Diving deep into B2B SaaS metrics with Rose Punkunus, CEO at Sudozi

Episode Description

In this episode of Diving Deep, Subscript's CEO, Sidharth, has an engaging conversation with Rose Punkunus, CEO at Sudozi.

Sidharth and Rose go deep into B2B SaaS metrics as they discuss:

In a challenging economy, what is the activity with the highest ROI for the finance team?
Practical advice for turning large amounts of data into insights that can guide a business.
How running Finance at a B2C company differs from a B2B SaaS company.
Wait...is Finance mainly a data job now?
And more!

Show Notes

Follow Sidharth: https://www.linkedin.com/in/sidharthkakkar/

Follow Rose: https://www.linkedin.com/in/rosezhong/

Follow Subscript: https://www.linkedin.com/company/subscript/

About Diving Deep with Subscript

Diving Deep with Subscript is a video series where we dive deep and explore SaaS metrics with leading investors, CEOs, and finance leaders.

Watch the entire series of Diving Deep with Subscript

Get caught up on the entire series right here: https://www.subscript.com/diving-deep

Episode Transcript

Sidharth Kakkar
One of the reasons I'm so, so excited, or I am so excited to chat with you today, is that you have this amazing background that's a combination of two things that make total sense together but I've not met that many people who have this combination of experiences both around data science and data analytics, and then as a CFO, and you've had both those sort of titles. And so I have to ask, how do you think about the combination of those two? Like the intersection of where data and finance roles meet?

Rose Punkunus
Yeah, thanks. Great question. For me, I've never really known it any other way. I know so many people can approach the CFO role and finance in general from different perspectives, right? Like accounting analytics. For me, it was very much data science and analytics heavy, really working from data tables and data upwards into for financial statements, accruals, GAAP metrics, things like that. And so I've never really known it the other way. And I think that fortunately for me and just the timing of the industry and the macro trends, you do see more data heavy CFOs, strategic CFO's being put in place, one that may or may not have had an accounting background. So I think that's maybe just luck. But then also it's partly because there is more complexity and there are more data components of a company and really trying to, one, understand how to even get the data in one place and just get your data right. Then the second part is where the models and the assumptions, correlation, causation, all of that sort of comes into play.

Sidharth Kakkar
Yeah, it's so interesting because you mentioned accounting and I often think of like accounting has always been kind of a data job, but a lot of the job has been getting the data perfect and now it feels like the CFO role is beyond just getting the data perfect. That's sort of like table stakes and then there's so much more you have to do here's. How you think of that?

Rose Punkunus
Yeah, I actually think accounting can be a superpower and a secret weapon that isn't always valued as much. Maybe it's just a simplistic way of thinking through accounting, but I actually think the building blocks of how your financials are put together are very important. Like how you structure your chart of accounts, what thresholds you decide to make allocations accruals, what policies you put in place to enforce that data accuracy super important and you can use those skills in good ways and bad, just depending on the ethics of the company and what you're trying to say. So accounting is super important. I think the difference that accounting historically hasn't touched is then what do you do with that data? Right? So yes, you spent this much on X over this period of time. Does that mean you're going to spend it again? Does that correlate to the revenue? Or maybe your revenue is just completely independent and randomly people came in and bought your product? So that's the place where I think is more data science, finance, if you will. But certainly the way that you structure your financials and the more accounting processes are really critical to that foundation.

Sidharth Kakkar
Oh, I love that. My background is an engineer and there's like a clear sort of engineering parallel I feel like there, which is around how do you set up your infrastructure, how do you pick your technologies, how do you create the right database scheme for the job that you're going to have to do and that type of stuff. And that's sort of similar. How do you lay the groundwork so you can do all the things that come after it?

Rose Punkunus
Yeah, like an infrastructure engineer or data engineer. Very different from an ai engineer or whatever. Right. Maybe they could have done both. But this is a job that's way different and each job is important in its unique way.

Sidharth Kakkar
That's cool. As far as the sort of competencies that finance leaders need to have as a result of this more data oriented world that they now live in, what do you think has become more important? Or what should like, finance leaders really start thinking about that maybe would not have been intuitive in the past?

