Crossbeam is a data-driven partner ecosystem platform often described as 'LinkedIn for data.' It helps companies find overlapping opportunities with their partners by enabling them to share and analyze data collaboratively. The platform replaces the manual process of account mapping, which traditionally involves emailing spreadsheets back and forth, with a secure, automated solution.
Bob Moore validated the idea for Crossbeam by pitching it to 20-30 founders rather than traditional end users or buyers. He sought their gut reactions and asked if they would start the company themselves. This approach allowed him to expand the scope of the idea and gauge long-term conviction rather than just immediate product-market fit.
Bob Moore chose to talk to founders because he trusted their ability to understand market evolution and the durability of a business idea. Founders have empathy for multiple personas and can provide insights into long-term sustainability, which is crucial at the early stages of a startup.
Bob Moore learned that product-market fit is not static; markets can move, and products can fall out of fit. RJMetrics initially struggled but eventually found fit when the market caught up to their analytics platform. However, they lost fit again when the market shifted toward cloud-based data warehouses like Amazon Redshift, making their bundled product less relevant.
RJMetrics grew from less than $100K in its first year to $1M in ARR over three years. During its peak, it grew from $1M to $5M in ARR, but growth slowed to 10% as the market shifted toward the modern data stack, leading to churn and flatlining revenue.
RJMetrics pivoted by isolating its data pipeline technology and creating a new product called RJMetrics Pipeline, which allowed data to be sent to any destination, not just their own servers. This pivot led to the creation of Stitch Data, which was later acquired by Talend.
The 'joint jam' is a sales tactic where Crossbeam conducts demos with two companies simultaneously. This approach ensures that both companies see the value of the platform immediately by connecting their data live during the demo, bypassing the single-player mode issue and driving virality.
Crossbeam merged with Reveal because customers were frustrated by having to use two separate platforms to connect with their partners. The merger unified the network graph, eliminating the need for customers to choose between the two and allowing both companies to focus on competing with larger players in the market.
Bob Moore was surprised by the strong work ethic and talent of the Reveal team, particularly given stereotypes about French work culture. He found the Paris startup scene vibrant and plans to invest further in growing Crossbeam's Paris office.
Intellectual honesty is crucial for recognizing when market dynamics are changing and when a product is falling out of fit. Bob Moore emphasizes the need to differentiate between noise and signal, especially when metrics like churn or outbound sales start to decline, as these can indicate deeper market shifts.
- A lot of founders, they talk about product market fit as though markets are stationary and your product is the thing that you move. It's almost like the market is sitting there like a pole in the ground, like this fixed thing, and you're kind of shooting arrows at it until one of the arrows hits the pole. And I think the thing that we learned at RJ Metrics was when we entered product market fit that like our arrows were just hitting, hitting, hitting, and we said, "Holy crap, this is amazing." What we did not realize was that the telephone pole keeps moving.
Welcome to In Depth, a show that surfaces tactical advice founders and startup leaders need to grow their teams, companies, and themselves. I'm Brett Berson, a partner at First Round, and we're a venture capital firm that helps startups like Notion, Roblox, Uber, and Square tackle company building firsts. On the In Depth podcast, we share weekly conversations with startup leaders that skip the talking points and go deeper
into not just what to do, but how to do it. Learn more and subscribe today at firstbrand.com.
For today's episode of In Depth, I'm really excited to be joined by Bob Moore, who's the co-founder and CEO of Crossbeam, a data-driven partner ecosystem platform. Before Crossbeam, Bob built two startups in some pretty different markets. He launched his first company in an analytics platform called RJ Metrics the day before Lehman Brothers collapsed in 2008. He then spun out a piece of the product into his next startup, Stitch, in 2016.
Despite battling headwinds, he managed to see both through to modest acquisitions. But for his third venture, he tried to take a bigger swing and move with the market instead of against it. In our conversation today, we dive into the parallel histories of the SaaS industry and Bob's own journey as a repeat founder. We start with how he validated the idea for Crossbeam by avoiding the traditional customer-driven approach and instead asking founders for their gut reactions.
From there, we rewind the clock for the lessons he learned at his first two startups, including the cautionary tale of how RJ Metrics found and then lost product market fit. He breaks down the warning signals he ignored and stresses why founders need to have intellectual honesty.
as they try to see around corners. We then bring things full circle to talk about Bob's current run at Frostbeam, where he shares how he's tackled everything from prototyping to pricing. Finally, he talks about Frostbeam's unconventional move to merge with a competitor and how he put his ego aside to make the right call for the business. Bob is an incredibly reflective founder, and he has a unique ability to get to the bottom of why things played out the way they did. He can zoom out and analyze the macro trends and market forces
just as easily as he can narrow in to pick apart his own decision-making as a founder. I really hope you enjoy the conversation. All right, well, let's get into it. Thanks so much for joining. Yeah, so happy to be here. It's a first time, long time. I've heard almost all of these. Well, given we're at the beginning of the conversation, maybe we start at the beginning of Crossbeam's journey. It's funny to talk about the beginning of Crossbeam being the beginning because in a lot of ways, the beginning of Crossbeam was the end. I...
I'm a three-time SaaS founder, and I started two companies before this one. One of them was called RJ Metrics. I started way back in 2008 when SaaS wasn't even SaaS yet, and that was an analytics platform for e-commerce companies. My second company was called Stitch Data. It was even nerdier than RJ Metrics. It was a data infrastructure company that kind of built... It was an early player in the modern data stack, building cloud-based data pipelines to help people get data into their data warehouses.
And these are both data companies. And we sold RJ Metrics to Magento in 2016. And then it got rolled into Adobe in that acquisition. We sold Stitch to Talent in 2018.
And there was this period that probably was about six months long between when things wrapped up with Stitch Experience and the Magento Adobe acquisition. And there was kind of this like calm before the storm. And I had over the course of the decade before building those last two companies, kept this old Evernote file. I guess it would be a Notion file these days, but it was an Evernote file. Every single business idea was
that I wished I could work on that I wasn't working on because I was doing RJ or Stitch. And you would just randomly every day, if something popped to mind, you would just scribble it down? Seriously, like a little thing, whether it was like a Skunk Works project we did and I said, holy crap, you could build a whole company out of this thing or like a problem we encountered that it seemed like there was no good solution for because
conversation I had over coffee somewhere where some kind of idea came up. There was stuff in there that was way outside the realm of B2B SaaS, right? There were consumer business in there. There was like, at one point I was like, maybe I should just open a chain of escape rooms, right? Like there was some stuff that was like way, way out of the ballpark there. But come 2018, like mid, you know, early, mid 2018, I cracked that thing open.
And frankly, this is where the repeat founder thing is I probably should have gone to therapy, but instead I cracked that Evernote thing open because I tried to sit on a beach for a couple of weeks and just lost my mind. I went through my own personal scan and I said, I'm gonna eliminate anything that doesn't sound like it
it would be extremely fun to work on. Literally, I think this is a founder market fit pre-screen, almost like this two by two matrix of how much fun would I have doing this? How much will it light up my brain? How intellectually interested am I in the problem is one dimension. And then the other dimension is what is it about my specific experience that makes me predisposed to being good at solving this problem?
I had varying levels of where I fell in the two by two matrix for RJ Metrics and Stitch. And I think it was part of why those businesses had, you know, base hit style outcomes. And neither one of them really got to like IPO scale or had the trajectory that hopefully Crossbeam is on now. That took that list of 100 down to
15, 20, maybe. Maybe it was even less than 10. And then I just started spending time with people that I know that were running other companies, like founders that were running companies at all different stages. And I just started having a meeting, not biasing them in any way and saying, hey, I've got three business ideas. I like them all equally. I want to kind of pitch you on them and see where it goes. And I probably had 20 or 30 of these calls and I distributed the 10 or so ideas across all these. And my goal was
It was probably to test the thesis as to whether or not I could really talk about and express these values in a big way, and also to test the thesis that I have fun talking about them. Crossbeam just stood out far and above. Why did you choose specifically to talk to founders versus sales leaders, product leaders? Pick any other potential end user or buyer. This may be a bias that doesn't actually prove true in the real world, but it's a bias that I have, which is that I trust them more.
Certainly, it's one thing to talk to your end buyer or user persona to try and identify like a strength of product market fit for a particular product idea. But to me, at this extremely baseline atomic level of a startup, while product market fit is really the whole ballgame as to whether it goes anywhere, there's more to life than just product market fit.
the level to which I can build and rally a team and a level of personal conviction and a belief in long-term sustainable product market fit due to the legitimacy of the problem we're solving on a bigger scale, on a bigger context. Like if you're at zero, stage zero or stage negative one, and you're just a heat-seeking missile for product market fit without regard to those other things, the problem is product market fit is extremely ephemeral.
