Vercel's CFO Marten Abrahamsen: Move Fast or Fall Behind

Summary

---

Run the Numbers: Vercel's CFO Marten Abrahamsen – Move Fast or Fall Behind

Podcast Host: CJ Gustafson
Guest: Marten Abrahamsen, CFO of Vercel
Date: June 8, 2026

---

Episode Overview

This episode features an in-depth conversation between CJ Gustafson and Marten Abrahamsen, CFO of Vercel, at the New York Stock Exchange. The discussion centers on leading finance in a hypergrowth, AI-driven tech company, the evolution of public and private markets, practical AI adoption in finance teams, the balance and tension between speed and precision, and building a business model at the intersection of product-led growth (PLG) and consumption-based revenue. Marten provides tactical insights into financial planning in uncertain environments, building internal AI tools, capital allocation, benchmarking in uncharted territory, and what truly drives durable enterprise value in today's landscape.

---

Key Discussion Points & Insights

1. Speeding Tickets vs. Parking Tickets: Bias for Action

[00:00, 19:43, 24:07]

• Concept Definition:
  Marten uses the analogy of “speeding tickets” (moving too fast and making mistakes) vs. “parking tickets” (sitting still, waiting, and missing opportunities) to explain Vercel's bias for action. The company prefers action and learning quickly over risking inertia.

  • Marten: “What we don't want is people sitting still and waiting. That's a parking ticket. …A speeding ticket is you move too fast, right? …We would rather have someone move too quickly on something, bias for action, than just sitting around and waiting.” (00:04)
  • CJ: “If you're going to be setting the pace for the rest of the org, I think you do have to take it upon yourself to show that you're willing to move fast.” (00:39)

• Leadership Tone: The finance team is expected to lead by example, showing that even in compliance-heavy areas, speed (paired with good judgment) is valued.

  • CJ: “If you're going to be setting the pace for the rest of the org… you do have to take it upon yourself to show that you're willing to move fast.” (23:49)

---

2. Evolution of the Public and Private Markets

[03:21, 05:08]

• IPO Benefits and Private Market Growth:

  • Marten contrasts the past, where IPOs were essential for access to large capital, with today’s massive private markets ($22-24 trillion in private company market cap).
  • Private company liquidity, large fundraises by companies like OpenAI and Anthropic, and widespread usage of tender offers have shifted the traditional incentive to go public.
  • Notable Moment: Recalling the Visa IPO during the 2008 financial crisis as a defining example of a company ready to buck the broader market:
    • Marten: “When people say that the IPO market is closed... that tends to refer to, like, the median company. Like, for most companies, it is closed. But if you're a Visa type of a company, you can go out even in a very, very difficult time.” (03:47)

• Consolidation and Public Company Trends:

  • Fewer public companies due to increased M&A and private equity activity; growing concentration of value among mega caps.
  • Marten: “I think there'll be more consolidation. I saw some other stat… the sum of the top 10 companies in the S&P 500 was 19% 10 years ago and today it's 38. So ... there's more and more value in market cap accruing to the larger companies.” (09:04)

---

3. AI Adoption in Finance: Building, Not Just Buying

[13:05–18:24]

• Internal AI Tools at Vercel:

  • The finance team dogfoods Vercel’s own platform, creating an internal KPI dashboard, a revenue flux analysis agent, and a custom revenue recognition tool.
  • KPI Dashboard (Mobile Optimized):
    • Pulls data directly from Snowflake with real-time metrics for signups, ARR, net dollar retention, etc.
    • Marten: “We built a...user interface with the exact metrics and the way we like to look at the business. Pulling data directly from Snowflake on my phone, it's...protected via SSO…” (13:20)
  • Revenue Flux Analysis:
    • Aggregates data from Netsuite, Snowflake, Slack, and Notion to explain month-over-month variances, highlighting the use of Slack context—channels recording deal wins—as part of the flux narrative.
    • Marten: “The agent will search all this data and...explain exactly why the flux has changed. We assess it and sometimes like, oh, well, it doesn't...get that piece quite right…but it gives us a really good first draft.” (14:53)
  • Automated RevRec Tool:
    • AI analyzes contract dumps, flags “vanilla” or “complex” items, drafts rev rec memos, suggests human review for edge cases.
    • Marten: “It will draft the actual revenue recognition for you and highlight which contracts should be reviewed by human because there are some intricacies…” (15:38)
    • CJ: “People talk about having a human in a loop. It sounds like you're actually using the tool to pull humans into the loop when needed.” (16:55)

• Build vs. Buy & AI Upskilling:

  • Marten notes that today, with tools like Codex or Vercel’s V0, anyone can build AI assistants—coding is no longer a strict requirement.
  • Vercel finance and accounting candidates are expected to know how to build with these tools as part of their interview process.
  • Marten: “It's part of our interviewing process and we encourage and celebrate everyone to use it...half the meeting is people giving demos of stuff that they build in AI…” (17:57)

