Self-Serve Growth: Marie Gassée on leading growth at Box & Confluent

October 23, 2023

Listen Now

Never miss an episode. Subscribe now.
Thanks for subscribing! We'll email you when we publish an episode.
Oops! Something went wrong while submitting the form.

Episode Summary

Coming out of Stanford Business School, Marie Gassée took part in a rotational leadership program, which led to the creation of Box’s online sales business unit.

There, Marie led a cross-functional initiative to add a self-serve motion to complement Box’s traditional sales model.

After 6 years at Box, she spent 2 years as Confluent’s VP of Growth — where her main focus was commercializing their thriving open-source developer community.

Now, she’s Head of Go-to-market at Column — a bootstrapped bank slash software startup that helps developers build financial products.

Marie Gassée has an impressive track record building self-serve growth programs.

She led the launch of Box’s product-led motion and helped Confluent turn a thriving community of open-source users into customers.

But none of that growth came easy.

On today's episode, join Alex & Marie as they discuss:

  • The growing pains of adding self-serve to Box's traditional enterprise motion.
  • Monetizing an open-source product at Confluent
  • Early growth lessons from Column
  • Scaling herself along her career journey

Links and References


Box's Rotational Leadership Program

Alex Kracov: I'd love to start today's conversation talking about your time at Box. You joined Box after you graduated from Stanford Business School, and you started at Box with the rotational leadership program. Can you talk about what it was like to participate in the program?

Marie Gassée: Yeah, definitely. It was really cool, actually. I think these types of programs are typically, I think, a lot of marketing but the substance is kind of crappy. It was a really cool opportunity. I basically spent two and a half years popping around the company, three different roles. I did work on Box's IPO — which I had no business being a part of, but it was so cool — focused on customer economics, which is great in that kind of IPO function and then another kind of sales strategy function. Then I worked on a new online sales self-serve initiative, which I think was probably one of the coolest experiences of my career and got me interested and got me plugged in to the growth space. So, yeah, it was great.

Alex Kracov: You eventually ended up picking online sales product management as the focus out of that rotational program. I don't know. You mentioned it was cool. What was so interesting about it to you?

Marie Gassée: I think I realized being close to revenue was what I enjoyed and what seemed to me to be one of the most interesting and valuable parts of the business. What I didn't like was this old school, no offense, Oracle sales motion that I had seen in b2b SaaS. Online sales seemed like, okay, you're close to revenue but you're data-driven. You're close to the numbers. That just felt like a perfect match. Then the idea of Box was not a small company and the idea of being in a growth area building still was really appealing. So, yeah, it was great.

Data Infrastructure for Self-Serve

Alex Kracov: From my understanding, I think one of the first things you did once you picked that focus was to put this data infrastructure in place to measure the impact of self-serve and just to mire the program. Can you talk a little bit about why did you focus on the data infrastructure to start, as opposed to running a bunch of programs to try and drive more self-serve signups?

Marie Gassée: Yeah, I wish it was a function of me being very smart and having a ton of foresight. But honestly, the current state was so bad we had to do manual recon at the end of every month to actually know what the revenue numbers were in self-serve. It just wasn't even an option not to invest in the numbers and in the tracking data infrastructure. We wouldn't even know if we were doing something right. And so we were almost stuck. It bore so much fruit down the line. If we hadn't spent all this time like first getting our act together, from a data infrastructure standpoint, it would have actually slowed us down so much, a year or two later. And so now when I do share thoughts and best practices on a growth motion or a self-serve motion, I'm a little obsessed with the tracking piece. Because I saw how valuable it was and how horrible and worthless all your efforts are going to be if you don't actually take the time to measure it. It's not sexy, and it's not fun. But it's the only way.

Alex Kracov: What size and stage was Box at the time? Do you remember how big the company was, or even revenue? Because it's so interesting. I'm sure it was pretty big, but it didn't have the stuff in place.

Marie Gassée: Yeah, we were probably definitely over a couple 100 million. We were public. We probably were like 1,500 employees. And I think there's like, the infrastructure the team had on Salesforce tracking, for example, for the sales team was good. There's decent product analytics. But because the self-serve motion had been neglected or not really important for a while, there was just absolutely nothing there. So, yeah, it was pretty dire. It wasn't great.

Alex Kracov: I always think people always say it's really hard. If you do the top-down sales motion, it's always really hard to go back and do PLG. Box pulled it off. But it makes sense why it's so hard. Data infrastructure is a very good example of the tedious work you have to do to go back and make that happen.