We experienced this at RJ and at Stitch in good ways and bad. You can build a product
and the market can move and you fall out of product market fit. You can build a product and the product can move, right? And you saturate your market. You get to a couple million in ARR and realize that's where the ceiling was. And you try to build that next feature on top of it to 10X again. And you've completely pushed yourself out of product market fit by pivoting your product. This happens so often, right? The companies that make it to the A and not to the B or the C. There are so many ways in which product market fit does not intrinsically persist once you find it.
So I think had I just spoken to executives or folks in the initial target persona that I thought would have worked for these ideas, I would have ended up kind of trapped in a prison of my own making of needing to build to and sell for those folks because it was the only feedback I'm getting. Founders are special. Founders, by definition, need to develop an extremely high level of empathy and understanding for the needs of people across multiple personas and also understand a
baseline grasp on how markets evolve over time and what makes for something that's more durable and versatile. And I think being able to speak to founders, not about would you buy this, but more about would you start this company? And what are the things that you think you might bump into? I think
Rather than taking these ideas and narrowing them down into the scope of how a persona would use it or what use case would work, I was able to take these core ideas and actually expand the scope of them to broaden my horizons of what this could become. Because I wasn't interested in building my next thing and having it be anything other than like an IPO scale business. I was not optimizing for another base hit or quick win or
you know, build it for a few years and sell it situation. I thought I had one more in me and I wanted to see how far I could take it. And I think that required this exercise of zooming out before I zoom in. And I think founders are a great profile of folks to talk to for that. Do you think that bringing them multiple ideas kept you from falling into the trap that happens over and over again, which is like a multi-time founder is thinking about building this widget and they talk to other smart people and the smart people...
either want to be nice or think the person is smart. So there's like this proxy dynamic of their idea might be smart. Or if there's an investor context, the number one thing that happens is there's this optionality dynamic where generally speaking, it's an investor's best interest to tell you how great you are
And it oftentimes leads founders in all sorts of non-optimal directions only because the investor is trying to preserve optionality. So one of the things that I thought was really interesting is that you took three different ideas as opposed to one. You are hitting the nail on the head so, so well. There is this absolute...
It's like, I suffer from this bias. If a friend of mine who started a company and sold it then comes to me and says, "Hey, I've got this new thing I'm thinking about working on." And then explains to me with a lot of conviction why they're so passionate about it. The thing that I perceive typically is that they're already sold on this idea and I'm not gonna talk them out of it. And the question is, even if I don't love the idea,
do I know this person well enough that I actually might want to write my angel check in and make a bet on them? Not because I think the idea is going to work, but because I think this person is not going to allow themselves to fail. And even if the idea, you know, ends up not being great, they may very well, you know, pivot the gaming company into the next slap, right? Or pivot the radio company into the next Twitter or whatever it is that's in that kind of, you know, chain of
great founders just not dying. And in that case, my angel check would be put to good use. So like the FOMO of not being in on their next thing is greater than my willingness to create confrontation with them, which is not a hard ratio to keep over one, then I'm not going to give them as candid feedback as I probably rationally should, or I'm not even going to allow myself into that headspace, right? Like that's the other thing with founders, right? They are default optimistic. Founders immediately think of every idea as like, oh, what if I was running that
would I be good at it? And that sounds like fun. Therefore, it's a good idea. So give it forcing them to compare and contrast and say, like, which one? Why this one? Why not? That one does create a little bit of like a forcing function of, you know, being able to criticize all these ideas. What were the other two ideas that you brought?
Oh, that's a great question. I had one that was all about this thing I had done at Stitch. Actually, all three of these came from Stitch. So Crossbeam, we'll talk more about Crossbeam, but Crossbeam originated with some pain points that we experienced most kind of concretely at Stitch. There was another one where I did this analysis at Stitch at one point where it was basically...
retroactive A-B testing. So I went into the enormous vault of historical data. Stitch was a PLG product. So people could sign up for free and then you plug a bunch of systems in. And then at some point you hit some data limit and you have to pay. So we just had all this data on who signs up,
what they do in the product and who converts and who sticks around and doesn't turn, right? So like what's the customer lifetime value? What are the cohort analysis? What's the strength and lifetime value of a company and a customer? And then what is the absolutely enormous fire hose of things that we capture from a combination of having snowplow in the product that captures every action and click and all the marketing stuff that happens up funnel and whatnot.
In a world where you want to say run an A/B test, what you have to do is go in and say, "Okay, I'm going to do marketing copy A for some and marketing copy B for others." Then you kind of figure out what copy is optimal. Well, we were able to find a bunch of things that we did have control over where there was built-in variability just in the population of people for some other reason. What source they came in from, what piece of content they read originally.
what partners they were already working with or that we knew they were customers of for various reasons, which would later become like a cross beamy idea. What the first actions they took were when they got into the product, what pages they went to first, what data source they connected first. We were able to actually go in and just basically run the analytics as though it had been an A/B test of sorts and say with a certain degree of confidence that better customers
are ones that go down this particular user journey or that connect to data source like this one first or that experience this type of content first. And then we were able to basically choose the way that we constructed the user journey and the way that we attracted and drove people through the content and discovery funnel based on our knowledge of where the best customers ended up kind of like routing through.
So it's not really A/B testing, it's just using data analytics to customize the way that you create user journeys. But it was this really interesting framing where we ran all the same analytics and presented it in a board deck that could have become a product in a way that felt as though you're running an A/B test. And it's like, oh, it turns out people that use MySQL
are 30% more likely to become paying customers than people that use Postgres. It's like, well, what's going on there? And does it say something about us trying to recruit more customers from a particular ecosystem? So anyway, didn't build that. There's probably 50 AI companies now that do a better version of that, but that was one idea. The other one was about microsites and using content for content marketing to create lead generation at scale for companies. And this is kind of an SEO play
We did this really cool thing at Stitch where what Stitch did as a product is it would move data like from one of your SaaS tools to your data warehouse. The data warehouses were Amazon Redshift, Google BigQuery, Snowflake, Azure SQL Data Warehouse. We bought all these domains that were like to Redshift, to Snowflake.
to BigQuery, right? Like T-O and then the name of the brand. And we built this content management system where every single data source, we pull from like 70 data sources. So, you know, Trello and HubSpot and Salesforce and whatever, GitHub, wherever all your data is, these actual operational applications, SaaS tools you use, we'd pull the data out of them
and created a content subdomain for every single one of those data sources. So if you went to, for example, MailChimp.toredshift.com, you would land on this page that was actually dynamically constructed at the moment that you accessed it. And the content of the page was basically, here's what MailChimp is, and here's their APIs and how to get the data out. Here's what Redshift is, and here's all the ways the data gets placed into there. You can basically build your own data pipeline by writing this kind of script to this API and then deposit it in. Or...
click here and sign up for Stitch and in a couple of clicks, the data will flow. It sounds very Zapier-like. Yeah, it's extreme. So Zapier has got a really, really good strategy that's very similar to this. The way that their product is operationalized is a little bit different, but they've done really, really well on an SEO strategy like this. But the cool thing is you only have to write A plus B number of articles or pieces of content to stand that up, but you get A times B number of
unique landing pages or microsites as a result of doing it. If we had 10 destinations and 70 sources, then we wrote 80 pieces of content and we got 700 pages out of it.
that became an enormous, enormous lead generation engine. And there was a business idea, getting back to your question about where did this land in that Evernote file, there was a business idea around trying to productize and operationalize that platform for finding like order of N squared or maybe that's just two N, right?
pieces of yield from generating a smaller number of pieces of content by cross-multiplying content against itself to generate more unique content. Again, now in the era of Gen AI, this all seems ridiculous, right? Because you can just like hit a button and generate...
and your infinite amount of content that probably is better than what our dynamic generation was. But we did a really similar play, by the way, at Crossbeam called Partner Base, which is like a giant database of all the partnerships between companies that exist in the world. That's an order of N squared play where like, we basically scraped the partner page of every company on the planet and draw a giant network graph of all those partnerships. And then
you get a landing page for every company that details their partnership, but also every combination of two companies that might have a partner or if they do or don't, you can show how their ecosystems kind of like overlapped or intersect or how many degrees of separation away they are from each other. Such that like if you Google Microsoft Atlassian partnership or something like that,
we are a landing page for that, that depending on the day might actually outrank the partner portals of either of those companies. I do random like consulting calls all the time for like fellow founders in a portfolio or whatever about strategies like these. And I can always find at least a way to do like an order of 2N outriggers
output SEO style like growth hack. These things require a lot of time and patience from accumulating domain history and like credibility. But when they work and you maintain them while they persist extremely well. But anyway, that was another one I was pitching. There was literally an escape room one that involved using like Oculus Rift headsets to like create, uh,
a near infinite number of escape room experiences, which someone also did. Yeah, which there's like the best of the best VR games are effectively just glorified escape rooms with cool content and stories wrapped around them. But yeah, a lot of these things did come to life. No regrets on the cross-stream path. I think it was the right one for me.
