Unknown Speaker 0:00:03.2: 

Welcome. This is the first, obviously, of our afternoon sessions, and Janani and I are going to talk a little bit about HubSpot and Anaplan. So we will probably take - well, we're planning to take about 30 minutes and leave room for questions, but if you have a burning question, I would say just shout out and help us along our journey here. The job here for us is to help you evaluate Anaplan and all the learnings that Janani has had, and just share. So, without further ado, Janani can you pronounce your last name for me, because I can't speak English. So how do we say your name, officially?  

 

Janani Venkatesan 0:00:41.6: 

Janani.  

 

Unknown Speaker 0:00:42.6: 

Janani? 

 

Janani Venkatesan 0:00:43.5 

Yes.  

 

Unknown Speaker 0:00:44.8: 

Was practicing that forever, and still managed to get it wrong. All right. So welcome, and thank you very much for spending time with us on stage. So as we get started, why don't you tell us just a little bit about you, and then we'll get on to a little bit about HubSpot. So where do you come from, bit of career, how the hell did you manage to be running Anaplan at HubSpot, and why were you crazy enough to join HubSpot in the first place? Something like that. 

 

Janani Venkatesan 0:01:06.8: 

Sure. First, absolutely thrilled to be here at Anaplan Connect. Thanks for having me. I am Janani. I am from HubSpot, and I lead the finance transformation team at HubSpot. I live in the Bay area in the East Bay. Prior to HubSpot - okay, I see some hands going up. Prior to HubSpot, I was at Dropbox, and I was leading the finance transformation team there for five years. Even before that, I was actually working in India in a company called Unilever, and I've been working across finance functions like FP&A, accounting, procurement and name it, and I've been there because that used to be Unilever's strategy, just letting people learn from experiences. I learnt a lot there, and that honestly helped me with finance transformation, because you really get a central lens to everything that's going on. You appreciate all functions, and you're really able to add value. So that's a little bit about me and my work experiences. You asked me why I joined HubSpot. I love HubSpot. I think HubSpot is an AI-first company, truly innovating in the space of product and AI. I love the leadership at HubSpot. For those of you who don't know, Yamini Rangan is the CEO of HubSpot, and I've worked with her before at Dropbox, and the leadership at HubSpot truly inspires me. So that was also one of the reasons why I joined the company. 

 

Unknown Speaker 0:02:37.8: 

Fantastic, and remind me, when did you join? What year? 

 

Janani Venkatesan 0:02:41.4: 

So it's been 18 months at HubSpot, yes.  

 

Unknown Speaker 0:02:43.0: 

Eighteen months, that's right. So tell us a little bit about when you walked in. You'd finished transformation at Dropbox, obviously. So now we have a new transformation journey at HubSpot. So tell us a little about if you walked in the door, all the skeletons jump out the closet, what was it like? What was the situation that you inherited? That will be useful to know. 

 

Janani Venkatesan 0:03:01.5: 

Yes. For sure. I think, for me, some of the biggest observations when I went into HubSpot, and my job was to start transforming the finance function at HubSpot. One of the early observations for me where - as I came in - so HubSpot obviously is a company focused on growth and all of that, which means every function is focused on optimizing how they work and support fast growth, which then means a lot of them picking their tools that work best for them. So there are definitely a lot of tools, fragmentation and then people optimizing their own processes, which are good to drive growth. But then in the long term, I think we also need to have that connected vision, and we need to have platforms that truly help us scale and help us be more productive and be more efficient. Also, I would say agility is so important in this age of the rapidly changing business environment, especially in the age of AI, right? You see new tools and you see news every day on changing technology and new features being released, and you need to be able to constantly adapt your processes and your systems to be able to quickly embrace all of that and really drive growth. So I would say for me, that was the biggest observation, how do you balance [over speaking]…? 

 

Unknown Speaker 0:04:26.7: 

Lots of different systems in different places, doing really good jobs, but not coming together at a consolidated level. So what's transformation then in the HubSpot sense? What was the catalyst to do it, and then how do you measure success? What were the initial success criteria? 

 

Janani Venkatesan 0:04:44.0: 

Sure. When I think about transformation for finance, I look at it from four key areas. It's about people. It's about process. It's about systems. It's about data. So first of all, I would say reimagining our processes to be really agile and like I said, support the ever-changing business needs, and also make sure that our processes boost productivity, boost efficiency, and make sure they are compliant with all of our regulatory requirements. From a systems perspective, I would say it's how do we make sure that our processes are integrated and how do we implement the right tools and technology that help us scale effectively and automate our processes? Then from a people point of view, how do we truly keep upskilling our people to make sure they are aware of the latest technology and they are aware of how to use the AI tools available to them? So we definitely pay a lot of attention to and spend a lot of time upskilling our workforce. Then data, I feel is becoming all the more important these days, especially that's the foundation for AI to work.  