Marie Gassée: Totally. What I tell folks who are a little earlier on, I'm like, do it now. It's so horrible to do it later. Just take a little bit of pain now. Because how disruptive it's going to be if you were Box — it was 1,500 employees — and not having kind of your sh*t together there? It's going to be so problematic. Just do it now when you're series A maybe and have 100 employees. It's going to be so much less painful to get some sort of structure now than doing it in two years when you have to unwind a bunch of legacy crap.

Alex Kracov: It's a good reminder. I need to do my stuff at Dock a little bit better. We have some infrastructure in place, but it definitely can be improved. It's always hard. It's like, alright, do you focus on the core product, or do you focus on the tracking? There's trade-offs in everything. Do you remember at the time what sort of the north star metric was when you're building out this data infrastructure? I assume revenue is a big component of it. But what other things were you thinking about tracking as you built out the infrastructure?

Marie Gassée: Our north star was definitely new ARR, new recurring revenue. So much more on the acquisition or the expansion side. We weren't thinking about churn, which we can talk about. It was kind of stupid. But we weren't for a while for a couple years. Then we kind of had a cascade. So there's new ARRs and North star metric. Then what are all the pieces that feed into that? All the way down, furthest away from revenue, you're going to have maybe ad impressions or unique visitors on the website, and then signups, and then conversion to paid, and then expansion. But we really were pretty focused on backing out of the revenue metric and not getting too cute with like, no, signups. We're going to optimize for signups. Okay. Well, you can cheat your way into getting 100,000 signups that are worthless. So let's always anchor on revenue. Because you can't get too cheap when you're focused on the ultimate metric that matters.

Alex Kracov: It makes sense. Was there a specific product activation goal or aha moment you were trying to get people to push towards?

Marie Gassée: We definitely had a threshold of activity that someone does during a trial that we wanted folks to meet in order to be much more likely to convert. I will say, we really did struggle to drive early onboarding activity and early adoption. If I reflect and I think, hey, what are the projects that were amazing and, from an impact standpoint, what are the ones that we just never quite figured out, I would say early activation onboarding we never found an obvious or quick fix. Maybe that's a signal of my impatience certainly at the time. But it's like getting people to use your product is not something you can just have a better button or a slightly better flow. That's deep, and that's pure. You have to really invest.

One of the maybe shortcomings I had was this impatience that was much more satisfied by a new pricing page or reducing friction and signup. The onboarding, the activation stuff, warrants this almost really academic patient, pure initiative that, for some reason, we never totally cracked when I was there. I'm sure folks who's followed me did a much better job at it.

Alex Kracov: It is such an interesting thing, and I'm dealing with this at Dock too. Because we get people to use the product. They self-serve, and they start playing around and building workspaces and stuff. But then, there's moments where people get stuck. And I'm like, why did they get stuck? I think what I've learned is, it's not even about the product itself and clicking the buttons. It's about the business outcomes that they're trying to drive. It's hard to set up an onboarding program, and they get slowed by the internal processes and the things around the product. That's really hard to think about and push towards as you think about self-serve motions. That's why I think sales assist stuff is getting even more popular. Because that can help push some of those more business outcome things when you have the self-serve side.

Marie Gassée: Totally. I sometimes think of the evolution of growth, or like self-serve, or online, whatever you want to call it, PLG. The focus has just shifted over time. There was maybe like five or seven years ago. We were all in this mindset of, if you have some great product tutorial at onboarding, it'll solve all your issues. Then I think we optimized as much as we could and realized, all right, we've hit some diminishing returns. Now we have to move on to another lever we can pull. And to your point, I think sales-assisted is now the thing of, like, okay, this is the next lever that can actually get impact. But it's going to look a lot different than a cute tooltip in your product dashboard.

Cross-Functional Teams

Alex Kracov: You eventually built a cross-functional team — I think it was called the online sales business unit — to work on self-serve program. I thought that was super interesting, because this stuff is super cross-functional. Can you talk a little bit about who is on the team and what that collaboration looked like?

Marie Gassée: Yeah, this part was awesome. This part, I think we got right at Box. This was my boss' doing. He kind of insisted and forced the executive staff to put money where their mouth was. Okay. You want to do this initiative. You want it to be successful. It hasn't in the past. You have to dedicate resources. You have to ring-fence teams to do this. And so we built a business unit that had eng systems, which it took me a while to realize how important that was — basically, making sure that all the systems are talking to each other, a big part of the data infrastructure piece. Product, analytics, we had design, marketing and then with support from finance and other functions as well. But yeah, bringing together all of those functions all focused on the same thing that we actually got sh*t done. It was pretty awesome.

Don't get me wrong. We didn't start with all this big fancy team. We started with, I think, three eng, me on product, and someone helping on systems. We had a few wins. Then we were able to make the point, okay, hey, let's bring in even more functions to move even faster, have more impact.