So share a little bit more. You get on a Zoom or you meet one of the 40 founders you were sort of hashing this out with. And you said, I have three different directions and you pitch them. What did you actually do? It sounds like the sort of general concept for Crossbeam bubbled significantly to the top. What did that actually sound like? Yeah, it sounded like there was always a next step with a lot of these, you know, the escape room one. People loved it. And they're like, oh, man, that'd be so cool. I'd probably have a lot of fun playing that.
Oh, well, see you later. With the crossbeam one, it was like, you know who you should really talk to about this? Because I know they've run into this problem is so-and-so. Or do a field or something where you can put me on your mailing list so I get updates on like when you've actually started building this thing. I felt like coming out of the conversations where I talked about crossbeam, I was actually...
cultivating a wait list or like some sense of pent up demand. And this is why Crossbeam is so cool as a business. The business itself is intrinsically viral. We're like LinkedIn for data. Like you can't use Crossbeam unless all your partners are also on it, or at least your most important ones are. So you are intrinsically motivated to...
invite your partners on and talk them into joining, even independent of us doing anything at Crossbeam. And I started seeing that virality kick in before the product even existed and we could build viral loops into the product. It just happened by a telephone chain. I would get text messages like, you know, the partner manager at our biggest partner was interested in talking to you about this idea because they desperately need some kind of solution like this. Anything that was like,
Product market fit just punches you in the face when you find it, right? And I think it was just very, very clearly there from the idea stage with Crossbeam. Before we sort of pick up what happened after those conversations, I wanted to go back to sort of your comment about sort of getting into and then out of product market fit, at least in the context of Stitch and RJ Metrics. Would love to dwell on that a little bit more and have you talk more about what you sort of meant by that and in a detailed way, what you saw maybe in both of those two journeys.
RJ Metrics is the best case study for this one because I think we actually, we started the business not in product market fit. And then we found a sustained multi-year period when we were very clearly in it. And then we slipped out of it again. And I think the big takeaway that I had from that, which is kind of like a first principle, maybe going into this story is that a lot of founders, they talk about product market fit,
as though markets are stationary and your product is the thing that you move. Classic means startup, build, measure, learn cycle, right? You have an idea, you take it out to market, something's not quite right about it, it's not quite clicking, what do you do? You pivot or you iterate on the idea or you follow the feedback and try to build a next version that's going to click better. And you do that basically until you
run out of money or patience or you throw in the towel or it actually clicks and something works and you have a real business on your hands. But that whole construct, it's almost like the market is sitting there like a pole in the ground, like this fixed thing. And you're kind of shooting arrows at it until one of the arrows hits the pole. And I think the thing that we learned at RJ Metrics was from the very first day there, my co-founder Jake and I worked at an adventure capital firm called Insight Partners from 2006 to 2008. And
We were big kind of technology and startup geeks. And this RJ Metrics idea came along and we decided to quit our jobs and go after this idea. And we quit our jobs on a Friday in September 2008. And on Saturday, Lehman Brothers collapsed. So we were dropped into a...
a market basically at the cusp of the great recession happening. And as much as we had all of our venture capital contacts and relationships and things from our couple of years working at Insight, they proved to be completely worthless. We were two unproven founders in a market where LPs were saying, make a capital call and it'll be your head, right? There was no capital flowing. So Jake and I had this really interesting extended period of about three years where we bootstrapped that business entirely.
And the business grew really slow. And I don't think we really actually had product market fit that entire time. I think what happened was we had hustle and patience. And what we did was build what ultimately was a services business with software wrapped around it for the first year, year and a half, just to generate enough revenue
to pay for our own health insurance. And we moved the business out of New York. We went down to Camden, New Jersey, which maybe has the lowest per square foot business real estate you can imagine, which is kind of part of my journey to Philadelphia. You know, we grew that business to a million bucks in ARR profitably, got our first dozen or so employees. And what was the early product, even if it was wrapped in services? We would connect to your backend database, which was usually your shopping cart.
which basically had all the tables about who your customers are and what they bought. And then we would suck that into a hosted MySQL database that was our early version of a data warehouse that was kind of indexed and optimized for analytics. And then we would build dashboards on top of that inside of our UI that would show you things like cohort analytics, customer lifetime value analytics, marketing ROI analytics. We were building software to automatically generate
SQL queries that adhere to your business logic based on the way your data was structured. We sold mostly to marketers in e-commerce companies who didn't have data teams. So the marketers could just log in and say, okay, here's not just what my sales were yesterday, which they could get from their shopping cart, but how many of those sales were from repeat customers? What did those repeat customers buy last time versus this time? Was it the same thing? Analytics like based on what you buy the first time, does it make you more or less likely to come back? What happens to
order size across multiple repeat orders. A lot of things just around like trying to drive repeat business inside of e-commerce, which is where all the margin is because new customer acquisition is so expensive. Maybe a better analog for what RJ Metrics was. RJ Metrics was Fivetran, Snowflake, DBT, and Looker.
all together in one product. We built all four of those products in order to build RJ Metrics. I should say we built the crappiest version you can imagine of all four of those products in order to bring RJ Metrics together. And it's no surprise, by the way, that you look at the diaspora of people from RJ Metrics and the original, you look at the 2013 RJ Metrics squad,
And it is the founders of DBT, the three co-founders of DBT all worked at RJ Metrics. The founders of Stitch, myself and Jake, which became the early competitor to Fivetran and played in that space. Omni, which is a hot new ticket in the BI space, their founding CTO, Chris Merrick, was our CTO at RJ Metrics.
Pretty good. Three out of four. I don't think we, I don't think we started any data warehouse companies, but three out of the four big players in the modern data stack of the deconstructed stack all started at RJ where we built the foundational versions of all these to stand up, stand apart J metrics. So anyway, we were figuring all that out, like very, very live. And we did not pivot that much during all that time. We just kind of like built the core of the business in this bootstrap way. And all of a sudden it just started working.
In a single year in 2011, it's like we landed all the cool companies of the day, which is funny because they're not the cool companies of today, right? But it's like, you know, fab.com, which is maybe one of the fastest growing e-commerce companies that ever existed. All the Groupon clones. This was in like the Groupon wave, right?
Subscription commerce just kind of like came into being overnight and everybody wanted to have subscription commerce as part of how they do business. And we were the best analytics vendor for subscription commerce. We had like Warby Parker, we had Paperless Post, we had, you name the hot internet brand that existed. Oh, a lot of the direct consumer stuff, right? Like Casper and Bonobos. We had them all. They were all our customers and we...
overnight, right? We just got into this extreme hyper growth mode. And also all the venture capital firms came knocking, right? And we were able to raise a seed round in 2011 and a series A in 2012 and a series B in 2013. It just was like almost too good to be true because we had built the bones of this thing. And then the market demand just like explosively showed up. What was the root cause of that? We kept shooting the same arrow in the same direction. We didn't realize we weren't hitting the telephone pole.
It took us three and a half years to get to a million dollars with ARR. Then the telephone pole moved. And actually, we fell into product market fit, not because of something we changed about our product, but because we were early to the market. And we didn't die because we had figured out a way to kind of like operate the company profitably and not run out of capital. And we were 23 or something. And we had all the time in the world on our hands, right? And felt like we were going to see this thing change.
through. Persistence or dumb loyalty or whatever else it was. We kept the lights on. I can make myself look like an idiot or a genius depending on how I pitch this story. And I try to go the idiot route sometimes, but here's the genius version. We saw it coming. We were at a venture firm that was investing in the most successful businesses that existed on the planet that had gotten to venture scale in this universe of e-commerce. And we saw these really, really early indications
the ways in which data was being used to actually fundamentally change the way these businesses were being run. One of the businesses I invested in at Insight was Fanatics. At the time, it was called Football Fanatics, and it was three guys in a strip mall in Jacksonville, Florida that had an e-commerce website that had pretty impressively good growth stats. And we invested in that company-
when it was minuscule in size and it's become this incredible industry-owning multi-billion dollar powerhouse at this point. But I was sitting on the front lines watching what 15 years later would become one of the largest, possibly the largest sports brand and merchandise company to ever exist and seeing what they were doing. And the reality was what I wanted to do was bring the way in which investors and the smartest entrepreneurs were thinking about data to the masses.
That is what we built and that is what we implemented. And the market was not ready for it when we came to market with it because they were still figuring out how to know how much they sold yesterday. And it took this two, three, four years of us saying, no, this is the way, this is the North Star, this is the way for the market to actually reach a state where it was ready to consume this in mass. We didn't cross the chasm, the chasm crossed us, right? Like we kind of stayed in the same place. The timeline shifted on the market and we suddenly became the right thing for the right market at that right moment.
