants to be at leading edge, and it will be a good partnership.
Mark Sarbiewski: I think you mentioned this either when we were talking before or maybe earlier today, that you all have setup a center of excellence around this.
Jeff Brobst: We have. I mean the training is really two days.
Mark Sarbiewski: Where did that come from?
Jeff Brobst: Again, in two days you can have training and get up and going but you are not going to be expert on how to do the modeling. So what we did is we get some experts that have built some of the early models, who now have become kind of an internal consulting group. We call it the Anaplan Center of Excellence. They consult with other people, so we have created basically a community of Anaplanners within and then we also leverage the Anaplan external community as well for learning. So that’s really the best way to go because there is a little bit of a hurdle of going from a spreadsheet environment where it’s cell reference to business rules. And to optimize, it really helps having people who have done it before.
Mark Sarbiewski: Right, right. And it’s funny, you did a nice job and it’s another point that I would bring up, which is that this is the power of the community and the people who have done this. Whether it’s an HP or another of our hundred plus customers solving different problems, we come out with new functionality, and people can leverage that. That’s actually a big part of what we want to do and so it’s a beautiful model from our perspective that you are creating that center of excellence in there but also reaching out and connecting in to the largeer community.
Jeff Brobst: I think having canned applications and wizards is a good progression for you guys.
Mark Sarbiewski: That’s right, that’s right. I mean that’s an important point for the audience to understand. This is a platform, hugely powerful, easy to use. You can build essentially whatever you want, but Anaplan’s vision is to augment that with finished applications as well as having partners build on top of this platform with applications. So we have a journey ahead of us but hopefully it has a great future. Just take that time to value for you all down to two months, three months to up and running in days.
Jeff Brobst: It’s still lightening speed from what we are used to.
Mark Sarbiewski: I am curious; Jeff, you mentioned that for planning in general that maybe you would be able to move completely off Hyperion. I know that’s not where you are Patrick, right? But when did you begin to think that and how do you get to that point when you are not using it?
Jeff Brobst: So I think once we realized that inherently this data could be shared across applications, we said if I had to do things over again, we probably would have enforced the common hierarchies of member names and definitions across all the applications. We are not trying to constrict you but please use essentially the members hierarchies. Once they can do that, you can basically integrate these applications. And so once we realized that three quarters of the company was already using Anaplan, why have them re-key into Hyperion or setup the interface? So, Hyperion has essentially become again a consolidation compliance tool for us, so we don’t see the value add there.
Mark Sarbiewski: Right, right, interesting. One of the other questions here that’s come in from the audience here is really for both of you. Where would you really like to go with this modelling capability that you have laid out? I will start with Jeff and then come to Patrick here. Think three years from now. What does that look like? Who is using this modeling capability, how does it proliferate potentially and become really valuable to maybe not just McAfee but Intel?
Jeff Brobst: Our next vision is really to have it integrated into operations. So I talked a little bit about how we have a model to optimize discounting. so our next step there is to actually have it pre-approved quotes. In other words, Anaplan can look at what someone is putting in for a quote, check all the measures against the business rules, and automatically workflow and send the quote on. So you are basically cutting half the volume of the deal desk away because there is stuff that they don’t need to approve. It also makes the users lot more savvy because they are seeing what’s being rejected, knowing these are the rules. So it’s that transparency; so you have that iteration. We are trying to really embed it then more in day-to-day decision making. I think the normal kind of planning, your planning cycles, your monthly and quarterly forecast; we have that nailed. How do you make it more a real time decision making model? It’s becoming a resource to people essentially along with the what-if.
Mark Sarbiewski: I know you are in early days, but challenge you to think through ahead.
Patrick O’Brien: We would like to use it as a planning and forecasting tool as well. In fact, we are starting that discussion now in the phase 2 of this implementation. And at least locally, with the sales teams and the retail teams – I am not talking corporately, that’s a decision made by somebody else further along down the road – but for our retail managers and our trade promotion programs, to be able to have a planning and forecasting tool would be really great, and it would eliminate another system that we currently use for that. And it would also allow for more collaboration between the operational teams as well on the sale side anyway. And I think we could achieve that in the next few years.
Mark Sarbiewski: So I am curious, I want to go back to the learnings that you had as you have gone through this. I am going to start with you this time because you are relatively fresh. So you went through; it sounds like you guys made a fairly deliberate decision to get to “we are going to take some extra time to make sure it works for our users,” but if you had some thoughts for the audience, if they are looking at going down this path, what are your best pieces of advice for them?
