In this week's episode of the Enterprise Monetization Podcast, Sandeep Jain sits down with Prafulla Patil, Head of Enterprise Applications at Mapbox
Hi. Welcome to the 17th episode of “The Enterprise Monetization” podcast. And this is your hosts Sandeep Jain. In this podcast, we invite thought leaders from enterprise monetization space, that CPQ billing usage, so that you can learn about challenges, opportunities and best practices in enterprise monetization. Today, I'm actually thrilled to invite a person that I met about four years ago, when I was exploring this this interesting field about CPQ. He and I had an excellent conversation then. And we have kept our conversations alive over the years. So it's absolutely thrilling to invite you Prafull to this podcast. Now, before we get into this, let me introduce Prafull to our audience. Prafull is an accomplished leader in the technology sector with an extensive background in enterprise tech. And he has a very strong passion for mentoring and advising startups. So we'll hear a little bit more about that. Currently, Prafull, serves as the head of enterprise technology at Mapbox, where he's responsible for leading the company's digital transformation. Prafull is going to tell a little bit more about Mapbox in a second. Prafull has a very impressive career trajectory. He has held several leadership positions at companies like LinkedIn, Atlassian, Flexport. And he has driven a lot of systems transformations. I think this is the time and I had met Prafull, when I was doing research in Quote-to-Cash. Through his journey, Prafull has also amassed an interesting experience across CPQs of all different types, starting with big machines, Apttus, and then Steelbricks or Salesforce CPQ. Now with that, it's very exciting to have you Prafull on the show.
Thank you. Thanks a lot for having me.
Awesome, Prafull. So before we start, can you share a quick fun fact about yourself?
So when I started in college, at that time internet was not that available to us. So I ended up doing programming in college without internet. And then also in start of my career, it was pretty much read that giant Java book, read Tomcat documentation and do Java programming. If you are stuck, ask your colleagues and that is how I started.
Understood. Share with me another fun fact about yourself. So this is an interesting one, but share something that you haven't shared with somebody else, because I'm pretty sure your friends would know about this one. But is there something else you want to share the other folks?
I can twist my palm and it will make this noise, I do not know if you can hear it. It is just like on-demand
Now, that's a fun fact I'm talking about. So the kids get scared, you can scare kids on demand.
Some of them are very surprised. They're like, do it again and again and some of them are like what's going on?
That's interesting. Let's get into a little bit more serious things about your background, like how do you came about at Mapbox? I think you have an interesting trajectory. So just give us a sense of how you decided to move from one role to another to where you are?
As I mentioned, I started as an engineer, and implemented end to end manufacturing systems, started working and had a lot of time, great fun, and did some transformations for comments, then moved to Silicon Valley and then joined LinkedIn. LinkedIn was very small at that point of time 2012, got opportunity to build a lot of systems from the ground up from millions of revenues to billions of revenues, how do you scale company, and I was lucky to be at the core of this transformation. So we ended up integrations with billing systems, setting up sales systems, we acquired companies, integrating their stack, work on customer master product master. We did a very thorough RFP for CPQ. And that is where we engage with few vendors and dig very deep dive, and also got opportunity to grow my career at that point of time from engineer to become a leader, and I also build global teams. I think after six years, I had great opportunity to do this again, and I thought let's do this again. And it will be really fun. Atlassian is a company which I love their product. So I went on there to work on commerce, sales and partner systems. Got opportunity to do some of the CPQ things there. And that is where I also got an opportunity to interact with Max Roodman, who started SteelBrick. So work with him on some aspects, did mergers and acquisitions for three of the Atlassian from a system standpoint, build team in Mountain View, Austin and Bangalore, also deep transformation in partner systems. Then later, I wanted to see how startup world works. So got opportunity to build systems at Flexport with GTM systems there, and also set up the integrations team. After that, then I came to Mapbox. And here I am head of enterprise technology. Along the way, I have been advising startups and returns a product strategy. And then a lot of folks reach out to talk about business systems. And I love to talk with posts, learn from them and share my experience. So that's overall where I am.
Awesome. Now, thank you for sharing this journey Prafull. And I'm pretty sure there's anybody from startups listening, they want to reach out to Prafull, they can do that through LinkedIn. I'm assuming your previous employer. By the way, my co-founder, Jeff, he comes from Atlassian. So we have a very friendly sort of relationship with Atlassian in some ways. Let's just deep dive on what Mapbox is, what Mapbox does, and your go to market your systems. But let's start with what is Mapbox do?