---

4. Impact of AI on Talent and Hiring

[18:24]

• Shift in Evaluation:
  • Old case studies emphasized Excel prowess, now replaced by judgment-based assessments and practical AI tool application.
  • Marten: “The case studies today are...much more about how would you assess X versus Y. And then we tend to throw a curveball...how does this change your conclusion? …You have to build something using a Vercel AI tool.” (18:38)

---

5. Moving Fast in Uncertainty: Decision-Making in Hypergrowth

[24:07, 25:02]

• The “One-Way Door” Fallacy:

  • Most business decisions aren’t irreversible. Vercel encourages action, experimentation, and iteration over perfectionism:
    • Marten: “There are actually exceptionally few one way doors out there...try something out. …Ship a new product, let's learn and improve it as opposed to let's just wait and see.” (24:17)

• Planning for Hypergrowth:

  • Annual plans are often obsolete “immediately.” Vercel now models three forecast scenarios (“like growth charts for your children”) and adjusts spend dynamically by tracking which “railway” the company is following based on performance.
  • Marten: “Every year we're basically off plan almost immediately, which is a good problem to have...So this year we're operating actually with three different models...and then we're spending accordingly.” (25:17, 26:33)
  • Marten: “It is absolutely not linear. It is also not perfectly exponential. …When you live this stuff day to day, you find yourself oscillating between…how amazing the company is doing and then sitting in sort of an internal yellow flag meeting.” (26:58)

• Budget Communication & Leadership Alignment:

  • Budgeting is collaborative, flexible, and transparent—with a focus on efficient growth, expectation-setting, and helping functional leaders who may have varied financial experience.
  • CJ: “You have to meet the leader where they are in their sophistication with the budget and help bring them along the way.” (28:54)
  • Marten: “We're trying to sort of give people visibility into also what's coming to help them make these decisions...You also need to bake in some flexibility.” (29:23)

---

6. Business Model Innovation: PLG Meets Consumption

[30:56–37:47]

• Benchmarking Challenges:

  • Vercel blends PLG (high-velocity, credit-card, self-serve) and usage-based billing—no perfect public-market comp exists.
  • Marten: “There hasn’t really been a consumption PLG self serve business in the public market yet.” (32:16)
  • Cohort volume and segmentation are leading indicators, with different expected growth curves for hobbyists, teams, and enterprise users.

• Performance Metrics & RPO:

  • Traditional SaaS metrics like RPO (remaining performance obligations) don’t fit. Most customers are monthly, non-committed, and must be “earned” every month.
    • Marten: “RPO is actually not so good for us either for the analysts…Majority of revenue is PLG revenue, which basically has no RPO.” (34:21, 34:39)

• Go-to-Market Observations:

  • Increasing shift from multiplayer/team usage to solo developer or even non-developer use cases.
  • Accidental consumer adoption (hobbyists using AI over holidays) now drives top-of-funnel, and business customers increasingly test tools individually before enterprise rollouts.

• Evolving TAM:

  • The expansion of addressable market beyond traditional “developer” profiles, as AI lowers the barrier to software creation.

---

7. Capital Allocation and Creating Enterprise Value

[39:20–49:01]

• Experimental Approach:

  • Investments are made with clear learning objectives (“what you need to believe” models), emphasizing rapid measurement and iteration, even when data is incomplete.
  • Marten: “You have to make decisions with perhaps less complete data than ever before. …We try to measure quickly and if something is not working out, we iterate.” (39:27)

• Growth vs. Margin:

  • Growth at scale drives compounding value more than early margin optimization—especially in fast-moving markets.
    • Marten: “If you optimize for margin too early, you're likely going to miss out on that growth curve. ...Don't make short term decisions that will reduce your growth in favor of margin.” (41:31)
    • CJ: “How do you know to keep accelerating into that? What are the signals?” (42:34)
    • Marten: “Are we able to maintain or accelerate our growth rate? …We do look at…do we have a healthy cash balance relative to burn? …If we don't have the biggest bucket and we possibly can to capture this wave, someone else will.” (42:37, 43:09)

• Industry Margin Expectations Are Shifting:

  • AI infrastructure raises the cost to serve; software’s classic 80-90% gross margins may naturally fall as “build-it-yourself” alternatives proliferate and price competition increases.
    • Marten: “If you sort of take a step back…what does like 80% gross margin mean in practice? …Now that we have AI and more compute and personalized experiences, it is going to be more expensive to serve customers.” (43:20)

• Terminal Value and Modern Cash Burn:

  • Modern AI companies may burn cash for longer but build more substantial terminal value than some of the 2021 “growth at all costs” cohort.
    • Marten: “...there is likely to be meaningfully more substance in that terminal value growth rate driver today than it was in 2021.” (44:32–44:51)