Patrick O’Brien: Measure twice, cut once. Solicit the feedback of all the user groups that they are going to be using it, don’t force it on them because you don’t need to. It’s not a canned package that you just have to accept and deal with. You can customize it how you want. So I think it’s really important up front to invest that time to get the feedback, and it’s a little bit painful. It might slow things down a little bit, but I think you will get more buy-in and more investment, personal investment from people who have to use the program to learn about it and to pick an active role in making it their own.
Mark Sarbiewski: I mean Jeff, that’s interesting, and it’s a lesson that you might have only really realized down the road a little bit as you started to see it go to more and more places, right?
Jeff Brobst: Yeah. Again, the hierarchy is if you want these models to talk to each other, they are going to have a common platform, and they have the hierarchies and member names the same. So I mean that’s probably our biggest learning for us. We are right in setting the center of excellence because that actually has helped move things along. And the thing is, we really want it to be as an open system that any group can play in; anyone can do modeling. We don’t want to just be a finance tool; it is a sales tool; it can be a R&D tool; it can be an IT tool).
Mark Sarbiewski: Actually, it’s interesting. It ties right to a question that just came in: Can you talk a little bit about more about how executives use this, what do they see, what do you put together for them versus a sales rep or an operational sort of user?
Jeff Brobst: We have the role essentially focused on dashboards that’s giving the measurements that they want. They are not going to be the type to really drill into the detail. So we actually try to hide that part for them. We try to make it as simple as possible, and we give them filters for sure.
Mark Sarbiewski: So they can interact with the data a little bit –
Jeff Brobst: But really, they are interested in how are you doing and where, and those are the questions that they are going to ask, and that’s what they would look at on a daily basis. If it’s more than that, then they are going to end up going to their controllers eventually.
Mark Sarbiewski: Yeah. One of the great customer stories that I heard was that an executive is coming, and they are looking at reorganizing a territory which, in an old approach with spreadsheets, was a ton of work – many, many hours of work across the team to make sure it got right. But now it’s a matter of changing a hierarchy and moving that stuff around, and there it is like 10 minutes later. Drag and drop is something that people who are in this audience and who know what is involved in that would be interested to see. So you have gotten those executives to their dashboard, but how do you react to those asks that they have? Have they learned to ask you tougher questions?
Jeff Brobst: They have, actually. They understand the details better now that it’s more transparent. The thing is because it’s real time in-memory, you could literally make the change in the fly and see it happen on the other guy’s screen. So you can actually have this interactive planning with someone as a customer of yours, if you are the planner and you have a customer, “Okay, we will make that change, what do you think?” You both are looking essentially at the same form.
Mark Sarbiewski: And to me, that’s one of those keys where you want to be able to do the what-ifs, check scenarios, look forward, what’s the impact of this – that’s actually the kinds of things in my experience, the finance folks I have always worked with, that’s what they want. They want to be able to deliver that kind of value and insight and not just keep score.
Jeff Brobst: One of the great features is the logs. You don’t have that in a spreadsheet, but you could see who made that change with changing those people around. You know exactly who and when they did it. Wait a second, I thought your code was x, who changed it to y. You can go right to it and know exactly who did it. So the audibility of the modeling is fantastic.
Mark Sarbiewski: And so the question to you then, is: For different roles, what do you put in front of them? What kind of dashboards you are building out?
Patrick O’Brien: I think you have to be careful with that. I think we would want to set up dashboards similarly, and that’s something we are planning to do, so that whatever that person might be interested in, whether it’s the retail function or the sale function or the finance function or all three, that they can just access that data really quickly in a friendly way without being able to manipulate it and change it with a little bit input from us.
Mark Sarbiewski: There is a question here for Jeff. What was the IT reaction to the center of excellence? And this is a more general question, maybe it may be beyond IT, like what was the pushback you got, if you got any pushback on any of the stuff that you worked with?
Jeff Brobst: So originally, they are defensive, I mean you have this Hyperion planning implementation, very mature since 2008, what are you doing and what does that do to our jobs. It’s natural for that to happen. And then of course, throughout the security card, are you sure this is SOX complaint and passes our security test. But once they realize the benefits, they actually have embraced it because again, they know what their core competencies are. And that’s in the data architecture; they don’t have to worry about, these upgrades and server. They hate that stuff, and then they don’t want to be in the report writing business either. So at the end of the day, they are pretty happy, but there was resistance at first for sure.