Alex Kracov: Can you talk through like an example project that you all worked on together?

Marie Gassée: Yeah, I think one of the wins we had was on an upgrade flow. We have a lot of folks who may be converted to trial. Maybe they're paying us like $500 a year. Not a lot. How do we get them to upgrade? That was super cross-functional. Because, obviously, there's a huge eng component, a big product component, design, what the UI is going to look like, how do we measure everything? So that one was pretty cool, especially because there were so many pieces to it. It wasn't just a flow. It was like, how do we show value in the product of things that users don't have access to yet? So we were previewing, like, hey, here's an insights tab that you could access but you'll need to upgrade. Then we had this really cool upgrade flow. That was one of the projects that I thought was really cool. Everyone was involved in it, and it happened to work really well, which is always great.

Pricing Page Design

Alex Kracov: That's awesome. That's really cool to be able to tease out the features that they could be getting and to provide a little bit of value, but not too much that they got to pay an upgrade. I've heard you also talk about pricing pages. How do you think about good pricing page design, and how does that impact the self-serve experience?

Marie Gassée: Yeah, I am a big fan of the pricing page as a lever in large part because we had such success. We launched a new pricing page early on in this business unit. Literally, overnight, the run rate increased by two and a half million. It was unbelievable. It defied what I thought was reasonable, that mainly a UI change could drive folks to purchase more. That's when I started to buy in and like, Oh, my God. This stuff is real. You can improve some experience, some UI, and actually get millions of dollars. It seemed crazy. That was, in large part, a design change frankly.

I think the other thing we learned there and that I still feel pretty strongly about is transparency, like pricing transparency. If people want to buy it, let them buy it. Don't force people on a trial if they just want to buy it. Don't like force people into an annual commitment if they don't want to. Offer maybe a slightly more expensive monthly option. And so that's kind of like a tenant I still hold today of like when in doubt, be transparently flexible. Set expectations. I'm not always right. I'm learning. I'm in a much more upmarket space now, and some of those principles are being challenged a little bit. But I certainly think in the SMB space and the self-service space, you can never go wrong by just being upfront.

Alex Kracov: I'm curious about the sales team who usually wants to hide pricing. How did Box's enterprise sales team interact with the self-serve program? Because I imagine you sort of had these two different motions. There's the self-serve motion, and then there's this more sales-led model. How did they interact with each other?

Marie Gassée: Initially, very poorly. One of the first initiatives we did with the sales team is, we recently took away credit. We're like, hey, you are no longer getting paid on trial conversions that you can't prove you helped. We basically took out, I think, 20% of their quota attainment on the SMB side, much less so on the enterprise side. So it was very bad initially. I think we changed how we thought of how sales get credit for things that come online. Then that smooth things over a bit. Then we really got in the mentality of, like, our job is not just to hit these revenue numbers but feed the sales team. So everything we do, we have to come to the sales team and say, like, "Hey, here's how it's going to help you. Here's how you can leverage all these new trials you're getting to hit your number." But yeah, it was not good initially.

We ended up actually moving the business unit to report into sales at one point as a way to try to smooth things over. It's like, hey, the sales leader has to own the AE-driven sales and has to own online sales. In that way, we can force us to get along. It did work, actually.

Working with Sales

Alex Kracov: Did you have a concept of product-qualified leads or something to pass over these good opportunities to the sales teams? How did you flag good accounts to the sales team?

Marie Gassée: Yeah, totally. We worked with the analytics team. We did a pretty basic analysis. Hey, what are the signals that we're seeing in a self-serve trial or a small deployment that signal higher potential? We found what those signals were. We used them to trigger a notification to the sales team to say, "Hey, this one looks high potential. It's worth your time." I like to think it was PQLs before PQLs were a thing. But yeah, that helped get us credibility and goodwill from the sales team. Because we really did try to say, "Hey, this can help feed you." We've tried to actually serve them up trials and self-serve accounts that had the ability or had the potential to have a meaningful impact on their quota.

Alex Kracov: It should be a good thing for the sales team, right? Getting warm leads who are already interacting with the product, that's like a sales person's dream. I always think of the best relationship is just sort of like an alley-oop, right? It's like marketing and growth tosses up the lead. Then hopefully, sales can slam dunk it, right? A bad analogy, but yeah.

Marie Gassée: I think I agree. I think where rubber meets the road is what people are getting paid on. That all sounds good. In our situation, we basically were paying on something and literally took it away overnight. And so it all should work. But then, when you're messing with people's livelihood, people rightfully get pretty mad.

Alex Kracov: Yeah, it's something that's good for the business because you have to payout comp a little bit less and automate it. But yeah, it's worse for the salespeople. But I think they adjust. Do you remember how the sales folks would reach out to these leads? Was there a specific hook or something to get customers to respond?