Rose Punkunus
Yeah, certainly. And what comes to mind is actually, I don't know if you've played like a World Cup is going on now, we are recording in November world cup is going on now. So if you ever played, like, a video game, you have this Pentagon thing, like, are they really fast or are they strength? What are you really strong at? What is each player really good at? And you have different players in different parts. And I think for the CFO or for people who are on the CFO path, being analytical and being good at Excel, typically you're already maxed out there. Right. And so talk about getting you to that next level in that official CFO role. Really think it's improving those other dots where you may not have as much strength or practice yet. And it's communication, storytelling, relationship building internally in the company and with all the external stakeholders, investors, customers, potentially competitors. So while there is more and more data, ironically, I actually think the emphasis is on how to communicate with maybe only two or three data points. Right. And which two or three data points do you use? How do you summarize all the other data into those two or three data points and tell the story in a way that makes sense and can help you build relationships with whichever parties you're trying to build a relationship with?

Sidharth Kakkar
Yeah, that makes a ton of sense to me in that storytelling is a critical part of making great use of data. And so it makes total sense that that would be the place where if you're amazing, this is maybe the next step to unlocking new value.

Rose Punkunus
Yeah. I'm sure you have had those experiences in engineering too, right? Like, you have logs and logs and logs of data, but what does your director actually want to see or know about what's happening in the systems? So I think similar situation in finance as well.

Sidharth Kakkar
Super cool. I'd love to learn about the finance data science function you started at Uber. Can you tell us more about it?

Rose Punkunus
Yeah. So when I got to Uber 2013, Uber was going from a black car model, which that by itself was a several billion dollar business, to UberX. Uber eats Uber pool, so they're ignoring the complexity of different cities and they're just like different products that the company was launching. In addition to that, you also had every trip is unique. The location, the distance, the time. I know. SaaS today we talk a lot about usage based pricing. Uber and the consumer model is a usage based pricing, and you don't know what the usage is going to be until someone requests the trip and actually goes on the trip or requests the meal. So you have these $10, 20 usage rates, pricing driven revenue streams that were adding up to literally like hundreds of billions of dollars. So this is not an Excel exercise. Right. At least on the revenue side. On the expense side, you can still model certain things in Excel, but it was breaking in a way. We were like, okay, well, how do we change this thing over here and actually have an impact to know? Not have to wait five days for our total revenue forecast update.

Rose Punkunus
So how do we know what's going on? So really starting to use data science and get more accurate predictions of revenue forecasts by product, by market, looking at historical trends, looping through, if you will, and trying to come up with some correlations, some general guesses on what investments would drive what type of revenue, and starting to do that at scale across hundreds of cities, across several different products in the city. So I had the opportunity and the budget, fortunately, to basically hire some data scientists that were solely focused on pricing and revenue forecasting. And then that expanded to the full P&L, right? Because it's like, oh, if you can forecast this, why not forecast data center costs? Why not forecast this other thing? And it just keeps going down and down the P&L. Of course, you still had Fp&a, sort of like providing the framework and the guardrails and even an alternative model, right? Just the gut check, different metrics that came out. But really the more dynamic forecasting was done by the data science team and in collaboration with a bunch of people. And then we also actually use TM One, and I believe the company still uses TM One today, but more for a budget repository rather than coming up with different scenarios and really being dynamic on that.

Sidharth Kakkar
That is super cool. What are the types of decisions that ended up being driven based on the learnings from building this?

Rose Punkunus
Yeah, absolutely. So there are a couple of decisions on some real key decisions on resource allocation. You have to rewind ourselves back to 2014. Through 2016, Uber had raised literally billions of dollars. Where do we deploy this to make the most gain for the company? On market share, on revenue and that brand? So understanding in which markets, if we deployed X amount of dollars, we would actually get more riders or get more drivers. And in some markets, the incentives didn't do anything right. The trip stayed the same volume. So why would you be pouring incentives and promotions into those dollars if it didn't change behavior? So there's a lot of sort of like machine learning data science assumptions about how behaviors can change based on the investments in the company. So I would say that was like an initial driver for having data science and having more dynamic forecasting in the business model. Then if you work more on the corporate side and more traditional Fp&a, it's important to understand sort of the full stack of your expenses, like fully burdened, whether it's CAC or other sales. We didn't really have a sales arm, but definitely marketing or even engineering teams building growth products.