• Gross Margin and Profitability:

  • Gross margin may contract, but EBITDA can remain robust via opex efficiency and scale—true value comes from absolute cash flow, not percent margin.
    • Marten: “...optimize for gross margin dollars and then…for EBITDA or free cash flow dollars.” (46:29)

• Unit Economics:

  • Still crucial as directional instruments, even when absolute numbers or ratios matter less at hyperscale:
    • Marten: “All the components of unit economics are really good metrics to track…if we're starting to see some deterioration…then we know where to dig in and focus.” (48:09)
    • CJ: “It's not that you're not tracking them, you're just caring about the direction and the velocity and pace of change.” (49:01)

---

8. Memorable Quotes & Moments

• On AI Building and Team Culture:
  • “Almost anyone can build these AI tools. ...We encourage and celebrate everyone to use it...people giving demos of stuff that they build in AI.” (17:53, 17:57)
• On Growth Compounding Value:
  • “Growth compounds value in a way that margin improvement can't.” (41:31)
• On “One Way Door” Decisions:
  • “There are actually exceptionally few one way doors out there.” (24:17)
• On Industry Shifts:
  • “AI is going to shift...the industry margin expectation structurally downward.” (43:20)
• On Hypergrowth Planning:
  • “Every year we're basically off plan almost immediately, which is a good problem to have.” (25:17)
• Expense Anecdote (Lightning Round):
  • “I once had someone try to expense an oven...because…they didn’t want to eat takeout food.” (52:41)

---

Important Timestamps

• Opening Leadership Principle — Speeding vs. Parking Tickets: 00:00–00:19, 19:43, 24:07
• Visa IPO Story / Public-Private Market Evolution: 03:21–07:57
• AI-Driven Finance Workflow Demos: 13:05–18:24
• AI-Driven Hiring & Judgment Over Mechanics: 18:24
• Dynamic Budgeting in Hypergrowth: 25:02–30:31
• PLG + Consumption Revenue Model: 30:56–37:47
• Capital Allocation / Event Experimentation: 39:20–41:21
• Growth Compounds Value Statement: 41:31
• Shifting Margin Expectations with AI: 43:20–46:13
• Lightning Round (Mistakes & Stack): 50:04–53:05

---

Episode Takeaways

• Move fast, but with judgment: Vercel’s culture prizes forward motion—even if it means occasional “speeding tickets”—over the risk of missed opportunity (“parking tickets”).
• AI for all: Advanced internal tools are now accessible to finance and accounting teams, democratizing automation and analytics.
• Financial planning in hypergrowth is a dance, not a blueprint: Consistent recalibration, scenario planning, and transparent internal communication enable more robust decision-making amid volatility.
• Business models are evolving beyond legacy benchmarks: PLG plus consumption revenue demands new cohort analysis, metrics, and investor communication practices.
• Growth is the compounding engine: In transformative markets, maximizing optionality and market share beats early focus on margin.
• AI is compressing gross margins, expanding addressable markets, and forcing organizations to rethink value creation.

---

Notable Quotes

• Marten: “It's not like we just had agents do all the work and the team is relaxing. It's that we can be much, much more efficient...” (16:55)
• CJ: “Do you think that there are companies out there who are growing really fast but they're potentially destroying value?” (47:25)
• Marten: “If you're a gross margin negative company...even if you're capturing a lot of market share, probably more challenging to create a high value business.” (47:30)

---

This episode is a must-listen for finance leaders, operators, and anyone building or investing in next-generation tech companies, particularly those grappling with the changing landscape of SaaS, AI, and modern capital markets.

Transcript

[CJ]

Speeding tickets versus parking tickets. I've heard you throw this around before. I got to know what that means.

[Martin Abrahamson]

It's basically a bias for action. Brazil is a very fast moving company. We may not at all times have perfect direction for what every single person should do. What we don't want is people sitting still and waiting. That's a parking ticket. You sat still for too long.

[CJ]

You didn't move.

[Martin Abrahamson]

A speeding ticket is. You move too fast, right? In a chaotic world, you went and did something perhaps you hadn't fully thought through or perhaps you hadn't like kept everyone in the loop. Obviously we always strive to communicate and do this a little things, but for the most part, we would rather have someone move too quickly on something, buy us for action than just sitting around and waiting.

[CJ]

If you're going to be setting the pace for the rest of the org, I think you do have to take it upon yourself to show that you're willing to move fast. Is this thing on?

[Martin Abrahamson]

Yesterday's price is not today's price.

[Podcast Host]

Welcome back to another special episode of Run the numbers at the New York Stock Exchange. Shout out to our friends over at

[CJ]

the NYSE for letting us in.