Marie Gassée: I wish I had a sexy answer to that. But no, discounts. Discounting and then like, hey, we'll help you. Like, hey, at least you'll have a person on the other side of an email or the phone that will answer your questions and things like that. So it wasn't super thoughtful, but it worked.

Alex Kracov: I always thought, and I think this is true, Box was this enterprise solution and then Dropbox was happening at roughly the same time. It was the self-serve version of the same product. It's so interesting to hear you talk about Box's efforts to kind of, I guess, compete directly with Dropbox in that segment. Is that how you thought about it internally? Was this a direct effort to go and try and eat into what Dropbox was doing at the low end of the market?

Marie Gassée: Definitely, I think the other piece was just even though we were strategically more focused on enterprise and certainly a lot of our product decisions were guided by that, actually, a lot of our revenue came from an SMB base. And so it was a little bit more like, hey, we have a lot of revenue coming from down market. We can actually do way better in this part of the market by just doubling down on these efforts.

It was a little bit more like, frankly, it was driven by how inefficient we were from a sales and marketing standpoint at the time. We were running around thinking, okay, how do we spend less? So self-serve, no commission. And how do we make more money? Duh? We actually thought, okay, this is a pretty crappy experience in self-serve, and we can actually probably double our money here by putting a little bit of effort. So it was really driven by that more than a direct Dropbox competition play.

Leading Growth at Confluent

Alex Kracov: Got you. It makes total sense. All right. Let's switch gears and start talking about Confluent where you went after Box, where you joined as the VP of growth. Can you talk a little bit about like, paint the picture of what Confluence product actually does? What was it like when you joined? How big was the team? If you remember rough revenue, I'm curious what it was like.

Marie Gassée: Team-wise, I think we were like 600 employees. Above 100 million, but I don't remember exactly where. Also, pre-IPO, we end up going public, I think, about two years after I joined in mid 2021. The company, very different certainly than what I experienced at Box. So open-source kind of commercialization product. Basically, this open-source software called Kafka which is focused on how you stream data — basically like a new way of thinking about data going from one place to another. Huge adoption in the OSS space, so tons of developer adoption. Basically, Confluent was the commercialized version. So you're using Kafka. You need help, or you need more features, or you want to unlock all these other capabilities. Then you would buy Confluent in order to access that. And yeah, it was very cool.

Definitely a little bit of a chasm between this very developer-friendly OSS facing part of the company that was all about adoption of Kafka — this thing that we didn't make any money on — and then a pretty traditional sales motion, pretty upmarket, initially very reliant on professional services, which was not great. So definitely like a little bit, it was hard to reconcile those two different parts of the company.

Alex Kracov: Yeah, it sounds like you almost had two different cultures. It's like this free, open Internet side helped the developers. Then it's like, alright, your hardcore SaaS enterprise folks. It's very, very different. How did you bring people over from that kind of open-source community to the paid side? Maybe we'll start at the top of the funnel. What did outreach look like, and how did you get those initial at bats with folks?

Marie Gassée: Yeah, I think the biggest thing — and this was a big learning. The biggest thing, the basis of what was successful from a growth demand gen pipeline generation standpoint was basically saying: finding people who were using Kafka, the underlying open source, and then pitching to them here's how that can be improved, made easier, made cheaper, by working with Confluent. Where we mis-stepped pretty meaningfully is data streaming, real-time data. That market is infinite, right? You can be like, oh, my God. Our total addressable market is enormous. Let's find all those people who could benefit from real-time data, and let's sell Confluent to them. What we realized really quickly was, if someone's not already using the open-source software, our conversion rates essentially are zero. And so we're spending all these resources trying to convince people that this new way of looking at data was the best thing for their business. Come to Confluent. We'll take you on this journey with us.

What we realized was like that's too much of a jump. Like, what we really needed was a two-pronged approach. One is this great developer advocacy group that's super developer-focused, just focused on Kafka adoption in the purest, most developer-friendly way. Then marketing and sales team's job was like, okay, once we have this addressable market of Kafka users, then let's show them the value of going from using Kafka to working with Confluent, as well. It took us a while to figure that out. It took us a while to, I think, bring — actually, there was no desire to bring the developer advocacy group to be more commercialization-focused, but at least bring the sales team and the marketing team a little bit more to appreciating that the foundation of everything we do has to be usage of Kafka, the open source.

Alex Kracov: Yeah, it's super interesting and a good lesson around how a niche works, right? You're trying to spread the word around the big idea that Confluent was trying to solve, but it actually worked better when you were just focusing on just the people who are using Kafka and then moving them over. You mentioned that two-pronged approach. Did you focus on both of them, or you were more just on the marketing and sales side in trying to bring the Kafka users over to the commercialized side?