Mark Sarbiewski: How about you, from IT or anyone else, was there a pushback or objections?
Patrick O’Brien: Well, I think from where we were coming from, a new system was working open arms. People had different perspectives on who we should potentially go with and using something new like Anaplan is new to the market made some folks nervous. But I think with some demos and some demonstrations about what was possible with it and why we picked it, it alleviated those fears.
Jeff Brobst: I have to say we had a strategy, and our strategy was go after the spreadsheets first because it’s hard to argue getting rid of spreadsheets. And then once you have proven the capabilities, then people are going, “Okay, I can see why.”
Mark Sarbiewski: That makes sense. There is a question that came in here about how do you control for human error and making sure that your models don’t get corrupted by human error? Maybe you answer that with the audits but I will put that out there anyways.
Jeff Brobst: Well, you don’t have the same type of errors like you are checking your spreadsheets because again, once you set out the business rules, it’s good for all. So you don’t have to test every cell. All you have to do is test your business rule and does that make sense. So yeah, you have the logs on input changes, and typically, the models are going to be combination of formulas. In our bookings, our forecast model is a very sophisticated regression analysis but also has some inputs for like known deals or known things that are happening.
So you have a combination of calculations plus freeform input. But it logs with it and everyone is able to put notes in the system, “Hey, I put this entry in because I am not expecting this deal this quarter.” So we encourage people to annotate their changes actually in the tool itself that makes it visible in the form. The form shows what notes you have. So that’s one way to get around it. But again, it’s not the same thing as spreadsheet error where you are referencing the wrong cell. I mean once you set the business rule, the business rule is a business rule.
Patrick O’Brien: Yeah, I mean it goes similar, I mean there is a nice audit trail there. I mean at the end of the day, no matter what system you have, I mean there has got to be some human level expertise that just makes sense.
Jeff Brobst: I mean you can’t stop someone from putting in a million when they meant a hundred thousand but that’s where the analytics come in play.
So on that data, we run the analytics and say, “Is anything out of balance?” and that’s something that Anaplan does very well. You can scorecard it, is anything out of balance, and highlight those things. It’s one way to catch those errors.
Mark Sarbiewski: Yeah. We just had a conference in Europe, and one of the things that we were demonstrating was what you are talking about: connecting plans across. And what we were talking about was planning by exception. So in our demo, we talked about opening a whole country. You are a company, and you are expanding and growing and what does that look like. And rather than build everything from scratch, can you pick up the last country that’s similar that you opened and use that as your baseline and then decide where you are going to change. I am curious if you do any of that or if you think that’s a place that you could imagine getting to.
Jeff Brobst: We do okay. So for example, we think a better way to forecast from the finance side like on sales is really looking at what’s in to date but what’s coming up for renewal. What kind of attach rate do you have, other competitive pressures and factor that in as your number. You are taking the emotion out of it, and you don’t have salespeople hedging one way or the other, “Oh, that’s going to affect my quota next year” that sort of thing. It’s statistical analysis that over three years has proven to be more correct than anything else we have done. It’s taking the emotion. But it is by exception because it’s literally a calc. We have these factors actually very few input into that forecast. There will be some known things that you are adjusting for anomalies, but you are only correcting for the anomalies. The rest of it is what it is.
Patrick O’Brien: Yeah, and I would say that at Activision we’re not there using as a forecasting tool, but that’s definitely something we would like to get to.
Mark Sarbiewski: I confess, the story you just talked about, using those analytics to derive the right sales plan, I was so fascinated by it, and it was so appealing. But at the same time, as a product person who has spent 20 years enabling sales force thinking that was the answer, I was a little depressed.
I just realized that that first thing you have to solve was, ‘I have got to put the right number for the right rep, right products, right mix, right territory, all that.’ And it’s interesting because I have talked to so many people that I have known through the years that are saying, “Well, yeah, we try to do that, and that’s what sales ops tries to do. It’s peanut butter and allocated, but it’s overly simplistic, and the reason is it’s a big analytic problem. You talked about regression analysis. It requires distributed expertise: ‘I have got to get the territory manager’s knowledge of what happened in the account there.’ So it’s a solvable problem. You have proven it, and then enablement comes in. So I understand where that’s coming from.
So now we have a question on how do people interact with the community. You talked about the center of excellence – you have new people coming in all the time; you are a big corporation; you get new salespeople. How do they get to where they need to be? I know it’s always training, but how does the community help?