So Mapbox is a technology company that provides location and data mapping tools for developers and businesses. What is the most impressive thing about Mapbox is the product. Developers love our product. Our API's are amazing user experience you are able to build through Mapbox API's is top notch, like amazing experience. Our platform offers customizable map geocoding, direction, navigation, special analysis, all things geo technology, we offer that. Our API and SDK enables integration of custom apps, location based services into web, mobile, AR VR. And our clients range from startups to enterprises. Some of the customers are BMW, Kia, Instacart, Strava, Tableau, all sorts of aspects. I think one cool thing I will tell which everyone can relate to is, vaccines, that girl, so that particular site is powered by Mapbox. And President launched that during the COVID. Also recreation.gov. A great site that is also powered by Mapbox. If you are on a phone and you see a map, if you look at right bottom or left bottom, you will probably see Mapbox there most of the time. I started observing this after joining Mapbox. Oh my God, it's everywhere.
Cool. What is your role at Mapbox? We know that you're leading the digital transformation there. But can you talk a little bit more about this? How's your overall team structured? And what do people do?
In summary, I'm playing like a CIO role at the moment. And what does that mean is I have IT enterprise applications, security and compliance, data engineering, business analytics. So these are the teams I lead. And really, what we are doing is, helping run company and making sure our internal and external customers are successful, and drive that digital transformation make ourselves successful and more efficient.
How many employees is like how big Mapbox just to swing?
Around 800 employees.
Understood, and you guys are still private?
Yeah.
And the employees based on the US or you guys geographically distributed?
We have a global presence. We have folks in Europe. We have folks in Japan, and then we have a strong North America presents as well.
Awesome. Let's go into your business model now. Can you talk about first what are the channels through which you sell?
So our main channel is self-survey as you go. And where you start free, we have very generous free tier. And that is where you will see developer signup. And they start hacking, building some cool apps and every product has a meaningful prettier, and they can sign up and build without friction. As users grows, they will get volume pricing built in, and there is no negotiation necessary. That's our one channel. If you want to assign some commitment, then you receive additional discounts. And that's kind of offline channel which you can go through. As far as our products are concerned, we have 12 skills, we have pretty cool pricing calculator on our website. So you can go and see what products you're going to use, and it will give you how much you're going to spend. So very transparent and upfront for customers, how it is gonna play out for them.
And in terms of channels before we move on from that. So you said you have selfserve, you have a sales lead motion, do you have resellers or marketplace?
No, that is not that significant.
Understood. And in terms of the actual business model, do you do subscriptions? Is it usage because you're talking about API? So I'm suspecting it's going to be usage?
It is mostly usage base. And as you can see on our website, some subscriptions are like a support plan, you can buy a support plan for a year that is the subscription aspects of our SKUs, but rest of them are consumption based model.
Let's talk a little bit about that in a second. But how many reps do you have if you have a sales now, just so that folks can understand how big is you?
I think around 50 plus, if I'm not wrong. That's our number.
And can you talk about what are your systems internally in a Quote-to-Cash perspective to serve your self-serve and sales lead channels?
So in terms of our Quote-to- Cash stack, we use Salesforce for CRM. For quoting, we have some internal thing built on top of Salesforce and we use that and then that is integrated with stripe and NetSuite. We use stripe billing. And then as far as how we get usage, we have some internal stack that kind of gets the usage for us.
And what are the systems in place before you joined or are these systems that you built as you went along?
So some of these things were in place, but in my two and a half years of tenure, there were a lot of things which we had to just scrap and rebuild from scratch, and some of the things we had to improve. So that is what we did.
Got it. And what are the biggest challenges for you, Prafull, as it stands to Quote-to-Cash?
One of the challenges will be like, like, up, there is a new product coming in. And that doesn't fit the current model, like, like how we do calculate usage and all that kind of stuff, then it needs to go through the whole thing, and it might take a lot of time. So I think that might be one last thing I will say.
I heard you say earlier that you build the coating system yourself. And I heard earlier that you have experiences across different CPQ. So tell me the story here. Like, why do you guys decide to build something ourselves because I don't hear that very often.
When I got here, we had a first version of that system and it was okay. What I saw is, we can make this better and it is workable. I think, as the company scales and grows, we probably need to do something else. But given current at the moment, I think this makes sense. Implementing CPQ is a very big undertaking. You got to have everyone across the board involved. And you need to be really sure whether this is the right call for you, this is the right moment and is there a right ROI on this particular aspect? We pulled that what we have right now. We are using Stripe and it has a lot of coating functionality. So we are probably in a good spot. And our structure is not that complicated. So those are the reasons.