And I think it's a little bit of both are true. There are probably, and I'll tell you why the idiot version of the story is the persistent timeline too, because here's the thing we didn't realize. I think we did realize when we entered product market fit that the telephone pole had moved and that like our arrows were just hitting, hitting, hitting. And we said, holy crap, this is amazing. What we did not realize was that the telephone pole keeps moving. And this is the other side of that product market fit coin, which is we were early to the market and then we crushed it for like three years.
And then what happens? By 2014 or 2015, Amazon Redshift becomes the fastest growing product in AWS history. The cloud-based data warehouse revolution hits. And what it does is it drives this incredible wedge right in the middle of the RG metric stack. And companies at the ops and engineering level start buying Amazon Redshift as a data warehouse, as like a data store for anything they want to do in analytics cross-functionally across their entire business. All the data starts flowing into there. And then
When people evaluate RJ Metrics, they say, why do I need to buy another redundant data warehouse that's embedded inside of this full stack product when I already have all the data in this warehouse over here? Why don't I just buy a BI product and plop it on top? And what were those BI products? Tableau, Looker. We at RJ Metrics fell out of product market fit because the market kept moving.
And we had that initial three years where we kind of like chugged our way along. What was the revenue ramp in those three years? Probably in that first year, year zero, I bet we were at less than 100K after a full year. It was probably 100K to 400K to a million, something like that. And then like a million to two and a half to...
five during the peak zone and then another decent growth year in there. And then it was like 100% growth, 100% growth, 50% growth, 10% growth. And still south of 10 million of ARR by the time that 10% hit. So we were on this very nicely shaped exponential looking curve with the luxury of that first three years being a real slow grind. And then
in the wake of the modern data stack revolution, more or less a complete flatlining. And a lot of that driven by churn. We were closing a decent amount of new business, but renewals were dying because all of these companies that we sold to, they continued their exponential rise and they needed something bigger and more robust. And the
Modern data stack was actually the answer. So what was the story that you told yourself at the time, which is, let's continue executing in this direction versus, well, something has changed in the market, and we need to go left instead of right?
straight. And this actually gets into where Stitch came from and where our Magento transaction came from, because we had, it probably took us, call it six months to a year of like really getting beat up and like scratching our heads and saying, why isn't the magic money machine printing money anymore to like really, really look introspectively and deeply understand that was going on. We did go through that exercise and I think we did get a very good handle on it. And we appreciated the magnitude of
what was emerging, which was like that old Jim Barksdale quote, like there's only two ways to make money in software bundling and unbundling. Right. And like a great a grand unbundling was happening and we were a bundled product, like a sweet solution. We all kind of came to the conclusion it's like we need to participate in
this modern unbundling economy, and we've built some really good technology here, what do we do about it? There were really three options on the table. One option was to blindly go forward and raise even more money and raise a Series C and try to just make our journey metrics work despite the fact that we think these market we fell out of product market fit these market forces were pushing against us now could we have raised that capital? Absolutely not.
Like, I'm sure that it would have been a real uphill battle. It would have probably been very ugly terms. So that was option one, but it didn't seem realistic. Option two was to hard pivot the business and take a piece of what we had built at RJ and what we felt like was maybe the best piece, the newest one, the best technology, the one that would fit the best into this market and just pivot into that really hard.
But that would also be a bit of a bloodbath because we would probably have to lay off 70% or 80% of the staff. We'd have to recalibrate. Our cap table was one that had raised a bunch of capital. So it's like, we're basically going back into startup mode, but with a cap table that's weighed down with a bunch of liquidation preference and all kinds of other stuff without being able to create any intermediate returns for anybody. And it was asking our team to sign up for a lot.
So a hard pivot, but that was on the table and that was like probably intellectually the right answer. It was just a very hard pill to swallow. And then option three was to just take our toys and go home, which is to try to sell the business. Like we had built a business that was at a decent revenue scale, especially for that time.
that had some really good marquee customers still, and that was still growing. It was a growing asset. It just wasn't a hyper growth asset. And that could potentially be very strategic to a lot of companies out there that we sold alongside or that we complimented. So we had this discussion as a board, and we decided to go collect data points on all three of these paths. So we did three things simultaneously in the last six months of 2015 into the first couple of months of 2016. One of them was
we started building and implementing a new product that was basically a completely carved out and isolated fork of just one piece of RJ Metrics, which was the data pipeline piece. Just the thing that moved data out of all these 70 different SaaS tools and historically just dropped it into RJ.
We made it so they could drop it into your data warehouse, drop it into whatever destination you want instead of us just pointing it into our servers. So that was phase one. And we called that product RJ Metrics Pipeline. And we started bringing it to market, doing the product marketing for it, getting initial customers on it, et cetera. That was like the setup for the pivot if we needed to go that route. B, Jake and I went around and pitched every investor you can imagine in Silicon Valley on raising a Series C and got laughed out of just about every single room we were in, right? Like we were... We tried...
in earnest to raise incremental capital. And it was just very clearly not gonna happen because the business had gone from 100% to 10% growth and it had a churn problem. The other thing we did was we had been cultivating over multiple years, relationships with senior executives at a lot of the companies that we were closely partnered with who are much larger than us. And there were a bunch that I won't name that we got into deeper talks with, but Magento,
was one that had just recently spun out of eBay. It was an independent private equity backed company. And it was being very acquisitive at the time because it had dry powder to spend and it was kind of setting itself up for a large strategic acquisition. Ultimately that would be by Adobe. Adobe bought them for 1.6 billion a couple of years later. But they were in this moment
where they were acquisitive and they kind of had their antenna up. And it was kismet because we at RJ went to them and they were very interested very rapidly. We were able to kind of engage in a couple months long negotiation and discussion about having Magento acquire RJ. And
The discussion actually fell apart the first time around due to differences in valuation. We were close, but it was like too much of a gap. And it just felt like for us to walk away for what the consideration was, it wouldn't have even returned like investors capital. So we both kind of walked away and, you know, another month or so went by and I don't
I don't even remember who reached out first, but I think we both kind of had some FOMO on having not done it. The conversation reignited and both sides were a little bit more reasonable. They were a little more reasonable on valuation and we were a little reasonable on valuation as well with a caveat. And that caveat was we will do this deal. But in addition to the consideration for the purchase, we want to carve out our geometrics pipeline and keep it.
as a piece of IP. That piece of the RJ stack that you need to operate RJ, you've got it. We'll fork the code, freeze it in time. You've got a perpetual forever license to use it, modify it, whatever you need to do to operate the platform. But
We're going to keep the IP that is RJ Metrics pipeline. And that was the deal that we got done. So we had this exit of RJ. We got acquired by Magento, but we retained pipeline. And that is what became Stitch. So Stitch was a spin out of RJ. That was just the data pipeline layer from RJ Metrics repositioned and rebranded and
that 70 or 80% of the company we would have had to lay off in order to do the pivot, they got acquired by Magento and went to work for Magento. So we didn't have to cut any teams or any people. We got an economic outcome for our investors and some of our shareholders. And
We had everybody around the table collectively committed to Stitch as the next journey. So we actually took the majority of that outcome from Magento and use it to bankroll Stitch. So Stitch never raised any additional outside capital. We basically like landed the plane of.
creating a really good home for RJ, bankrolling Stitch, creating a little bit of outcome for folks that had been involved in the RJ journey along the way. And then we got to operate this new pivoted company and Stitch in two years, in 19 months actually, got up to almost the same revenue that RJ Metrics had in eight years. And that's when Talon came along with an acquisition offer that had a really big premium attached to it for how long we'd been running that business and created an outcome for everybody that made the whole 10 years worthwhile.
You mentioned you were kind of dual tracking these. You're basically gathering data on the three potential directions to go. And part of it was the pipeline product and doing product marketing and bringing it to market. In that process, did you get really strong customer resonance? That's a great point. That's a really great point. Yes, we did. And in fact, what was interesting is the form that it took because we certainly started...
selling the product directly, like as new business and kind of finding these traction points where we were having a lot of conversations in the sales pipeline that were getting closed out as close loss because people ended up buying Looker or ended up buying some other thing. And what we were able to do is basically pivot all those conversations, even ones that we had lost in the past into sales opportunities for the Argiometrics pipeline product. Because if you are using Looker sitting on top of Redshift at the time and you were like, oh crap, I really need to get my MailChimp data in here
There was no good way to do that in an automated kind of cloud-based fashion without your engineering team writing a bunch of scripts to pull the data out of the APIs and drop it into Redshift. We were that middleware glue. So we actually made the Looker installs more valuable. We made the Redshift usage go up and the storage and compute that happened there go up by just kind of making the amount of data that was available more robust. You know, Looker, who used to be our biggest competitor over at RJ, started becoming our biggest referrer of business at Stitch because the Looker sales reps knew
If they were going to close a deal, their client wanted to see all the dashboards they cared about in the demo. And that meant they needed all the data. And that's not what Looker does. Looker needed a partner to get the data into the warehouse. And we became that de facto partner. So really, really quickly in that time period, it was actually... And the churns and the never-closed lost pipeline from RJ literally became the pipeline for Stitch. It was like a perfect...
inverted version of the growth curve from RJ. Everything we had ever lost became the opportunity because the reason we lost it was because we lost this difference in philosophy that now we had adapted to. If you can't beat them, join them, right? And we joined the revolution. Now that you're running Crossbeam or that you have lots of talented friends who are building companies, they have got into product market fit. They're at seven or eight million in ARR. What is the situational awareness that they need?