Jeff Brobst: We have road shows to really try to evangelize the product, internally. And so that’s probably the biggest way, I mean it’s pretty simple. Again, it’s a spreadsheet interface, so it doesn’t take a lot. The questions are typically what are the business rules; that’s what people are questioning or maybe they are debating. But the use itself is pretty easy, and I am not sure if maybe the question also had to do with the external Anaplan community. So you have multiple methods, so you have chat and people have posts and there are just certain community pages that people can log certain problems essentially. It’s very similar to the way Microsoft does it.
Patrick O’Brien: And for us, I mean when we have an issue or something we like to explore, we do check out to see if the topic is in the community and has been explored before.
Jeff Brobst: There is no reason to bang your head. Somebody has come across these same issues before. So I mean as long as you understand that, then you definitely have people you can reach out to.
Mark Sarbiewski: One of the things that I think you each talked about that was a little bit unique is bringing in a bunch of data to do that analytics. Without disclosing anything proprietary and competitive secret here, what data are you bringing in to build that sales plan and how do you do that?
Jeff Brobst: Our entire transaction log is assigned to sales rep and then all the hierarchies end up aggregating, but you can drill down to I sold to this customer this SKU. The model handles it, and so you don’t have to worry about aggregating it yourself into a more compressed form like you would in a spreadsheet. You subscribe; you can drill down.
Mark Sarbiewski: Have you come across any limitations of the Anaplan system; so dimensions, model size, number of users?
Jeff Brobst: Yes, and they get addressed. I mean so from the beginning till now, we have asked for, again, the numbered list is something, many-to-many relationships. Like it’s normal to have one to many, but many to many, it’s needed to cure sparsity. So that’s something we have asked for that’s being delivered upon. There are certain Excel formulas that we say ,”Hey, I used this one Excel formula. Can we get that in?” So we put our list out of things that we are trying to do and do we want big, what I call, Big Data, Hadoop type data eventually? Absolutely, I think you are not there yet, but I think you will get there, and you know that’s the way it’s going to. So our data will get bigger.
Mark Sarbiewski: I actually have no doubt, I mean you think of the amount of data you have brought in to just do the sales planning. So then the rest of the operational planning, similar amounts of data, you get these huge models. How about you, you’re in early stages so you probably haven’t bumped up against anything?
Patrick O’Brien: No, I mean nothing significant but just simple things like get the print function operated a little bit differently, I guess small things like that. And I know that there is a refresh coming up in a couple weeks, and we are looking for that to be solved in that refresh. And I think the important point is that there is pretty good two-way conversation and dialog, and then if we have come across issues, you guys will address them.
Jeff Brobst: Works better on Chrome than IE; there are small things like that.
Mark Sarbiewski: I am going to take one more question here from the audience; I know we are coming up on the top of the hour. The question here is, once everyone was looking at that same set of data, what was the response? You have touched on the transparency. Did you see fundamentally different kind of behaviors because of that transparency?
Jeff Brobst: Yeah, across the board depending on what application you are talking about. So if you are looking at profitability, you would look at it and you didn’t realize maybe some products you are spending a lot more money than you thought you were and some really high flying products you are not spending what you thought was enough. I mean you can really drill down all the way to the P&L transaction if you want to. So people explore a lot more.
Patrick O’Brien: I think it’s just promoted better communication between headquarters and the field, etc. So I think we have seen an immediate dividend there. So the feedback across the different groups that use it now has all been really strong.
Jeff Brobst: It also highlights inconsistencies in your data because you can see it. It’s like where is the last quote for that order, I don’t see one, something happened, like you didn’t have the close loop on that particular order or that process. So it highlights those things.
Mark Sarbiewski: Yeah. I know when we run our entire company on it, and everything you say is exactly right, marketing budget is part of the overall planning, if assumptions change at any level, it reflects through. I have the power to do what I need to do, but it’s not a disconnected thing from everybody else. So it does make communication – I think you point right. If people can see it, it’s all visible, and that’s what we are going for anyway.
Respondent: It’s real time. You don’t need to have someone send you something.
Mark Sarbiewski: Well, no doubt, we could talk for a long time. I want to thank both you Jeff and Patrick for joining us today as well as the audience. I hope you got a lot out of it. I know there were a bunch of questions we didn’t get to. We will answer all of those. The recording will be available for replay, I think, within a few days here. We will make sure everybody has a link to that, and in that same notice, we will send you link to something where you can see the product and a little bit what we were talking about today. So with that, I will close today’s session, and thanks again for joining us. We hope to see you again soon, and have a great rest of your day.