Rose Punkunus
It's not free to create that button in your iOS app, right? You have to dedicate engineers to doing it. So thinking through, well, if you're spending X million dollars on this. Is it going to drive the same outcome or should we reallocate money to something else? And I think those models become really complex if you are just trying to manually do them in Excel. I think the organization was more thoughtful about where to allocate resources to really impact growth. Having much more easily built scenario plans with more technology on that.

Sidharth Kakkar
It's amazing. I mean, I often think of the job of the CEO and the CFO is to decide on resource allocation. That's kind of the core job they have. And you basically built the ultimate tool to help make the resource allocation decisions.

Rose Punkunus
Yes, exactly. I mean, it was not perfect and still is not perfect. I think if any companies like we have a perfect forecasting model, we can maybe call BS, but maybe we can make the levers a little bit easier to pull or poke and have more visibility to some range of scenarios that would happen in each case.

Sidharth Kakkar
That's really cool. You recently shared a really compelling post on LinkedIn where you said something like finance plays a critical role in getting the data right and aggregating the right dimension, aggregating in the right dimensions and surfacing for the right audience. We love that because at Subscript that's what we do as well. And so I wanted to unpack that a bit. What do you think are some of the challenges that finance leaders face in doing that?

Rose Punkunus
Yeah, certainly. Let me share an anecdote that I had at itunes actually. And I think there's a generalizable way to see how this impacts all the finance at itunes. I joined itunes, to lead analytics and data science for TV, movies and podcasts because they didn't staff podcast since it wasn't revenue generating. So I picked it up as a side hobby within the team. But my main role there was to produce analytics for the TV and movie business partners, which are studios, to help them put more content digitally. Because at that time, 2011, 2012, everything was really still physical. There was some netflix was obviously growing, but in terms of like real box office, real live, near time digital reduces, that wasn't really a thing yet. So one of the interesting challenges we had was to report back to a studio how many copies of a movie was sold and why it was interesting because if you think about the 30 or 40 different foreign translations of a movie, the title is different, the actors are actually different. And you really have to have a human go through and look at the movie and the data to actually know which movies are actually the same movie in different countries.

Rose Punkunus
Right. So I would parallel that with the accounting GL structure where you are logging everything. Everything is logged, all the data is there. But really to drive an insight for us at the movie level, you actually had to tie together different pieces of data that didn't have a common ID or linkage. And I think that's basically what finance people are doing, which is accounting has logged everything. Everything is stored, everything is kept properly in the accounting format. But to make business decisions that are relevant for your organization, you might need to look at two or three different pieces of data together in the right timeframe to actually draw a conclusion. If we go back to the marketing example, at Uber, we combined rider promotions, digital ads and money we paid for the Growth team on Rider app together as Rider Growth money. Right. And those are very distinct places in the P and L from an accounting perspective that you wouldn't naturally look at together from a traditional financial. So yeah, I think, you know, the aggregation, I don't know, the rolling up is another way of saying that the rolling up and the different time slicing and dicing is like super important to make decisions.

Sidharth Kakkar
That's such a good point. The time things makes so much sense in that you have to look at data points compared at different points in time to be able to get some sort of insight out of it and then those need to be aggregated in an intelligible way for that insight to be meaningful rather than just two roll together. So those two things makes more sense aggregating in their dimensions and then like getting the time dimension right. What are things that you think that finance teams could be doing better, to be able to manage this easier or how can finance team set themselves up for success in doing this?