Marie Gassée: Yeah, more so on the commercialization side. One thing I think Jay, our CEO, did really well was he siloed the developer advocacy group and let them do their thing and not be poisoned, frankly, by the sales and marketing team. And so we were just there waiting for the Kafka community to expand, and then seeing how we could action that. I think that was a really good decision and kept it pretty pure. So yeah, I was just on the greedy monetization side.

Alex Kracov: Makes sense. Then what does it actually look like to bring people over and sift through the community? Are you just looking through GitHub and seeing who's really active in open source and who's contributing to project and things like that? How did that go-to-market motion work?

Marie Gassée: Essentially, the biggest successes we had were, okay, what are scalable? Obviously, not clicking around a community. But what are scalable ways to identify what we called Kafka signal? What are the ways that we can identify that a company, a persona, is using Kafka, and then market to them usually more so technical personas? How do we market to them? Like, hey, here's how your life, or your job, or your outcomes can be better by using Confluent? Examples of that, we use LinkedIn skills a lot. Because people would talk about Kafka as a skill they had. We used job postings, scraping job postings to identify if they were using Kafka at an organization. Then through a number of channels — ads, SEM, sales outreach, et cetera — we would target them.

But again, another mistake we made is getting a little too — in my opinion, this was not everyone's belief at Confluent. We got a little cute in trying to go after these more senior business-like personas. Like, oh, we're going to really pitch them the dream of how digital transformation will change everything. I always found that really getting it like, hey, a slightly more senior technical persona, and the conversion rates agreed, was the way in to starting a deal cycle and getting pipeline generated. But yeah, not everyone agreed with me.

Alex Kracov: It makes sense. I'm sure the technical folks really feel the pain of using Kafka and the open-source technology and then really feel the benefit and the value that Confluent was providing. Whereas the other style messaging is more pie in the sky and harder for people to grasp. Wait, what are we actually — buying your digital transformation sounds cool? But it's like, okay, what's the problem you're going to solve for me right now?

Marie Gassée: Totally.

Building Confluent's SDR Team

Alex Kracov: It sounded like you built an SDR team to go out and reach out to those folks. Can you talk about that process and what that was like?

Marie Gassée: Yeah, we had a pretty traditional — initially, a traditional SDR team. We had an inbound team that was focused on ingesting leads and converting those, and then an outbound team that worked across with the sales team and outbound it into key accounts or priority accounts. So that was not anything revolutionary.

What I thought was cool is, we also built a sales-assist team — we called them cloud SDRs — as we were really trying to push the Confluent Cloud product. Their role actually was bridging the gap between the developer advocacy team and then the more commercialization team, and that we really look to them to be super helpful, not salesy at all and be focused exclusively on adoption. Their key metric, at least, when I was there, was, okay, how do we get them to activate and use the product regardless of whether they're going to monetize or hit a certain threshold that's worth much from a dollar standpoint? So that was the first foray into a very pure sales-assisted motion.

Alex Kracov: Would you comp the SDRs on the product activation metrics?

Marie Gassée: Yes, gosh, I'm forgetting the term. I think we called it production activated orgs (PAOs). Yes, PAOs, production activated orgs. And that was the North Star. How many folks who signed up for Confluent Cloud can you get activated? And that's the goal. That was a certain threshold of usage, but it was not linked to pipeline. It was not linked to a certain monetary value. It really was — we're trying to increase the base of folks who are using our cloud product in some sort of regular monthly active or whatever way.

Content at Confluent

Alex Kracov: You touched on this a little bit before, but how did you think about the role of search in the growth program, both paid and organic? Because I imagine you have a lot of developers out there who are using the open-source stuff and looking for improvement. That probably feeds really nicely into a search program. How did you think about it?

Marie Gassée: Totally. Yeah, our search program was great. We had a great person leading it on the team. We saw a ton of success on branded search, in that we said Kafka was part of our brand. Yeah, we saw a ton of success there. I think the foundation of that is, obviously, the structure of the search program was great. But the foundation of that was the amount of content that we had. We had on the organic search side. But also, just in terms of providing content for the paid search side and all kind of our paid programs, we had so many Kafka technical experts. They produced so much gold. So we had so much content.

Honestly, a lot of what we did was on technical SEO, just making sure things were appended the right way or organized the right way. Because there was so much valuable content in our docs, all sorts of materials, white pages, whatever. And so the foundation of it and what I tell folks often is like, invest in content. But don't do BS shallow content that's going to maybe come up on a Google search but not actually provide value. And so it made our lives so much easier on the acquisition, what we call digital growth side. Because there was like a plethora, a never-ending amount of content, that we could just adjust a little bit to make it more lucrative in terms of getting folks into the website.