 

Janani Venkatesan 0:05:56.7: 

We need clean, structured, consistent data, and we also need unstructured data, and you can truly unlock the power of AI by bringing together all your structured and unstructured data. So for us, there is definitely a lot of focus on data and insights, and that's where we are also looking to leverage tools like Anaplan to help us really achieve our vision of connected data and collaborative planning and execution. We feel that that's really the foundation for future intelligent AI-powered growth. 

 

Unknown Speaker 0:06:27.2: 

All right. So we have a look at the landscape, the four criteria you talked about. So some early learnings about how to set this up. So we heard [?Zaf] earlier eat the elephant. He took a great big mouthful, or the team, and oh, that's a bit too much, and we need to break that down a little bit. So what was your approach? Who was helping you from the executive sponsorship? What was the first thing that you set your eyes on achieving? 

 

Janani Venkatesan 0:06:51.0: 

Yes. I think one thing I would say, before I get into like a which functions did we prioritize, or what is the order of execution of transformation. One thing I would emphasize is while execution can happen in phases, I do feel like we need to have a central lens to planning. We need to really understand what is the why of transformation. What are our overall objectives and goals? We need to make sure that there is a central orchestration of the design. The initial design phases are so critical, and we need to make sure we bring in all the right stakeholders during that phase, because every process, if you think of it, has upstream and downstream dependencies, and in a way, they are all interdependent, interconnected. So while we could take, let's say, SP&A as a priority for Anaplan, as we are thinking through and designing that system, we've always adopted a more central approach where we also look at all the other connected areas, like for example, procurement looks at budgets before approving new purchases, right?  

 

Janani Venkatesan 0:07:59.3: 

So in a way, procurement workflows are dependent on the information from SP&A, and there are so many such examples, like tax maybe is a more downstream process, but again, it needs data, it needs all of this data for their tax reporting. So there are always dependencies. So I think that we definitely need a more centrally orchestrated designing of solutions, and then we can approach implementation in phases, and as you approach implementation, pick the ones, again, that tie back to your overall transformation goals. Which could be different for every organization. Maybe for one organization, it's all about cost efficiencies. For someone else, it's about speed to data insights. For someone else, it could be driving revenue growth and supporting that. Depending on what your goals are, then you could pick okay, which areas or functions are you prioritizing? So at HubSpot, for example, we are working on a few things right now. We recently completed our implementation of Anaplan for SP&A, and there were other functions already using Anaplan, but now we are really looking to unlock value from connected planning. 

 

Janani Venkatesan 0:09:03.8:  

How do we bring these teams together and truly unlock the power of data and insights using Anaplan as one of our tools? We are also looking at other areas like reimagining our entire structure to pay platform using AI-powered tools, and how do we connect those back to systems like Anaplan for all the process dependencies. So it's about also integrating our solutions. Yes, those are our priorities at the moment. Like I said before, we always have a central lens [?engine]. 

 

Unknown Speaker 0:09:38.2: 

All right, so instead of eating the elephant one bite at a time, it's painting the picture one color at a time, [unclear words] the corner. Okay. Just give us the flavor of the team that you've got to do all this hard work with. What's your size? Do you use any outside services? What's the structure of the team? 

 

Janani Venkatesan 0:09:54.4: 

Yes, so I have a team that is focused on process, where they partner with our business stakeholders, which are mostly all the finance functions like tax and treasury and SP&A and procurement, and all of these teams. They are trying to understand each of the teams' processes, the gaps, the pain points, the opportunities. Then, like I said, when we get to designing solutions, we have all that process knowledge and we're able to design the right solutions for the business. This team also partners very closely with our tech org, when it comes to picking the right tools and designing the right solutions. I also have a data team that is really focused on building models, reporting and facilitating self-serve reporting for stakeholders and data insights. Also, more recently, we are developing a [?fin] board, what we call a fin board, which is really a finance knowledge board that connects all of the structured and unstructured data in finance, like we connect to Snowflake, we also connect to all of the unstructured data like QBR DAX or Earnings DAX, and things like that, which facilitates easy Q&A for the finance team.  

 

Janani Venkatesan 0:11:06.9: 

So that's something that we're working on. I also have a systems team, which is primarily the Anaplan team. So we have a couple of model builders who support us with the FP&A planning, implementation and model enhancements. So that's my team, and we are about 20 in my team, working on all of the transformation efforts. 