Rose Punkunus
Yeah, I think there's more automation that could be leveraged and I kind of say that lightly because I know there are a lot of tools out there to do a lot of things. I'll focus just on the data side, both on revenue and on expenses and I think this is kind of where both of us are actually trying to make an impact in the market where again, you have all the data with low tech, right? You could throw all the data into Excel or Google Sheets and just have multiple tabs that automatically spit out with different macros, different views and you basically like human goes and looks at this tab and that's happened, and that's happened and to see where there are insights or things that are different than what you expected. So there's a way to automate with the existing tool set, which a lot of times, as you know, Google Sheets and Excel. I think the next level, there is some software to actually do that stack for you. So that people on the finance team can focus more on those relationships and have the machine spit out the insights and the metrics, and then take those metrics and talk to the marketing lead or talk to the engineering lead.

Rose Punkunus
Talk to whoever else in the team to make decisions about what to do next. So that's where I think the finance teams can focus more of their time, assuming that they have some resources and some technical skills to actually automate some of the data work. That is a starting point.

Sidharth Kakkar
Yeah. And this goes back to your earlier point about the communication and the storytelling. Right?

Rose Punkunus
Right. Because if you are afraid or if you're not good at communicating or approaching your investor or approaching some other person or company, you might just fill your day up with busy work and Excel, and that might be okay. Right. But really where the highest ROI is communicating the insights across the company to actually drive better business decisions.

Sidharth Kakkar
In your experience, what was the right sort of approach and forum there? Did you receive pushback as a CFO? Like, what are you talking about? Or you don't know my world. I don't know. I imagine there's art to this.

Rose Punkunus
Yeah, certainly. So it's funny, I believe that if you're not getting enough pushback, you don't know where the line is. You're not actually crossing the line if your marketing lead isn't pushing back to you. I won't speak for everyone at Uber, but there was certainly a cast of characters, and I was not the one in finance getting the most pushback, let me just say that much. But certainly you have potentially product teams or marketing teams like, hey, finance doesn't know what they're talking about. Why are they recommending this or that? And I think that's where there is some EQ and building up to those recommendations. Maybe the first meeting you have, your CMO isn't like, hey, this campaign isn't working. Go do this podcast instead. It's more like, hey, here's the data. Here is some information for you. I'd love to hear what you think about this data. How would you use it? It is helpful should I be providing you more of this information. So just like any other introduction and meeting, really, like trying to be helpful for what they're already doing and letting them make a decision that hopefully you're aligned with as well.

Rose Punkunus
So to minimize push back. But I'm not sure minimizing pushback needs to be the goal. This is, of course, different culturally in different companies.

Sidharth Kakkar
Yeah. Being aware of that context in the culture is probably really important as well. Zooming out a bit. Why do you think it's important for leadership teams and leaders, like marketing leaders and product leaders and so on to understand their business and subscription metrics?

Rose Punkunus
Yeah, this is obvious, but I can understand why it's not right. It goes back to our earlier conversation. Like, accounting data is great. You absolutely need accounting data, but it is maybe one of the most lagging indicators of the business. By the time the revenue is booked and the revenue is collected, it's not a time that you can influence the customer in any way or influence the pipeline in any way. Understanding leading metrics, particularly of revenue super critical and revenue proxy to that is demand. Like understanding where your demand is coming from, how it's evolving from the full customer lifecycle, from like maybe even some place you don't even know you're interacting with a potential customer or someone who would be in your community or in your ecosystem. So you start somewhere there. It doesn't even get to your full subscription metric, but it starts there and then at some point they become a customer and certainly you need to understand what they bought, what's recurring, what's not recurring. This has an implication on how you provide the service.

Sidharth Kakkar
Right?

Rose Punkunus
Is it something that's just technology? Do you need to have some people staffed towards this revenue that you just sold? So understanding what you sold and also then understanding, well, what are the implications moving forward? Right? Is this going to be continuously ongoing? Is this going to stop in six months? So having a real understanding on those revenue subscription metrics, how it leads to cash and where the focus can be on that customer engagement, customer relationship moving forward, that's kind of the high level. I mean I can go down and talk about the importance of that for a while.

Sidharth Kakkar
Yeah, it sounds like ultimately it's sort of like anything that you're using or any time you're thinking about your company's strategy and the implications of that strategy, you kind of need to understand the metrics behind the business to be able to do that effectively. Right? Is that a fair summary?