And how do you manage against the inverse problem, which is there are often things that look like it could get in the way of your product market fit and it never materializes. What's sort of the meta actionable insight from that whole experience of the market moving away from you and going from really strong product market fit to sort of being out of product market fit? If I had to distill it down to something that I...
wish I had more of back then and I worked on a lot. And I think all founders need to really, really care deeply about. I think it's as simple as the idea of intellectual honesty, a willingness to understand the difference between noise and signal in the things that are not working in your business. And for us at RJ, it was churn.
which like we always had a problem with the churn at RJ. We never had great churn numbers. Part of it was because we had a fairly low entry level price point and there was just a lot of noise and like companies going out of business and getting acquired. And like we used to call it structural churn. And we existed for a long time with this structural churn problem. And then
In the modern data stack era, call it early 2015, when we started to suffer from growth slowing down, all these things just started, they seem like they started to break at once. And one of them was like churn got a little bit worse. And one of them was that our outbound SDR program, which used to be like this huge ROI positive thing, completely stopped working. And the economics of it just kind of went upside down. And I think we were so analytical and so data-driven that I think we tried
too hard to like contort to the data and the narrative to something that made us feel better and sleep better at night. Maybe churn was going up because there was some kind of bigger structural churn component, right? And maybe the outbound stuff wasn't working because we hired two great outbound reps as the first two. And then we like kind of got to sleep at the switch and just hired a bunch of folks that weren't good at their jobs. And like, maybe it was a people problem. Maybe it was some kind of a macro problem.
when in reality, the market was really changing under our feet and the data was there and we didn't want to hear it. And I'll give you two examples of that. If you feel like the people
the people around you are getting dumber and dumber. Like maybe you're the one getting dumber. Like the great example would be we had a bunch of customers that churned because they decided to like go to Looker or because they decided that their engineering team was going to take over the analytics stack and the marketing team was kind of losing control of that budget. Our consistent reaction to that was very often like, "Oh, these guys are dumb."
It's like, oh, they don't know how to run their business. They don't know how this analytics stuff works. Like they're going to be back in a year. You're kidding me. You're going to have your engineers telling your marketers how to run the company. Like these guys are making like a fundamentally flawed decision about how to run their company. It was one thing that for the sake of pattern recognition, it was like, it happened enough times that at some point in time, it clicked that it's like, oh wait, I think we're the dumb ones. And this was a pattern that we could have picked up on earlier. The other example, which occurred to me years later, was when we announced our series B round, Amazon had just launched Redshift as a
as a data warehouse product. It couldn't be farther in my mind from something that was competitive. It was a infrastructure product sold to engineering teams to store data. And we were an analytics product that sold to marketing teams that was full stack and just so happened to have like a MySQL database in the middle.
And the TechCrunch reporter asked me this question and like would not give up on this thread, no matter how dismissive I was about it, which is like, aren't you scared of Redshift? Don't you think Redshift is going to like be a problem for y'all? And I was so adamant. This was two completely different worlds. And I remember talking to Jake. I was like, I feel like this reporter is like maybe on the wrong beat or something. They're trying to make this story a story about Amazon because it'll generate more clicks if you're talking about Amazon instead of talking about some startup.
I don't think they actually understand how our business works. Again, right? I thought that somebody was dumb, but I think they were actually the smart one. They might have actually seen around a corner that I was refusing to see around the corner of. So that intellectual honesty bit, it shows up in a bunch of places. But when you feel like your rate at which you are being dismissive of people and things that you feel like don't know as well, because they haven't been doing this as long as you goes up, it's time to do that intellectual honesty check and see if maybe there's something else really going on.
But isn't there a little bit about being an entrepreneur and trying to take the world to a future state that you've imagined that suspending disbelief is a little bit of a part of it? And so it's at times difficult to have the right amount of intellectual honesty. Like I said, differently, I'm sure there's a lot of people before you got into product market fit that didn't get it. And you're like, no, no, no, you don't get it. You're going to get it. And then people started to get it.
And it's sort of the same thing happened in the inverse, basically. Yeah, and I think this is where, this is great because it kind of brings it full circle to that Crosby-Morgen story and the three ideas and everything else and what I was looking for. Because when I talk about
an idea that I could really geek out on and would light up my brain and that I would care about for a really long time multiplied by something that I was like personally predisposed to being good at because of where my career had taken me and what my experience was. I think what I was seeking was an idea where I could have a level of conviction on like a first principles basis that could cut through
the anomalous noise that might emerge and distract me if I went down an overly analysis paralysis kind of direction and tried to run the business. Now, the difference is when we started RJ Metrics, I didn't have that level of conviction. Maybe it was something that lit up my brain. I didn't even know the word dashboard until after I started the company.
So I built this thing and demoed it for somebody and it's a cool dashboard. And we were like, oh, did you come up with that? We should use that word. Turns out it's been around for decades. We didn't know business intelligence as a market. We didn't know that it was a category. We didn't even know like who Gartner was. Because of that, I think we kind of went in a little bit as like mercenaries. We had worked at a venture firm and we felt like we wanted in on this startup game and big data was increasingly a thing. And I knew how to write a damn good SQL query. And that was kind of all we had.
And we went out in that market and we almost, you know, we put sweat and hustle in for the first couple of years and we thought about things really analytically. But, you know, when we got stuck or there was a hard problem or there was a, hey, should we build product direction A or product direction B? More often than not, we would like turn to the data and just look at the customer feedback and say, well, eight people asked for that and only six people asked for that. So we're going to build the eight thing. And I think what happened then was that RJ Metrics kind of became designed by committee.
a way where the lack of this core conviction that a certain specific thing ought to exist because that is where the market is going that was not there this is where the why i actually believe the dumb guy narrative about us falling into product market fit any awareness of where the market was going was subconscious and was a byproduct of me being almost skewed by being surrounded by very forward-thinking people and seeing what those people were doing and building for them they were the ones that led me into a place where i happened to pick up on a theme that would get really big in a couple of
years with crossbeam i wanted to do that in the inverse which was to to be very very conscious of it and have that as a north star to really help dictate what we are and what we are not building here and why we're doing it maybe you index a little higher on the extreme outcomes version like if your north star is wrong your north star is wrong and you're really gonna stick yourself but you
If you don't do that, you cap your upside really severely because your whole company becomes just designed by some consensus of people that don't actually know where anything's going when you average them all out. So it's a good chance to sort of loop back to where we started the conversation around Crossbeam and what
And when we last left off, you went to 20 or 30 founders and kind of workshopped a few different directions. Crossbeam was resonant, at least in the sort of forward momentum. You intersected somebody and they would say, oh, you need to talk to Jane or Sally or this person or that person about it. In sort of as a detailed way as you can, what was then the following six months like?
This is where the repeat founder thing gets really interesting because after talking with a lot of founders and getting some conviction around the idea, I did a couple of things in parallel. One of them was that I initially funded the company. I wrote a check into a bank account, created the C Corp, basically issued a safe note to myself to initially stake the company. And
One of my close coworkers from the RJ Metrics Magento days had left Magento and started his own development shop. His name was Buck Ryan. His company is called the Buck Codes here. And I hired Buck at the Buck Codes here to build an initial prototype of this product that I wanted to use to then
A, understand the materiality of the technical challenges that existed. And I wanted to pick something that was actually going to be hard enough to create a defensibility mode in the technology itself. And I knew Buck was the guy to really, really rapidly do that. So I went into a fairly technical mode, really for just an initial month or six weeks.
And then when we had like the most bare bones version of a prototype, so this is immediately next, right after this six week period is when we started meeting with the personas that we thought were going to be the buyers. It started out with RevOps people and then RevOps people led us to partnership teams. This is where I actually use just
locality to my advantage, which is, you know, I'm based in Philadelphia. There's not a ton of startups in Philly, but there were enough. And I met with every single one of them that was large enough to like have a have a rev ops function. And then I went to New York and I met with everybody that I knew from New York, from all the portfolios that I had been a part of in the past and gotten to know all these founders. And I went back to the founders that I had talked with about these ideas. And I said, OK, I'm ready to talk to your rev ops person. OK, I'm ready to talk to your partnerships person. What was the rev ops conversation like and how did it lead you to partnerships?