And another thing I want to double click on is since you have a primary usage based billing model, maybe we have talked to public companies that do usage based model. And there is minimum commits that apply. There is a wraps of minimum commits that is prepaid credits. So there's a lot of different levers on consumption based billing. And I hear people talking about how do I operationalize this? Like, that's a big mess. I didn't hear you say that. So there could be two reasons. One is, you have somehow figured the magic sauce. And in which case, I want to talk about that. A second is, your usage billing could be simple which is here is the red card, here is credits, just eat it until you're done and pay me more. So what is the story here?
So our usage model is very simple. So what does that mean is, customers will come and they will buy a prepaid particular specific credit. Let's say there will be customer who will buy $20,000 worth of credit, and they can use any product we have and on monthly basis, they will get invoice and their credit will get reduced. Once their credit is expired, they can get more or they can go on to pay as you go model.
Got it. Do customers come back to you and say, “Look, I've exhausted all my credits, and my credits are increasing. So I want better discounting, I want to renegotiate the contract.” How does that work?
That is typical sales motion. I think that is a typical sales motion and it will happen.
In this case, one of the things is because you have an internally built CPQ. And the billing is happening where this usage billing happening. Are you dumping usage data directly into stripe?
Yeah, it goes to stripe.
It goes to stripe. Then how does your CPQ or the internal system talk to stripe? Is it through the CRM, like Salesforce CRM? Is that where the connection happens or you’re CPQ?
We have usage information in Salesforce as well. I think the team has done a great job bringing and implementing customer 360 and getting all data together in a warehouse, in snowflake and sharing it across the system. So we have insert data in CRM, we have a reference of that in stripe as well. So that helps us.
Another related thing is one of the problems that I've heard, and you and I talked about in first talk is the product catalog proliferation. So the product catalog is in multiple places. I think he talked about launching products quickly as one of the challenges. How does that discrepancy play out? Like, you must have a product catalog and stripe in your CPQ tool. I'm assuming that you've built and probably in Salesforce CRM as well.
When the company grows, this is the real problem. I think most of the companies have this problem, which is the prices are not in sync, one changes, and another doesn't know. Right now, I will say it's a manual. And we it is maintainable given the SKUs we have. But I think with the scale, this is gonna be a problem.
And, Prafull, on a related note, what are the top priorities for you in the next few quarters?
I think some of the integrations like integrate tax system with billing and all that kind of stuff. And then support revenue acceleration programs, as they come, I think that's what we will be focused on.
Can you talk a little bit about what does revenue acceleration mean? Is it like, a new go to market motion that you're planning or is it something else?
It will be. Like, if there are new SKUs launching, and there is gonna be some different ways of pricing. And that might be a driver for revenue acceleration, so there might be some other aspects, for example, unlocking specific currencies. I don't have any foresight into that, but the example I just shared, that's a good one.
A related one, which I was thinking about asking, but I forgot, because you have a very well developed self-serve motion, which was what Atlassian had your previous employer. One of the challenges I've heard is how do you merge the self-serve in the sales lead channel? So when I sell serve customer comes in? They grow big, and they want to be now doing a deal with you? How do you manage like expansion or contraction from sales lead to do a self-serve?
This is a great question. And this is definitely a challenge across the board in the industry I see. And I will say to address this problem, you gotta start with data. And what I mean by that is, you need to know what is happening in your online sales motion, who your customer is, you got to identify. So what I will say is you can use some data tools like zoom info clear with whoever comes through your online motion, you got to identify. Sometimes what happens in online motion, what I have seen in companies is you just have very limited data about that particular customer, you got to enrich that. Then after that, you need to get them in a CRM and if based on information you have gathered, suppose there is some particular user who signed up with a credit card, but they are a fortune 100 company, your CRM should tell you that. There should be some alerting mechanism to tell that person and company person just joined. So that you can have a rep assigned and kind of start working with that person. Now you need to have a good alerting mechanism built in to say, there is particular customer who is on your pay as you go channel, but their usage continues to grow. Then you need a way to alert your sales team that Hey there, this particular customer, their usage is going growing month over month since last three months, probably good opportunity for us to engage in conversation and talk with them how we can make you successful, what are you trying to do and engage in that conversation? So building these kinds of alerting, which we have built. I built this at Mapbox as well as I built at my previous companies is very helpful in those particular cases.
This makes a lot of sense. Thank you for sharing this, Prafull. So when a person joins a self-serve, he directly don't go into Salesforce CRM. It is a certain criteria before they get an entry into Salesforce CRM lead, or something like that?
I think they go to CRM, but what triggers for them so that they are flagged to your sales team. For example, you might have 20,000 people coming through your channel every single month. But how do you look at that data make sense of that information? And then have a refined list for your salespeople.