Alex Kracov: Do you remember how you went about creating that more technical content? Because it's hard, right? We're talking about Kafka. I barely understand what that actually means, right? Then the people who usually know the technical stuff or maybe don't have time in their day to write or aren't good writers and don't want to create content, how did you bridge that gap at Confluent and create that more technical content?

Marie Gassée: We actually were pretty lucky in that we just — because it's a very pretty technical product and that there was this group of focused developer advocates really focused on writing content related to Kafka, the open source. We had so much to work with. We didn't have to hire technical writers to put content that was going to help bring people in to the ecosystem. We did some of that. But we were really lucky that it was pretty natural to the business to just have a lot of things written down about Kafka, which is what folks would search and look for. And so it made it so easy for us to be the number one destination for anything that's Kafka related. It was almost a no-brainer. If you want to look anything up about Kafka, you're going to come to the Confluent website. Imagine how easy then of a jump it is to be like, "Oh, hey, by the way, if you have any issue with Kafka or any aspirations, we're the folks that you can come to." So that was awesome.

Alex Kracov: It's almost like it's a content marketer's dream because it's all just so naturally down funnel. It's like we're thinking of all this stuff at Dock to write, and there's like these more top of funnel keywords like sales or b2b sales which are hard to go after and don't even provide that much value to us. But you have so many different avenues where it's like very down funnel, and it's an easier conversion. So yeah, that's cool.

Joining Column

Alex Kracov: I would love to switch gears and talk about Column, where you are today, which I think is a really interesting company. Can you talk a little bit about what is Column? Why did you join? Who's the founding team? How did you get involved over there?

Marie Gassée: Yeah, totally. Column is so different than anything I've done before. I still, almost two years in, struggle to properly explain what we do. Essentially, we're a bank and a tech company. So any kind of company that wants to operate a financial product — if you think of Chime or even Robinhood, any money movement, if you think of, anything that involves moving money, holding money, lending money — you have to actually work with a bank somewhere in your stack. You cannot do that if you're not working with a bank.

The current ecosystem is horrible, a bunch of legacy banks that have no idea what it means to work with tech companies. And so William Hockey, our Founder and CEO, he was a co-founder at Plaid. He identified this opportunity. So he bought a bank and then built from scratch a banking core and APIs to plug into, basically, the banking, the financial system. And so we work with fintechs, with payment companies, with insurance companies, with any company that's doing any sort of financial capability, and essentially are the infrastructure stack for them. And yes, so different. Certainly, I think, I joined Column out of a slight existential crisis around the idea of doing the same growth playbook for a different company, a different product, gave me existential dread. And so the idea of doing something totally different, smaller, with a proven founder and a great team, I was just like, you got to take a swing. It's been awesome. It's been really great.

Alex Kracov: It's way more early stage than the other companies, right? How many people were at the company when you joined?

Marie Gassée: Oh, gosh. Maybe 30 or 40. Yeah, maybe. Maybe less. Yeah, it's so different. But I think one of the biggest learnings I've had, because I hadn't gotten so small before, is like I don't think I realized how much my soul was getting sucked out from the corporate BS, some of the politics you inevitably are involved with at a bigger company. I think when I started working here, I was like, wait a second. I work a lot. But it feels light because most of my energy is focused on just doing stuff and moving the ball forward, versus 80% of my job being focused on getting ready for e-staff or arguing with some director of sales around their pipeline targets. And so that's been incredibly rewarding. I suspect you might feel the same way too.

Alex Kracov: Yeah, pretty much. As soon as Lattice got big, I was like, "Okay. I'm going to go quit and start a company and go back to super small." That's funny. It's good. There's pros and cons of both sides. But I find it never gets easier. But the work changes. I enjoy doing real work myself as opposed to managing always through people. Although, I realized, hopefully, as Dock is successful, I will have to manage some people. Because that's what a CEO does. It's a funny balance, but I guess I'll always have my weird side projects that I do myself as I think about it.

Marie Gassée: I think what I'm learning here at Column and what I think William has done a good job of is, we're now 100 people, I think. Yes, still small. But there are ways to grow and to have the org expand and not get this political dynamic or these inefficient bureaucracies. Obviously, at some point, if you get to 1,000 people, it's almost impossible to fully avoid. But I think there are ways if you're really disciplined, really transparent, to actually have a company where like, okay, you can scale a good amount without being this big corporate bureaucratic nightmare.