So the RevOps conversation was a mechanical conversation about how the heck is this data structured and what about that structure really matters to you when you think about a full cycle, data is going to leave your CRM, something is going to happen to it that makes it more valuable or enhanced or richer, and then you're going to want it to land back in your CRM.
walk us through the raw mechanics of that data egress, the API access, the objects that matter, what requires a custom object versus a custom field. A lot of these, the minutiae of the implementation details, that was all in service of making sure that we had something that was actually plausibly going to be buildable and was going to make sense and get us through the rev ops evaluation, the security evaluation, all the things that were going to make it really sticky to get this thing stood up.
And then in talking about like who would actually use this or own it internally with those folks and with the founders, we consistently got passed to the partnership group as like the first set of conversations, largely because they own the relationships with the companies that they might be connecting to out there.
That initial prototype, then we were able to kind of build it right and start demoing it. We're kind of probably three to four months in now. We're starting to meet with partnership people. And there's a lot of reasons why I'm not sure Crossbeam could have ever been started by a first-time founder, but
This is a big one, which is that I was able to really just go back to like the first company ever on cross beam was stitch. And the second company ever on cross beam was looker. And it's because a stitch was a company I co-founded and I knew the people really well, and they were willing to take a chance on me and looker was their biggest partner and they were willing to
to take the plunge with Stitch and experimenting with this co-selling tool. So yeah, we started basically building the very first atomic network of companies on the Crossbeam platform at this point. So this is probably, we're at the end of 2018. The product's not even live yet. We don't even have a website. Anybody that's onboarded is onboarded manually, but maybe we have four or five companies. We get
guru the knowledge management software company which is another philly company headed up by this great entrepreneur rick nucci who was very generous with his time and suggestions and was an early adopter nick meta a gainsight who i had met along the way in the rj metrics days got gainsight on board really early we had a couple of people come in through like the sales enablement universe so we had this company called sendoso which is kind of a gifting and sending platform that came on early and what was interesting is all these companies our goal with them
was basically to get them to a place where they were willing to invite as many partners as possible onto the platform. There's a lot that we learned in like how to build atomic networks and like spike virality, but the way that it played out was there were a lot of kind of founder-led sales, direct personal work. We got them all to a point of conviction where they said, "Okay, I'll invite some partners on."
And we had these initial waves of companies where I manually got five on and then that five led to 15. What was amazing was of those 15, 10 of them invited another three on and then we got 30, right? And then like this incredible viral loop started
before we had even started optimizing on the way to drive virality through the product or anything like that. One of the things that sort of came to mind is because you were using your first handful of customers were people who knew you and liked you. You would run the risk of them doing this because they like you, not because you're solving a really important problem for them.
Did you think about that? Or you're so convicted in the direction, you didn't need customer validation. You were going to spend the next 10 years working on this, and it was just about finding a means to an end. There's a really important difference between a network effects driven business like this one and a traditional SaaS business where each individual user or player is contained to their own silo, even if they're in like a multi-tenant system. And that big difference is
In the case of something that has a proper network effect to it, the person that you are asking to take a chance on you is not just spending their own time and their own energy to do you a favor and do you a solid. They are also expending their social capital
that they have with other people in order to do it. And that to me is an incredible product market fit test, because it's one thing to have my friend who started another company who doesn't want to hurt my feelings say, Hey, this is really cool. I'll give you a 50 bucks a month just to like make you go away and not have to have an awkward conversation with you and like, let this tool sit here and not get used for other. It's another thing to say, okay, I'm going to go to the most important partners in my business who are responsible for driving a material amount
of new sales pipeline or customer retention or expansion for my business. And I'm going to go to them and I'm going to say, Hey, I know you're busy and I know we have a lot of stuff to work on, but I think you should use this thing and we're going to use it alongside you. And at that point,
that ratio of the, you know, that I talked about earlier, like the FOMO of not working with you versus the cost of like having an awkward conversation, that ratio is actually flipped because now that cost of having an awkward conversation is also the cost of you burning social capital with an extremely important business relationship that you have, which is a really critical partner. So my thesis there was, I don't think that we could delude ourselves into thinking we have something when we don't for very long, because I
By the time we get more than one degree of separation away from me and my personal relationships, any virality that we see has nothing to do with my relationships and everything to do with people being willing to spend social capital just to use the product more. The only upside would be that there's more people on the product and more data they can get access to. So I think if this would have fallen flat, it would have fallen flat so incredibly immediately because my bluff would have been called the second I said to Rick
Rick Nucci, hey, I know you said you like this, but can you also invite Zendesk on here to connect? That's a real show your cards moment, right? Because Rick's a nice guy, but he's not going to burn a relationship with a publicly traded partner company. Yeah, but I don't know that that dynamic is always true, right? That's because of the nature of how viral the product is.
Once you left those early conversations with founders, was your attitude, I'm going to go make Crossbeam work? Or was it I'm going to spend the next six or 12 months to figure out if there's a there there? I got pretty high conviction pretty fast. It was I'm going to make Crossbeam work. The proof in the pudding to that is that I went out
in July of 2018. So this is before Stitch and Looker had even signed up, but I probably had the first wave of conversations with them. And we'd had those RevOps conversations and we were building the version of the product that would actually work. And I raised a $3.5 million seed round from you guys and had that initial set of conversations with First Round who co-led the round with Uncork. And we got a handful of other folks around the table who had invested in my previous companies and things like that.
And I think that was kind of me signing up for the point of no return. Like I had written that safe note in and we probably could have made it to the end of the year, just kind of spending down my money. But it felt like while the idea had maybe only been being kicked around for six months, it had shown up in that Evernote file three years before that. And my ability to, I guess, pattern recognize as a side hustle, creating the validation around this multiplied by what I was seeing out on the market, just
put me in a place where I didn't want to slow play it. And look, I also wasn't in the same situation I was with RJ Metrics, right? Which was like, I didn't show up to necessarily do a classic lean startup, build, measure, learn playbook on my next thing. Like I wanted to pick something that would be very, very hard to get started, which network effects businesses are. There's this huge cold start problem. And by definition, doing that rapidly would be capital intensive.
I think I had a handle on that pretty early on and went down the road of being venture backed almost immediately with this one. You touched on this a little bit, but what was the first version of the product? How did you figure out how minimum it could actually be? One of the things I learned really quickly was that
people actually already do this at almost all companies and i had no idea the process of like cross beaming back in the day before cross beam was called account mapping and the account mapping process is an exercise by which two companies who are collaborating with each other basically
email a bunch of spreadsheets back and forth to one another. And those spreadsheets are typically sent either by partner teams or by individual sales reps or by partner managers to sales reps, and they contain lists of accounts. And those accounts, depending on your situation, they could be customers, they could be prospects, they could be open opportunities, but they're very often like
It's not like every open opportunity in your Salesforce instance. It's like whittled down to a very, very small specific target list. And what you're basically doing with your partners is like playing battleship. It's like, all right, I've got 500 open opportunities right now. I'm going to pick 17 of them because I think maybe they're in the same market. And I'm going to send them over to this partner because I don't want to send all 500. I'd be like, you know, giving up the Glenn Gary leads here. I love my partner, but not that much. So
I think I'm going to take a shot at these 17, send them over and they say, oh yeah, there's actually four of these that we're working right now as well. Maybe we can coordinate on, you know, get our reps connected, et cetera. Very, very manual process. The data is overwhelmingly incomplete. It happens rarely enough that the data is always extremely stale and it's impossible to act on like real-time changes or things being dynamic in the pipeline. There's huge security and compliance related issues with doing this at all, which is crazy that it's such the gold standard. This is really, really fraught process that was basically true.
completely run even in multi-billion dollar market cap businesses by spreadsheets and email, spreadsheets and writing VLOOKUP functions in Excel. Our goal, the initial version of Crossbeam was to replace that and basically just to create a spreadsheet interface where both sides could
connect data. And we had a Salesforce connector and we had a CSV upload. And those were the only two things you could do. And then once the data was in, you built these things called populations, which were basically like segments of data. And your populations could be your prospects, your opportunities, or your customers. That's like your standard groupings of companies that might exist in your funnel. And then
That was a way of creating like a universal data mapping kind of very loosely. We sat in the middle. When you partner with somebody, we would show you what was at the center of the Venn diagram. And we had this thing called the account mapping matrix, a three by three matrix. So prospects, opportunities, customers on one side, prospects, opportunities, customers on the other. And we just give you the count of
of how many overlapped in those nine boxes. So how many of my prospects are your customers? How many of my opportunities are your opportunities and so forth. And then you could click on the box and boom, it drops you in the spreadsheet. And here's the spreadsheet of all the ones that overlap, but you can't see what doesn't overlap. Each side keeps the rest of their stuff private. At its core, that was the idea behind the engine. It was even more simple than that in some ways. Standard populations actually came a little later. It was basically a collaborative spreadsheet where you could only see the rows that overlapped, right?