Rose Punkunus
Yeah, absolutely. And I think depending on what stage you're at too. I know the markets are a bit crazy right now. So really if you need to be fundraising in the next twelve to 18 months, really having a great handle on what your revenue ideally recurring revenue is what the likelihood of churn and just getting ahead of any investor questions that might come super valuable. And that's obviously also serving a different objective than just building the company and knowing your customer. But there are multiple objectives. That's why we can keep on going all day about these metrics but that's certainly a very important object, especially if you're venture funded or you're not profitable yet as well.

Sidharth Kakkar
Yeah. From sort of a data and metrics perspective, what are the differences in running finance at a consumer company versus a subscription or a SaaS company?

Rose Punkunus
Yes, super interesting. I've been thinking about this and one of the key differences even though in a subscription company you do have this contract that says like twelve months, et cetera, et cetera, I actually think the difference is not as like it's more continuous, it's not as like a strict one. Right. Like nothing and very few things in life are purely dictated by a legal contract. If a customer really doesn't like using your product, they're going to stop using your products even if they have six months left and then hopefully you can identify that before the six months is up, but it's something that the contract is not going to keep the customer there forever. So I think starting on what's similar, I actually think just fundamentals of getting to know your customer, making sure that you're solving the pain for your customer. Those are the same from a product perspective. What's different then? Very much so is the go to market motion. Do you have, how do you talk to your customers, how do you even reach your customers? So that whole workflow, very different customer size. How many stakeholders obviously Uber mainly need to sell the one person who's taking the ride.

Rose Punkunus
So very different sort of selling. And you're actually like Uber should not be talking to anyone, right? The CAC does not justify us talking to anyone. But if you're selling 30, 40, $50,000 worth of software, absolutely you should be getting on the phone with your buyer, with other stakeholders involved in decisionmaking process. So I think that then helps you inform how you should set up your company internally, where you should be investing in what type of talent you should be investing in, how you should be recruiting for that talent. So it really affects how you run the company, but ultimately you still want to be caring for customers, building things that they use. At the end of the day, how.

Sidharth Kakkar
How is the finance team's job impacted in the two?

Rose Punkunus
Yeah, I actually think first of all, things are evolving every day. So these are just my experiences. Also, I had these very heavy consumer finance roles like five to ten years ago. And the last few years I've been much more on the B to B to true enterprise side. I think running finance at a consumer company, broadly speaking, is more like running a science experiment where you can push buttons, you can deploy capital in marketing or other things. And because the volume of numbers is so large, you will get some signal back. I think running finance, again, this is like the top of funnel revenue side. Like go to market finance on B to B through SMBs through Enterprise a lot more art and all a lot more bets with higher variance outcomes involved. I think. Then also, if you work down the P&L like staffing, because it might take more marketing to get to a customer with your ACVs of $50,000, how much money do you actually have left for engineers and for product managers? Right, so not saying that they're not important, but you also need to have cash. Right. Thinking through that split is much more important in a B to B world.

Rose Punkunus
So whether it's a science experiment or not, it's all tricky. There are a lot of metrics in all of them, I'll say, going through Subscript and what you're building. And I was thinking through like, how did I analyze revenue at these other consumer companies? And I know not every consumer company is like this, but I think when you have consumers that are not businesses, you're more likely to build out internal systems that have the data structure and schema and that sort of CRM if you will, that's more internal. Right. And again, these days, who knows, maybe there are SaaS products for that. So maybe you're not building that in house. But I think any consumer business at scale does build out that in their own engineering tech stack. So then you can pull that data into a data warehouse, dump it into snowflake wherever, put a tableau on top of it and pull together a lot of what Subscript is doing internally at a company. Whereas if you are going to the PM, you may not have that many resources for so many engineers and data engineers to be doing this for you. Right. So you have a Salesforce, you have some other CRM, you have some data there, you have an ERP.

Rose Punkunus
Again, the data is not always aggregated in a way that's relevant for you and so there is an opportunity to look at the data, but you need to either hack it together in your own Excel or there are products out there these days that you CAC get to the same consumer type analytics, revenue analytics in a SaaS product.