Column's Growth Strategy

Alex Kracov: Yeah, the crazy example that comes to mind is Amazon's pizza teams where it's like, alright, you should have a team that's the size that was like two pizza boxes. They can all eat. I'm sure Amazon is very bureaucratic and corporate, but you can tell they're trying to hold on to it. They're trying, at least, which I give them credit for. Can you talk a little bit about Column's growth strategy? Is it a very small ICP, very defined, or is it wider than I think? Because in my head, it's like a very small group of people who would buy Column services. But maybe I'm wrong about that.

Marie Gassée: You know, I thought so too. And it is wrong. I joined thinking, okay, we're basically selling to fintechs that's not that large of a market or, at least, it's a large market but the ICP is fairly focused. I think what I've found is it's a blessing and a curse actually. Because we're infrastructure, financial infrastructure, our TAM is enormous. I think that's great as a shareholder and whatever. It just means it's hard to focus. We've found that the most innovative fintech companies are an ideal customer. But also, these legacy insurance companies, they have super complex payments and a lot of pain associated to that. And so I will say it's a struggle to really narrow in on the ICP, because we find pain that we can solve in so many different kind of company profiles.

I think the way we focus right now a little bit is, we are a premium solution. And so our deal sizes are fairly big. That's a little bit of how we, frankly, filter and qualify. And I think what we're also trying to do, which can be hard, is just like we're early. Keep an open mind. There's obviously a balance between we want focus, because that's going to allow us to execute really well. But we also still need to approach go-to-market with a lot of curiosity so that we don't come across an account and say, "Oh, no, they're not our ICP." But we say, okay, well, are they? Oh, wow. Okay. All of a sudden, we uncover a whole new segment of the market that could be extremely lucrative. And so, for me, that's been a little bit tough to really keep that open-mind curiosity. Because I'm tempted to do the go-to-market thing of finding the ICP, focus on it, execute on it, but rather like, okay, we're still learning. We're still kind of expanding. Hopefully, it's what gets us to be an incredible $100-billion company. But it's definitely a give and a take.

Alex Kracov: Yeah, you need to have two sides of your brain working at the same time. It's like you got to execute to hit the short-term numbers, but then obviously keep experimenting. Yeah, it totally makes sense. I assume you have some minimum or something to make sure that there's supposed to be so much noise of all the fintech startups who just raised a ton of money. Maybe you want to use it but don't want to pay. And so yeah, like a juicy minimum contract size probably weeds out a lot of the noise.

Marie Gassée: It's tough, right? Because we want to work with the most innovative newest companies, too. That's hard to focus as well. Gosh, the idea of the coolest new fintech that we might price them out, that doesn't feel right either. Anyway, we're still figuring out where our ICP is going to be.

Alex Kracov: The thing we did at Lattice was like, we had minimums. But then if we noticed that a company was a grower, like they raised a bunch of money or something like that — cruise, I think, was a good example of this where they started really small and then ballooned to 1,000 employees. If you can catch those growers early on, it's huge. Those are the best ones. They'll grow up with you, and they'll never rip you out because it's just like part of the story. Yeah, super interesting. So is it more of like enterprise sales motion?

Marie Gassée: Yeah, I would say so. I think, again, part of what drew me to Column was, there still, again, is both. This comes from, I think, William's DNA and a lot of the folks who joined the company. We still want to be the most developer-friendly banking/financial infrastructure company out there. I think our docs are amazing. The tech is amazing. We try to be extremely friendly to developers who have questions, who want to play around in Sandbox. We often will allow folks to open a bank account just for themselves or their company and play around, see what it's like to move real money with Column. So we still have these very developer-friendly roots. But the deal cycles are very enterprisey, six, seven-figure deals, more involved, very complex.

Again, I'm finding myself trying to marry those two motions. But I think maybe that's what I like most. I think the idea of being it like just doing the Oracle big enterprise sales stuff is just, I just can't get on board with that. Then trying to find that marriage of so developer-friendly, great product that the user actually values, and then the enterprise sales motion that warrants these giant deal sizes certainly feels great as well.

Alex Kracov: Yeah, I mean, I think it's the best go-to-market motion. It's the same thing I'm trying to do at Dock. That's what worked for Slack, and Twilio, and Confluent, and all of these companies, right? You have this self-serve that automates the low end of the market and as a lead gen for the rest of the business. Then you have the big enterprise sales-assist motion which closes those big deals. It works. It worked for a bunch of companies. I'm sure it'll work for Column, too.

I guess maybe one last question on Column is like, it's a super regulated industry, I imagine. It's in banking. It must be crazy to go by a bank charter. I can't believe William did that. What's it like operating in a regulated industry? How does that impact your growth strategy?