That's probably the easiest way to put it. You brought that to how many customers and what was the reaction? Where did that leave you? Network effects was like the whole ballgame in the early days. So we have one of two outcomes when those situations came up where we brought it to a customer. One of them was that it was dead on arrival and went nowhere. And the other one was that it was explosively and exponentially
growing. There's a pretty stark difference between the two. It's a little bit of this, the product market fit test that you guys have featured that has been really popular, right? To find the 40% of the people that are just ridiculously passionate about using it and then focus in on that profile. That was very much our situation. Like more than half the companies that we showed this to or that signed up kind of went nowhere, especially once it just randomly landed on our website because they saw like a funding announcement or whatever.
And what we learned really quickly, and this is obvious, is imagine that you sign up for LinkedIn, land in LinkedIn, and none of your contacts are there. You're going to have a bad day. There is no value proposition associated with joining a social network where none of your friends are there unless you are willing to do the heavy lifting. Again, spending your personal or social capital to recruit them onto there.
And people that don't really have any context or investment or knowledge about what this thing is are probably not going to do that. The ones that were dead in the water were the ones who desperately needed a single player mode. And there was no single player mode. And they kind of went nowhere. So what were the ones that were explosive? Well, it turns out the ones that were explosive was anyone who was invited in by somebody else. When you're invited in by somebody else, your onboarding is a completely different experience. Because you get an email that says so-and-so company wants to share data with you right now. Come sign up.
And then it says like, you're one step away from getting that data. Just connect your Salesforce. And then it's like, you're just two clicks away from getting that data. Just define these populations. And there is this incredibly valuable, real carrot that can be dangled in front of someone to guide them through onboarding because they're constantly marching toward that first connection and that first realization of the data. And then when they get it, they see the value of getting access to that data. And they say, wow, I wish my other partners were on here. And they invite all their other partners on.
And then all those partners, how did they get onboarded? By receiving an invite. And it perpetuates this really incredible thing. So pretty early on, we realized all the stuff that always worked at RJ and Stitch wasn't going to work.
work. Content marketing, buying ads, hold outbound, anything that lands a fresh new person that's never seen it before on our website to just sign up directly through our funnel would fail miserably because they'd show up, they would have no connections, and they would be in a single player mode environment with no single player mode. The only way to make this thing go was to get the people who were already on it inviting more people and
to seed new clusters of people to start that viral mechanic going, right? So then you have two things going at once. One of them is product work to make sure that people that are already on it are disproportionately incentivized to invite more people on it and also aware of who is on it and is not on it and upping the connections. The other thing is how do you get more nodes in to like kind of kickstart this in different verticals? And this is where we pioneered this thing called a joint jam session.
We didn't do product demos or sales calls unless two companies were on them at the same time. So in the early like founder led sales days, if we had a new person show up and came to the website or like try to sign up for the product, but they had no partners and we could get them on the phone, we'd say, Hey,
And they say, hey, can you get a demo? And I'd say, yes, but please bring your biggest partner. When you bring your biggest partner and we'll do this kind of like three-way demo. And what we would do on the joint jam is we would get both companies signed up. We'd get their data connected and we would actually live on the phone, show them both. We would bypass single player mode that either of them individually would have experienced live on the demo call to get them to that first atomic state. And there wasn't an issue where like these were partnership managers or...
I see sellers or sales leaders and you also need sales ops or somebody to actually authorize the Salesforce connection? In the early days, shockingly, it was less of an issue than you might expect. And a lot of that was because so many of the companies we were onboarding were
early stage businesses. I mean, this wasn't like a snowflake in Adobe coming on and Looker certainly wasn't anomaly, but Looker was probably the biggest company that they, they happened to be the second company that signed up. They were probably the biggest company on there for the first year and a half of the business. Right. So we had a lot of like
seed stage and series A stage venture-backed companies. We'd occasionally get some big fish due to my personal relationships, like the Gainsites and the Lookers of the World. And those ones would have a very different journey. We would go through RevOps. We would go through kind of security reviews and things. And this also was like in the earliest of early days of like GDPR and CCPA and like
all things that we've been able to, you know, kind of cover our compliance requirements around on an enterprise basis that come up a lot more now, they just didn't come up as much back then, right? Oh, and also there was always the CSV uploads, right? So even if the initial data wasn't connected on the CRM, if you could get into your CRM and export the same information and then upload it, then you could also make it work long enough to prove the value and then take it to RevOps to get it connected. So, you know, we were able to kind of get
past that in enough cases that the viral machine started going. I was just pulling up this data, right? So like first customers were in September, 2018. At the end of 2018, there were four registered companies. At the end of 2019, we got up to 300. At the end of 2020, we were at 1800. At the end of 2021, we were at 5800. The end of 2022, we got up to 12,000.
And then we were over 20,000 by the end of 2023. And seven out of every 10 of those companies that signed up for the platform signed up because they were invited in by somebody else who was already on there. It's obviously very unique for B2B software to have a network effect.
It's very rare. How did that influence the way you thought about early pricing? Because obviously the durability of the business is generated by having every single potential customer on the platform. But you also need to make sure that you're capturing economic value to even as some sort of proxy metric to show the value you're creating. Because if you have a huge network and nobody's willing to pay you anything, that's not really worth very much. Pricing was like a huge
albatross hanging around my neck for the first four or five years of this business. The whole thing, when I talked about earlier, it being potentially like a capital intensive business where we wanted to raise venture early and be able to have a very long-term view the whole time was because of this exact thing you're asking about. So, you know, spending social capital is one thing, but if everybody had to pay to be on the platform to get their very first look at
the data, then we were going to introduce an incredible friction that was going to bypass the whole reason for trying to build this thing in the first place. And the whole idea here is to build something that creates an extreme amount of value because it is ubiquitous, right? Because it is a LinkedIn for data. We...
basically didn't charge anybody anything for the first two and a half years of this business. We had a couple of paying clients, but I used to say in the venture meetings when we were raising our A and a little bit, we had a little revenue by the time we did the B, but there was a slide in our series A deck that said, you might look at our P&L and see positive numbers. To be clear, we don't have ARR, we just have R. And the idea is every dollar that we had flowing in was because I
I had custom negotiated some kind of payment from someone that was not always necessarily guaranteed to be annual and not as always necessarily guaranteed to be recurring. We had a lot of these design partners that we embarked into bigger feature development and kind of ideation with, and we might charge them 10 grand or 20 grand or something just to kind of have a
a little bit of a stake in the game for them and wanting to see it through and get some value out. But we were not optimizing around like conventional SaaS metrics. We were very much treating this almost like an open source software product would be treated, which is you want to get your GitHub stars going. You want to get the people that are like extremely passionate about this thing. They want to see it exist. And by virtue of doing that,
They themselves are willing to actually spend their own time and energy in making it better and basically contributing to it. Our network graph is an open source software product with 30,000 contributors on it now. And that philosophy allowed us to kind of map out the expected growth curve of the company and the funding strategy for the company to mirror the way that you would approach funding and growth for an open source software company.
which kind of has this very big upfront community development style investment period that's not extremely lucrative from a revenue standpoint. But then you get to the point where you're ubiquitous enough that there are actually very material enterprise applications and kind of lightweight cloud-based deliverability paths that could be more PLG in nature. And you bring those things to life and the business side of the business blossoms. And that really, you know, the last two years or so in the business has been that
part of the journey for us. - Talk about the reveal merger, because it's so rare to have
of two scaling private companies merge. It basically, one, never happens, and historically when it does, it doesn't always lead to great outcomes. - Yeah. - And it seems like, obviously, you didn't do that without thinking about it carefully, and now you're sort of on the early innings of the other side. Would be great to hear your sort of founder lens on why you chose to do it, particularly knowing that it's so atypical. - We've talked a lot about how
Crossbeam is not your standard B2B SaaS company. The actual proper network effects dynamic is extremely unique. We're also a business that kind of like came up in the zero interest rate phenomenon era and was able to raise capital prior to having revenue at like a pretty grandiose scale relative to what a lot of companies are capable of doing. And I think that was repeat founder multiplied by incredibly compelling virality across
in the core usage of the free product multiplied by the Zerp era, right? Kind of equals ability to raise a lot of capital. Yeah, or like building a business that's like really working and like fulfilling what it was supposed to be at this part of the journey. But this other thing happened, which was basically at the same exact time within six months of me starting this company, another guy named Simon Boucher, who...