This is like your self-serve leads, like how do you want leads flow into your salespeople? I think one of the challenges that I've heard is, people don't want to dump all their self-serve leads into Salesforce CRM, because it does not scale is what I've heard. And that's really my question was that sometimes companies don't even add leads. Is that a problem that you have or do you ever really go into Salesforce?
For us, I don't think it's a challenge. But with the growth and scale, I can see that it might become a problem. And at that point of time, you probably want to limit those to marketing system, and then based on certain criteria’s bring them to CRM, that makes sense.
Shifting gears a little bit. You have worked at several companies worked with several tools, you must know like what works or does not. So do you have any advice or recommendation for vendors in this space about how they can do things better for operators like yourself and your team?
I have a lot to say, actually, because I started from big machine like dealing with XML files, and then at SteelBreak, and then I looked at that at some point of time as well. So, one thing I will say is simulation. So CPQ is a great tool but it is also very dangerous tool. And what I mean by that is, if you have a junior CPQ administrator configure some rule and that rule ends up giving huge discount to specific customers. The code might be already in front of customer at the point when deal just catches it, it's already late. So there can there is a potential of a lot of accidents happening. So this is one thing what I have not seen in CPQ space is, if I change this pricing rule, how does this affect my existing customers? How is it's gonna affect me? There should be a way to simulate. This is a very hard problem to solve. I think this is very key for big customers, because the stakes are really high. And this will come in really handy for pricing teams to come up with pricing strategies like, what they need to do there.
And what do you do today that it does not exist? So you're saying that what if the price of this usage API was ‘X’ instead of ‘Y’, what would happen?
I'm not talking about this in a Mapbox context, I'm talking more in a general context. I think what I will do if I have to do this today is I will go to Data Warehouse and I will through some SQL or through some data transformations, I will try to simulate. But what does that mean is then you gotta have whole team involved in this activity.
Understood. And you want this to be like a no quote, just change the pricing, like create a new pricing, and then pretend this was the pricing for the last year? What would have happened?
There should be some way of getting some information. I know, this is not an easy problem.
Vendors are not paid to solve easy problems that you can do yourself.
I think the other thing I will say is preview pricing. This is like you are launching a beta product. Six months, you want it to be in the wild. You can use this for six months. But we are also going to give you insights into how much this is costing you, so that you are better informed. And what I mean by that, let's say, I launch product one as a beta. And now the customer keeps using it, their usage keeps piling up, then they go to their portal, and they see I am using $2,000 per month for this particular feature. So it makes sense to me. So providing that transparency, I think that is a great way to build trust.
Let me just understand that. This is interesting what you're saying. So if I'm your user, I come to Mapbox or any website. It's a free product, because it's beta. Now you can say, “Sandeep, this API’s priced whatever ‘X’ cents or dollars per API call, we're gonna give it to you for free, because we are in beta trial. You still want when I go to the product to see that accrual.” So one thing is I see this is $0. And I have one API’s price this but what you're saying is based on my usage at the end of the month, create a sample invoice. So that I can understand well, I would have paid this much amount. So I have a sense of value that I'm receiving from you. And what I'm getting back in dollars. I've seen that thing before.
You’re right. And this is a challenge for consumption based businesses. Because with subscription, what you're gonna buy or what's gonna happen? It's very hard for me as a customer to figure out well, I will start using your thing.
Understood. What else? This is good.
I think some of the things are CPQ and billing transformations are big undertaking. I think there needs to be deep discovery that there is there is a strong will and executive support from IT and business side and a whole company. Otherwise, what I have seen is implementations get stalled, they are not successful. So that will be my advice to vendors as well as customers or companies who are, who are looking to do these kinds of implementations. Also, some of the other things I will mention is UX, user experience should be front and center for any vendor in the space. Sales teams are really great at spinning up Excel sheets, the moment your coding experience is not great, what will happen is there will be a pricing spreadsheet, and that will start circulating. So I think we are competing against that.
Another related question. There has been a talk about this product catalog proliferation with CPQ billing and usage systems. So integrating that in one platform, do you have any thoughts on this unified architecture versus having a split architecture?
In terms of unified architecture, it can offer several benefits, such as consistency easier report to reduce integration challenges. And by having single system that can handle both CPQ and billing processes, I think business can eliminate need for separate skills and it can lead to fewer data discrepancies. Also, it can be seamless experience for customers and reduce complexity. I think at the same time, unified architecture may not be suitable for businesses especially those with CRM, and then complex M&A is happening with unique requirements. And then best of breed approach, you can't implement. For example, you get best CPQ and best billing system, you can't do that because they are integrated. So for example, during my time working on CPQ and billing at LinkedIn, we faced several challenges of syncing data between multiple systems. Part of the product catalog, we ended up building product master, I think it was the right call, because the company just grew in a big way but that challenge can be solved with unified architecture. So overall, I will say, the answer is it depends, depending on the size and we need to have careful evaluation of pros and cons of each approach. Consider factors such as integration requirements, scalability, vendor support, and then we can make a right choice.