Sidharth Kakkar
Totally. It's really interesting because to your point about when you're talking to customers versus not talking to customers, I find that that governs the complexity of the contract. When you don't talk to customers, they don't really have that much choice. Right. They can't be like well, I'll pay you after POC for two months at a time and only when it's a blue moon or something like that, they can't do that and so they can just tap the button whether they are or not, who is the highest salesperson on the other end? You can negotiate. So you do and things look a little more complex.

Rose Punkunus
Yeah, well, I will say the consumers do have choice. It's just that it's not worth the company's time to convince the consumers who are not choosing your product. I think they still have choice. They can choose. But definitely on the B to B side there is more of that tango and negotiations going on.

Sidharth Kakkar
Yes, totally. So you've had all these experiences in consumer companies and B2B companies as a data person, as a finance person of both and how did all of those things sort of culminate in you being inspired to start Sudozi?

Rose Punkunus
Yeah, so I'm like an accidental entrepreneur. I never really thought about starting my own company. I always enjoyed being creative and enjoyed looking at data. And after reflecting on some of these experiences as VP of Finance, CFO, I really thought that there is this data stack within finance that could be automated in a pretty scalable and consistent way across different companies. And of course there are some customizations, your chart of accounts are different, but there is a big chunk of that that could be automated. And so at Uber, we built out internal processes for procurement. And this goes to the relationship building. We wanted everyone who should have access to actually see what was being approved. We didn't need all their approvals. We wanted them to see that this was being approved. We're spending X million dollars here, you know, Y million there. These are the reasons we're spending the money and being really transparent across the company about what we're spending money on and why then linking that to the budget and having visibility on how you're tracking against your budget, even if the budget was very large, like having some visibility on how people are tracking to the budget.

Rose Punkunus
We had resources to build that out internally, and I think as companies don't have resources to do that internally. My personal experience is your relationships with external business holders suffer if you are making them go through this really annoying procurement process where 10 million people get pinged every time they want to buy a subscription and they don't really need to know. And it's just really a bad use of resources for finance and for your business stakeholders to be like, spending time on these expenses where a lot of that can be automated, the insights from those investments can be automated, and you can be spending more time making decisions for the business. So that's kind of how I strongly felt in my last CFO role. I actually wanted to buy a product that did essentially procurement, budget tracking, department relationship building on the finance side. And I couldn't find it. So I did what I think you might have done too, which is, let's just try to start this business, see how it goes. And we're still early, we're a Seed Stage company, but our goal really is to allow finance teams to focus on building those skills and building those relationships with different business leaders.

Rose Punkunus
Not spending time in the meeting, looking for some random contract that's about to renew. It's not a good use of time. That should all be automated and basically surfaced to you. Then you make a decision if you want to continue buying this software, let's call it software, or engaging with this consultant, what that consultant is going to do to drive more business for you. And so that's what we're all about at Sudozi.

Sidharth Kakkar
That's awesome. I think there's this theme in how you think about the world. And I think also how I think about the world that I really like, which is the finance team's time can be spent in, like, a much better thing then it's spent on right now because you get these brilliant people working in the finance team and then they're, I don't know, staring at reading the same clause in the contract for the 15th time or going cross site on a spreadsheet where they can be doing something so much more meaningful for the business. And I really appreciate that.

Rose Punkunus
Yeah. And I parallel it with the trends and investment in the ecosystem we've seen in, like, ML Ops. Right. I think finance leaders at companies don't do a great job marketing themselves and showing how much work they need to do to get the financial reports to the way they are. And there's a lot of operational work that goes into doing that. And so, just like any machine learning engineer, if the data set is not clean, if it's not in the right schema or right place, they can't actually do their job.

Rose Punkunus
so finance can't actually do their job of telling the story, of helping them make decisions if the data is not in the right place. And so I think that there's an opportunity to really accelerate those individuals, particularly because you only have a few in a company. You don't have, like, armies of finance people.

Sidharth Kakkar
Yes. Totally agree. This has been awesome. Thank you so much for taking the time.