He's a multi-time founder based in Paris who had most recently sold his last company to SAP. It was in the HR tech space. Started a company called ShareWork, which would later change its name to Reveal, doing basically the exact same thing. From a core North Star building of a graph of connected companies to allow for a modern replacement for account mapping to power the co-selling and cross-selling motions traditionally owned and taken on by partner teams.
all in service of an ecosystem-led growth strategy being deployed among the modern cohort of companies. It was kind of a textbook example of them doing this. I would call it a copycat business, but I authentically don't think it was. I think we came up with the same idea at the same time because we observed the same things at the same moment.
and he was based in Europe. So they had collected a lot of business in Europe that where I was strong and Crossroom was strong in the analytics vertical, they were really strong in the HR vertical. No surprise. This is where our embedded networks are. This is where our atomic networks came from. Oh, and where we had raised $100 million from Andreessen Horowitz and others during the Zerp era, they raised $50 million from Insight Partners and others during the Zerp era. Yep, that's right. Plot twist. The same Insight Partners that I worked at many, many years before.
About a year and a half before the deal happened, I met Simon at an event for partnership leaders that was in Miami. And we had a brief conversation there that was about nothing to do with the business, but a little bit about the silliness of being repeat founders. We both have young kids that are like almost the same exact age. And our stories, despite being theoretically arch rivals out there on the market, we're basically like the same. He's like the French version of me. And I'm like the
Philadelphia version of him, right? We have such an incredible amount in common. And I think one of the things that we discovered in that 15 minute conversation is sometimes you meet like a CEO at a conference, whether they're a competitor or not. And you just like from a personality standpoint, don't vibe with them. And it's usually a mismatch and something like a fundamental values level, right? Simon and I both have that intellectual honesty thing that we're kind of somewhat obsessed with not necessarily doing this for the glamour, or even necessarily for the spoils, but
because of the way that it lights up our brains and how interesting the problem space is. We got along really well, well enough that we obviously, you know, care a lot about keeping a giant air gap between us, but we felt like we had formed the basis to like, hey, if there ever was something for us to talk about, then we'd feel comfortable kind of like initiating that conversation.
Meanwhile, we both went back to work, you know, for another several months. And, you know, we're looking at like our product roadmaps and what to build next and everything else. And I happen to notice in our product feedback channel and in all of the data collecting that we do among our customers that the number one requested feature throughout all of Crossbeam was for us to integrate with Reveal and to support the companies that were on the Reveal network.
Getting back to the LinkedIn analogy, imagine if you signed up for LinkedIn and all the people that you worked with at your first job were on your LinkedIn, but all the people that you worked with at your current job were on the other LinkedIn. Which LinkedIn sales navigator do you buy? Uh, which LinkedIn recruiter do you buy? Um,
Where do you go and spend your time when you want to upload that viral selfie from the conference? Like we had a massive universe of customers who were having a compromised user experience and a degraded value proposition because their network of companies they wanted to connect with was actually split between these two distinct network grads.
And it was leading to not only an extreme amount of confusion and frustration among our customers who had to hop between different tools, but a capping on the amount of investment that any of them would make in one tool or the other. Because why buy our Salesforce integration to push the data back into your custom object if we only have half the companies that you want to connect with? And why buy both if your RevOps team doesn't want to integrate two different products whose custom objects aren't compatible with one another?
It was broken. The whole market was broken because there were two. Fundamentally, very, very, very different than if you just had two companies that were building competing customer success software or building competing marketing automation software where the value proposition is contained into like your instance or your silo. The
The customer started pounding the table and it was like, it was, I saved a bunch of these. There are all these LinkedIn threads. It's like, for the love of God, why don't Crossbeam and Reveal just merge? Right. And like we'd host a webinar and people would put in the chat, they'd be like, when is the Reveal integration coming? And it got to a point where it felt like we both arrived at a spot where we said, we've got to put egos aside and
do what's right for the customer here. And actually, neither one of us is going to realize the true potential of our business unless these businesses are brought together. A natural response might be, well, why don't you just go out and compete like hell and take the other one by winning, by kind of on the virtues of
We're all just like building a better product, et cetera. But what it boiled down to, we did that for years, but both of us had raised so much capital that the other one was clearly not going away. At the same time, we were both in a spot where I guess like the important thing to us was making this as big as we could possibly make it. And not having our particular company's name on the, on the top of the business card or
or going down in history as the one that won the big battle. We were going to destroy each other by coexisting for so long in
in such a poor market dynamic that neither one of us would necessarily rise to the top. And we were both willing to set aside some dilution and some pride, some sense of control in order to bring this thing together into being one cohesive thing that will actually create what we wanted to create for our customers in the first place. And by the way, still compete with big
big tech, right? We're two very small kind of venture-backed player. We've raised a lot of money, but we're not enormous companies at like an IPO scale just yet. We're out there competing for budget with an entire universe of revenue orchestration tools and data enrichment tools and intense signal tools and other products or partnership technology tools. So there's competition, don't get me wrong. We just did this one particular thing that necessitated a network graph
And we had to come together. And we both had like five plus years worth of runway in the bank. And we decided to do this deal as an equity deal with no cash exchange and just bring it all together under one hood. We are in this position that is just extremely strong from an ability to go operate for a really long amount of time on a really big problem as a company that's positioned in a way where we can actually tackle it. And that seemed really, really exciting. What surprised you now that it's closed?
You know, this is a funny answer that I hope is not, uh, that no one takes the wrong way. When you talk to Americans about merging with like a French company, you kind of get this reaction that's like, Oh boy, Oh boy, better get ready. They're all going to be smoking cigarettes and, uh, you know, uh, sitting on the roof at 2 PM and taking their month long vacations in August. And, uh,
It's impossible to part ways with anybody. You're obligated to employ them forever. Some of those things may have some partial truths to them, but us Americans aren't perfect either. The thing that surprised me the most is the incredible work ethic of the people from the reveal team.
This team works hard and is extremely smart and absolutely goes toe to toe with anybody that I've employed at any company that I've built along the way. They have built an incredible company there. And I think, and I spent a good amount of time in Paris and increasingly in the Paris startup scene in the last year or so, Paris in general has this incredible energy to it from a startup perspective.
perspective in general, there is a very real generation of knowledge workers and technical workers in Paris. A huge amount of them working on AI stuff, by the way. The AI ecosystem in France is really strong. And Matt Turk, one of our investors from Firstmark, is on the board of a bunch of companies there that are at the helm of a lot of that. I feel lucky to be working with so many French people.
And I think that like, I want to continue investing in growing our Paris office. And I think that that was not, that was one thing that was going to be impossible to do diligence, right? You kind of had to sign up for knowing that there were maybe some complexities or some nuances there, but man, some really, really great people have come along with this deal. In wrapping up, I want to end where we normally do, which is the question of who across this sort of set of crazy company building journeys that you've
been on, would you say has kind of had the biggest outsized impact or imparted
something on you that has really stuck and what's like the thing that you picked up from that person? I always jump to Jake Stein, who is my co-founder from RJ Metrics and Stitch, who's now the CEO of a really great company called Common Paper, which is kind of the they're treating legal documents like APIs and kind of building a next generation technology on top of helping folks construct and close more efficient legal agreements, mostly commercial contracts.
Jake and I are very different on just about every possible dimension. But we ended up working together for a decade, like really arm in arm. And I think we learned a lot from each other and complemented each other really well. But one thing that always this is a silly example, but it's like this Jake Stein story that just sticks with me so much, which is like in the early days of RJ, I think it was just the two of us in an office in Camden, I had like, gone home over the weekend and read whatever the the
the hot startup book of the day was right in 2009 or whatever. And I came into the office on Monday with like all these ideas and like a couple of post-it notes in the book. And I was like, this book has so much good stuff on it. Like we need to dive in and like, we should do this and we should do this. And he was like, that's super cool. I'm so excited you read that book. What in that book did you not agree with? It kind of like took me back a little bit because I don't think that
that I had thought about that or that I had allowed myself to exist in a space where I could hold two thoughts at the same time, which is that there might be some extremely good
and insights sitting inside of this book I picked up at Barnes and Noble. And there might also be some really bad advice that doesn't work for us or doesn't apply to our business or that is informed by something that was relevant for somebody in some circumstance but doesn't apply to us. And I think I treated it all on the same level, like as gospel. And I think the thing that Jake brought to me and my thinking was not being more pessimistic or being more critical, but being extremely capable of holding two conflicting thoughts in my head at the same time.
And he was always very, very good at that. And I think that when you talk about intellectual honesty, I think that is like a really core ingredient of being able to do that and not turn out as a blind optimist who says yes to every idea or as Jake would sometimes call himself the Charlie Munger to my Warren Buffett, like the abominable no man, right? His job is just to say no to everything because I said yes to everything.
The reality is that that's how it started out. I was the yes man and he was the no man. But over the course of the decade, I think we kind of converged into both really being able to hold those two conflicting thoughts at once. And I think it's been so, so critical to like not losing my mind doing this job over the course of the time I've been doing it. Great place to end. Great conversation. Thanks, Brad. Always a pleasure. Yeah, great one.