So if I just summarize, what do you say. Are you saying if you're early stage company, then unified architecture probably makes more sense? But if you're a bigger company, with already have bigger systems, then you probably have to think a little bit more before you go towards that architecture?
Absolutely. I will say you are a small company, it's a no brainer. You don't want to deal with systems complexity, and you want to focus on growing revenue and supporting that.
Got it. Any thoughts on intelligence and CPQ or maybe intelligence is what ChatGPT these days? Or any things that you want to tease or give a teaser about? What have you been thinking about this? How intelligence can play a role?
This is the time everyone is thinking about it. I think, there can be interesting ways we can use AI in CPQ space. I think on customer facing aspects as well as on the internal like a pricing strategy aspects. And what I mean by that is, you can really build a sophisticated recommendation engine based on data you get from your warehouse usage, customer behavior, marketing signals you get, and then recommend them additional products. And that can take a lot of burden out of salespeople, they can have all these recommendations ready for them to pitch to customers in the sales promotion, you can highlight those to customers in the portal or whatnot. As far as pricing strategies are concerned, value based pricing is something you can figure out new ways of how you can position different prices to different customers. So there is a lot we can do there.
That's interesting. We think about this a lot internally. So it's good to hear what operators think about that, as well. So you're right about its pricing. It's like what discounts can I offer to my customers? Like, is my customer even credit worthy for this deal? Have they been paying their things on time? And that intelligence is sometimes with the finance people, but that information doesn't feed back into the sales on amendments or renewals. So that's good to know. By the way, we didn't cover this topic earlier. So let me just quickly cover that one about amendments and renewals. Are these pain points for you in your business because that's what I something here consistently for SaaS businesses?
I don't think are pain points for us as well as the renewal. So amendments in our context are like a customer buys certain credit, their credit is exhausted and they will buy a new credit. So that's the kind of amendment we have. As far as renewably is again, there is a certain period they buy credit for and it ends at that point of time. So they do renewal. So given the nature of business, I don't think it's a challenge, but based on my previous experience where I work and companies were like subscription based. And there are a lot of amendments happening in the middle of the tenure, and then you do debug, put back and then it gets complicated with bigger customers, there are like six opportunities happen during the year. And at the time of renewal, you are trying to make sense of what really happened? Which product this customer has? How do I make sense of all that? So when the company is very much subscription focus, it's a real challenge if there are so many changes happening during the renewal, during the one year cycle of the term of the contract.
That's actually very interesting insight that I don't think is talked about often. And what I mean by that is, the subscriptions versus usage. There's been a debate whether we get a subscription company or a usage company, but I don't think what is talked about is and enough has been talked about on the subject. But the fact that with subscriptions, you have a term to deal with. And whenever you're making amendments, there is core term, which means you have proration, and credits and debits. And if you're in the usage based model, which is where you are right now, you're not pricing it on a subscription term. So it just eases your way to do business. And I don't think that fact is widely understood. So we'll call that out.
Yeah, that is a great point. I think a lot of this complexity with the usage base goes away.
That's interesting. Prafull, this was an amazing conversation, because we dealt with a lot of different subjects around self-serve, usage, CPQ, and billing. But before we let you go, could you recommend or do you have any recommendations on a resource that you will like to share with our audience, maybe a blog or something?
I would love to share two podcasts that I'm a big fan of. I believe most people are familiar with these podcasts, particularly "How I Built This" by Guy Raz, which happens to be my favorite.This is a valuable resource for anyone who is interested in starting their own business. Also, it's a great way to learn from experience of others and the way Guy interviews people, it's just amazing. So I will not miss a single episode. And then another one, this has been with me for many, many years, The Moth, probably not that known by “PRX”. I love 'The Moth' podcast, because it's a great way to hear true stories from real people. And I'm always impressed by the courage and vulnerability of the storyteller. So this is another one as well.
Thank you so much for sharing these two resources. Actually, I knew what the first one because I used to be in podcasting earlier as a distributor and whatnot. But anyways, the second one was an interesting one, actually. And I have talked to the “PRX” people in the past. So with somehow I missed this one. So thank you for sharing that, Prafull. And thank you for your time today.
I enjoyed our chat and thanks a lot for having me.