Marie Gassée: I think the polite answer is, a regulated space is a competitive moat. It's regulated. People don't really want to be in unregulated industry. It's really hard to acquire a bank. That competitive moat is incredible. And so, every day, I'm thankful for it. The less polite answer is, being regulated sucks. It was new for me, so I stepped in. I was like, oh, my gosh. We have amazing product market fit. This is a game changer. Well, you pay for it by being regulated.

And so it wouldn't allow us, I think, to be as aspirational and successful as we've been if we weren't regulated, because we wouldn't have that moat. But it just means documentation and thinking about risk and how you mitigate risk and thinking about this stakeholder that has never been part, at least, of my career which is the regulators. You have customers. You have employees, stakeholders, maybe investors which we happen not to have here, which is great. But now there's this other giant stakeholder in the room, which is the regulator. And so that's been an adjustment for me. We hired a good amount of tech folks. That's always an adjustment of like, oh, wow. Okay, this is a whole new element that we have to take into consideration.

Alex Kracov: And so it's like when you're building the product, there are specific ways you have to build the product to comply with laws. Is that kind of how it plays into the picture?

Marie Gassée: That definitely plays into the picture. That's not the hard part or, at least, it's not the part maybe that I think about every day. But every customer that we bring on, the banks take on a risk. Anyone they work with, they take on a risk that they have to mitigate document. So every time we bring on a customer, we're not just trying to sell to them, show how much value we can provide. But we also have to be thoughtful of, okay, what is the risk we're taking on with this customer? How do we mitigate that? What is the due diligence we do?

And so for me, it was like such a change in my operating model. I spent my entire life thinking about generating demand, getting people to buy. Then here, there's some of that. But there's a lot of like, okay, now I need to qualify really carefully and mitigate and make sure we set expectations and due diligence. And so it's just flipping it a little bit on its head in terms of the things we think about. I guess it's intellectually challenging, but it's different.

Alex Kracov: Yeah, I imagine it's like the know-your-customer stuff that I have to do when I create a bank account.

Marie Gassée: Exactly.

Alex Kracov: Interesting. So it's like qualifying actually is really important. It's not just ways to get good customers. It's legally important. Yeah, very interesting.

Marie Gassée: Yes, and we make it part of everything we do. We're always thinking, okay, it's not just this could be a great customer. It's like, okay, well, do they have the processes in place to stay compliant?

Alex Kracov: Super interesting. Yeah, I know. We have a few cannabis companies who use Dock. Yeah, I imagine those companies, I don't know why they would use Column. But they probably couldn't because of all those.

Marie Gassée: Not yet, but maybe one day. I like that.

Scaling With Your Career

Alex Kracov: All right. I'd love to end today's conversation maybe on a little more personal note. You've had a lot of amazing growth over your career. You went from a B-School rotational program to being a VP of growth at a couple awesome companies. How have you thought about scaling yourself along the career journey?

Marie Gassée: I think I have a couple big learnings. I think one of the biggest ones is, there's nothing better for your career than being at a company that has great product-market fit. I know that sounds kind of VC-ish. I think I'm great, sure. But I think I've had a lot of luck in terms of being at companies where people want to buy the product. It turns out that nothing replaces that. That's the best thing to optimize for, I think. So that's definitely big. I think the other learning key, maybe best practice, hiring people that are better than you. I think we learned that in B-School, and I didn't really get it. Then now I get it. I'm like, oh, yeah, only way to scale is to hire folks who are smarter, more thoughtful, better than me and who will remind me of it every day so that I hopefully can get a little bit better.

The other piece that's been a learning for me at Column in terms of maybe more tough moments or more recent big learnings is, like, you have to learn the product. I think if I'm very honest, I got away with it at Box. Box, not a super complex product. But certainly, at confluent, where I got away with knowing just enough to have productive conversations. When I joined Column in such a complex system, the financial system payments, I had a pretty tough come to Jesus of, like, oh, I can't cheat here. I can't be cute and know just enough. And so really, for the first time in my career, I had to go really deep into the product and really, really know and understand the market in a way that was, frankly, pretty humbling. And so that, for me, has been maybe a mid-career learning of like, don't cheat. Don't try to just be lazy. You certainly can't do it in these complex spaces. Then my ability to be helpful as a sales leader or a go-to-market leader now is because I did the the real work. And so I wish I internalized that earlier in my career.

Alex Kracov: Thank you so much for an awesome conversation, Marie. If people want to follow up with you, reach out, have specific questions, where should they find you?

Marie Gassée: LinkedIn. God. I wish I could say Twitter, but no. Yeah, just LinkedIn.

Alex Kracov: Cool. Awesome. Thank you.

Marie Gassée: Thanks.

Never miss an episode.
Subscribe now.

Thanks for subscribing! We'll email you when we publish an episode.
Oops! Something went wrong while submitting the form.