 

Unknown Speaker 0:11:29.9: 

Gotcha. You report to finance or IT? 

 

Janani Venkatesan 0:11:32.9: 

Finance.  

 

Unknown Speaker 0:11:33.5: 

You report to finance. Okay. Just finished FP&A. Just give us a flavor of the FP&A application. What depth do you go to? Dimensions. How do you plan FP&A at HubSpot? 

 

Janani Venkatesan 0:11:47.8: 

Yes. So our FP&A team is the central team, obviously, planning, leading the annual planning and all of the monthly forecasting at HubSpot. So we also have Anaplan being utilized by our operational teams. Like for example, the sales and the sales teams and also the workforce planning teams. So each of these teams obviously have their own operational goals. For example, the sales teams get into detailed quota planning, and they go into territory and geography and segment level planning and all of that. For FP&A, it's more about consolidating at the company level and translating that to financial planning and results. So today, in the way we are structured is each team has their own Anaplan models, and then we have the SP&A model bringing all the pieces together, and that's where we are striving more for connected planning by unifying our data hub. I think today, we all saw a little bit about the data hub. We saw the areas. So two key things that we are trying to immediately work on is one, we want our data hub to see if we can really centralize, unify our data instead of bringing data into siloed models, can we just have one unified data hub, and that data hub then talks to all of the other Anaplan models?  

 

Janani Venkatesan 0:13:10.2: 

The other thing is, how do we utilize ADO, and how do we actually seamlessly integrate with our other applications, like for example, Workday? It is so important for us, I think, at this stage to get to position. I think during the Workforce demo, we also saw the position ID level planning and insights and all of that. Now, I think with the technology and with AI and all that, you truly can get to that level and unlock deeper insights, once you're able to integrate with tools like Workday. So we want to make sure we are working through the entire position ID lifecycle and automating that whole process right from creation, and then following through all the regional transfers and [?attritions 0:13:50.4] and everything, promotions or whatever happens with the position ID, you're truly able to follow the entire life cycle of a position ID and you have deeper insights. With headcount being the bigger spend for us, it's so important and critical to actually unlock value by integrating with Workday. 

 

Unknown Speaker 0:14:08.4: 

Okay, cool. Then what's next? So you talked about unified data, still doing infrastructure transformation, but from an Anaplan perspective, the next one or two applications, what's the most important areas you're going to expand on? 

 

Janani Venkatesan 0:14:21.5: 

Yes, first I think with Anaplan, the number one goal is enabling connected planning and getting all these teams together, unifying our data sets, revamping our data hub. I think once we have done this, obviously, the next thing we are going to get to is deeper insights, and how can we get more value, quicker data, faster insights and all of that, and that's where we're really excited to start looking at all of the wonderful agents and also, the things like the co-model or co-planner that we saw today. I think it was all very exciting, and we are definitely looking to at least start experimenting with some of those products and see what adds value to us. So we are definitely going to experiment a lot with AI. I think with AI, all of us are learning, and it's important to actually start experimenting and really start [?iterating 0:15:14.0]. We have to start small, but then also learn quickly, iterate, and then be able to scale very soon.  

 

Janani Venkatesan 0:15:22.5: 

Then I would say, finally, I think we are also focused on governance and data quality, and all of that as well. So how do we implement the right access controls, which is so important, which I think most of us tend to overlook during the implementation phases, but it's so important. Like building in the right data quality measures, governance standards, and all that, are so critical, and that's something we constantly keep looking at. Also partner with our internal audit teams, bring them early on and make sure we are following all of those policies and making sure that we have really good compliance, high-quality, accurate data, and especially because these are all tied to financial results. 

 

Unknown Speaker 0:16:06.6: 

Gotcha. Yes, governance is something we've heard a lot more recently as a built-in requirement, and particularly with AI. So how do you know AI is doing the right thing, giving me the right answer with the right confidence, and a human being looked at it and helped the AI train and said, yes, this is the right answer. Those are the things we're looking at from a governance framework perspective. So not just users, but AI as users. Certainly, over the last two or three months, we've heard a lot more about that. I was going to ask you about AI. Has HubSpot picked - you don't have to tell us which one - but have you picked a corporate standard yet, or are you still early in the experimentation? I know within HubSpot itself, the application are doing a lot, but for customers. Internally, have you picked a standard yet or still learning? 

 

Janani Venkatesan 0:16:50.4: 

Yes, I can definitely talk a little bit about HubSpot AI and what we're doing internally for HubSpotters. First of all, I think externally, probably all of you are aware, HubSpot is definitely AI-first, and recently, if you all have followed inbound, which happened in September, there were a lot of wonderful AI product launches. We introduced our Loop, which is a new marketing framework, and it talks through how the entire marketing lifecycle is now powered with AI tools within HubSpot. We have all our AI-powered hubs, and we also launched AI assistance, companion, which is called Breeze. We have a lot of AI agents as well. Internally, within HubSpot, I think, again, the focus is on how do we get every HubSpotter to adopt AI tools, and finance is no exception. We want 100 per cent of our finance team members to start using AI tools, at least on a weekly basis. So we obviously have a few enterprise tools. Right now, the usage is spread more widely, I would say. Some of us use ChatGPT Enterprise. Some of us use Claude.  

 

Janani Venkatesan 0:18:01.2: 

Each of them, I think, have their own strengths, pros and cons. We also have several other tools, like Gemini. We have our own internal versions of these that work behind the scenes with all the APIs from these wonderful tools. So we have our own chatbots and everything as well. I think right now, while it is all really fragmented, I'm sure in a few years from now, people will start consolidating to see value like we are all speaking about, how do we connect all our tools today, we'll be talking about, okay, how do we get to a consolidated AI platform? So I would say we are in the experimentation phases, where we are learning a lot by utilizing all of these tools. It really depends on how quickly we learn, we iterate and how quickly we end up making those decisions on what will help us scale effectively. 

 

Unknown Speaker 0:18:55.0: 

So I'm very curious about your [?sinbot], which I think is hilarious. I've decided I'm going to use that at some point. I like that name. So tell us just a little bit about that, because I'm sure everybody is interested. Obvious idea, but how did you execute against that? What skills do you need? How could a member of the audience maybe replicate that, or to experiment in their own organization? 

 

Janani Venkatesan 0:19:17.4: 

Sure, yes. Like I said, I have a data team that is working on this sinbot, and in our data team, we have data scientists and machine learning engineers. We also have analytics engineers. All of these resources, they all come together and they have built this [?finbot]. Just to give you some examples of what it does behind the scenes. So this team works on ML models, for example, if you pick, let's say, an ML model that will help you spot anomalies in trends. Or it could be an ML model that helps you predict risk scores like customer collections. Or it could be a model that can help you predict what is going to be your future focused on a certain type of spend, whether it's opex or it could be a capital software expense. So these are ML models which are helping you with predictive analytics or spotting anomalies and things like that. What it also does is, with thanks to all the gen AI tools we have, it also enables conversational analytics on top of that. 

  

Janani Venkatesan 0:20:25.8: 

So you could go just ask a question, instead of you having to open a dashboard or open some report, you could just do a Q&A with that bot and you could ask, okay, can you explain the trends for a certain type of expense in the past one year? Can you explain the drivers behind that variance? So it can help with, for example, BVA commentaries and things like that. So that's what we're trying to get to. Like I said, we want to use structured as well as unstructured data because we feel there's a lot of content or information in the unstructured data. You have it in all those slides or your nodes, and you have to find a way to tap into those resources to truly leverage the power of AI. 

 

Unknown Speaker 0:21:11.4: 

Gotcha. Yes, we're doing a simple similar thing internally where it's going against Anaplan, going against email, going against PowerPoints, everything. It is mildly disturbing what it can get to and how quickly. But that's okay, it's all secure. So as you have been here for a little while, a couple of years with HubSpot transformation at [unclear words 0:21:29.4], what are your words of wisdom for somebody setting out on the journey, and then maybe halfway through the journey, how do they know they're on a good path or not? 

 

Janani Venkatesan 0:21:37.6: 

Yes, maybe I'll talk a little bit about overall transformation and then specifically about Anaplan, because we have also recently gone through this implementation journey with Anaplan. I think, first of all, the broader finance transformation, if you're really looking to transform a specific area, I would say always start with the why? I think we first need to be convinced that there has to be a change. We should not just go ahead and make a change just for the sake of change. So first, start with the why and be convinced on why we need to change or transform that process. Then executive sponsorship is so important. I think this will really help drive adoption and change management later on. Change management is 80 per cent of the work, right? So we can implement all the truly wonderful tools and all that, but without adoption, it's not going to get us anywhere. I think it requires a change in the mindset behavior of our users, and that's really most of the work, and we should not underestimate that work. So plan for change management much ahead in your transformation efforts.  

 

Janani Venkatesan 0:22:48.0: 

Then I think I already mentioned this before, designing is so critical. So in all of my transformation projects, I think projects that have been successful are projects where we have spent considerable amount of time in the planning and design phases, bringing together all our stakeholders from all our upstream, downstream processes, our tech teams and everyone, and making sure we are just not looking at the problem that we are solving, but we are looking with a more broader lens, and we are getting in all of those unintended consequences or risks or everything, much ahead in the planning phases and being prepared for it. Then finally, I would say also do it in phases. Just don't go big bang and try to deliver everything in one go. Start small. Iterate in phases. Celebrate all the quick wins along the way. So overall, I think I would say for transformation, this has worked for us in the past, and that will continue to be our approach.  

 

Janani Venkatesan 0:23:47.8: 

Specifically for Anaplan, I just wanted to call out a few learnings that we have had along the journey. I think one of the biggest things I would call out is start thinking about data reconciliation and validation much earlier in the process, and it cannot be an afterthought, especially if we are switching from a legacy system to an Anaplan system. I think most of the work, the implementation is all done within, I don't know, three months or you said, the sixteen weeks or so for larger, complex companies. So the implementation gets done by then, but then what you realize is for switching over from legacy systems to your new system, you need to make sure your data is accurate, and that is the hardest part, right? You spend hours and hours reconciling data, and that cannot be an afterthought. So this was one of our learnings. So we would have loved it if we had thought about it a little earlier in the process, rather than after implementation. We have started doing this already for the new modules, the workforce planning connected modules that we're trying to build.  

 

Janani Venkatesan 0:24:54.7: 

Right now, it's part of the development where at every phase we are making sure we reconcile data, make sure data is of the right quality. So that's one biggest learning that I wanted to share. Then I would say it's just that we need to make sure we bring in all the right teams early in the process, to truly get the right, perfect design for your art and make sure that it doesn't result in siloed solutions for each team. 

 

Unknown Speaker 0:25:24.0: 

Okay, cool. You talked about going on a transformation journey. So at the end of HubSpot or even today, how are you explaining the value that's been achieved or is going to be achieved? So planning is hard, right? Planning is how do you know you made a better decision than you did before? Did you save any - how are you guys doing it? How are you thinking about it? 

 

Janani Venkatesan 0:25:44.9:  

For sure. I think we definitely are trying to pick a few KPIs that can help us track efficiencies over a period of time. For forecasting accuracy and forecast, improving the forecasting efficiencies, we do have a few metrics like the forecasting accuracy percentage. We have a baseline, and before, prior to implementing Anaplan, what was the baseline? After implementing Anaplan, are we starting to see improvements in the forecasting accuracy? That could be one metric to track. We also try to track time savings. How much time did it take for someone to manually do it? They were doing it in a G Sheet prior, or maybe in another legacy system, but now with moving to an Anaplan - or maybe this is true for any other tool, not just Anaplan, any transformation effort, what was it before and what is it after?  

 

Janani Venkatesan 0:26:39.1: 

Overall, I think with AI, definitely one thing that's part of our vision is are we able to shift considerable percentage of work from all the manual, transactional, repetitive tasks to more strategic work that supports decision making? So I think that's going to be our overall goal, tracking the shift in the percentages over time. We definitely want to leverage AI to move from more the transactional work to more strategic work. 

 

Unknown Speaker 0:27:10.5: 

Okay. Good. All right. So any anything you'd like to add, before maybe we open up for Q&A. Anything we missed? 

 

Janani Venkatesan 0:27:18.4: 

No. I think I would just conclude by saying that for us, if you think about finance transformation itself, I feel like the last two years I have done finance transformation for, what, seven to eight years now, and in the last two years, the amount of true transformational shifts that we've experienced, I think is so drastic than what we have seen in any of the prior years. I feel like it definitely helps us improve the speed and also the impact of our results with finance transformation. Maybe it could have taken us several months to get to a certain KPI. Now maybe it's down to days to get to that KPI. If you're looking at a tool to unlock some time savings with a certain process, maybe two years earlier, we would have said, okay, it will take us a year to get to that goal. But now with all the AI and tools that we have, maybe it's a question of a few days or months to get to that goal.  

 

Janani Venkatesan 0:28:23.0: 

So definitely things have changed a lot, and this is such an exciting time to be in finance transformation. So I just wanted to conclude by saying that we have to all start experimenting with new tools, just explore and learn. That is what is the most important. Nobody has perfect answers yet, in this age of AI, but I think who will win is whoever is ready to learn and experiment, be willing to fail, but be willing to learn and start moving faster. 

 

Unknown Speaker 0:28:54.0: 

Fantastic. All right. So thank you everybody for your time and attention.