[Podcast Host]

I recorded this interview with Martin Abrahamson, the CFO of Vercel. Bet you heard of them. They're growing pretty fast. We talk about AI adoption and finance. He gives some tactical use cases that his team at Vercel is doing to apply finance to revrec, to forecasting, to dashboards they're creating to discuss on a weekly basis. And then we get into what it's like running a company in hypergrowth. A lot of the forecasting norms that we're used to get thrown out the window. So how do you figure out the direction and speed of a company that is changing by the minute? And what's really interesting is that they're at the intersection of PLG and consumption. So these are two different things that like historically PLG was subscription led and

[CJ]

consumption or usage based was a whole different bag of burritos.

[Podcast Host]

So he explains how he thinks about monetizing this business and and how they can come up with comps that may not be there yet. And finally, what drives durable enterprise value? This guy really explains it well and how he thinks about what a point of growth is worth, especially when you have a market opportunity to seize. For all those out there who are wondering what it's like to run an AI company that is breaking all sorts

[CJ]

of records, tune in for this one.

[Podcast Host]

It's a special one if you like this. I Did a deep dive on a cfo. Explains the history of stock exchanges. Yes. Yes. The very place I'm sitting in to do this interview, I went deep, deep, deep, deep.

[CJ]

Martin, thank you so much for making time and coming out to the New York Stock Exchange.

[Martin Abrahamson]

Thank you for having me.

[CJ]

I want to start with some quick hitters here. Do you remember the first stock you ever bought?

[Martin Abrahamson]

I didn't technically buy it, but I was gifted a stock.

[CJ]

Okay.

[Martin Abrahamson]

Which I do remember, which was. I was born and raised in Norway, and the stock that I received from my grandfather was a stock of Telenor, which is a Norwegian telco company.

[Podcast Host]

Yeah.

[Martin Abrahamson]

And I talked to him about it because, you know, I was like, 13 years at the time. I got my first cell phone and I was texting a lot. So I was really thinking I was contributing to telehealth said revenue. And he was the only person that I could, like, talk to about that. My parents had no interest for season, my birthday in November or Christmas. I got one share of teh.

[CJ]

One share?

[Martin Abrahamson]

Yes.

[CJ]

Are they still around today?

[Martin Abrahamson]

They are still around today.

[CJ]

Did you keep the share?

[Martin Abrahamson]

I actually don't exactly know when I sold it, but I do not currently own that share.

[CJ]

When you think of IPOs and ringing the bell, is there one that comes to mind that you're like, okay, when I think of public offerings, I remember that company in that moment during the

[Martin Abrahamson]

financial crisis, I was an analyst at Goldman Sachs. As you can imagine, the stock market was completely shut, and it was the largest US IPO ever at that time. Do you know which one it is? Financial services company. In the middle of the financial crisis goes Visa.

[Podcast Host]

Oh, okay.

[Martin Abrahamson]

Which is just this incredible story. And the reason why I was, like, sort of shocked that they could do it was everyone were just talking about, like, there'll be no IPOs and, like, the window is closed and all that stuff. And here we are, arguably one of the best businesses and business models out there who are ready to go public in the middle of the financial crisis at arguably the worst possible time to go public. And they basically said, don't care. We're going to go public anyway. And we're the largest ever IPO ever in the US to do it. And I think there's sort of a good, good sign of, like, yeah, when people say that the IPO market is closed, Right. That tends to refer to, like, the median company. Like, for most companies, it is closed. But if you're a Visa type of a company, you can go out even in a very, very difficult Time as to prove that they can do it

[CJ]

also sends a great signal about the confidence you have in your own business model.

[Martin Abrahamson]

Absolutely.

[CJ]

Why does that one stick with you when you think about Visa? Just because it's performed over time and it's continued to compound. Like, it's a pretty amazing story.

[Martin Abrahamson]

They went public when they were ready and when they wanted to. Not necessarily when, like the market said that this is like a good time to go public. I have to imagine that people that were advising them on the IPO might have said, hey, kind of choppy waters out there right now, like, maybe you should wait. And instead, I'm guessing Hevil's not in the room, but they said, you know what? No, we're going to go.

[CJ]

Do you think the benefits of going public are the same as they were maybe eight, 10 years ago?

[Martin Abrahamson]

I think they're quite different. So for starters, the private liquidity market is much, much bigger. I had someone do some stats for me ahead of this podcast and I think the sum of all the market caps of all the private companies is like 22 to 24 trillion dollars. So this market is now massive. I think the Russell 3000 is like 3 trillion dollars worth of market cap. So the difference between, you know, some of all, like the small and mid cap companies is dwarfed by the sum of the market cap of all these private companies. So I think the scale of the private market is now able to actually have companies now some of them are going public this year. Right. But like the SpaceX of the world would probably be more difficult for them to access the amount of capital that they need. And same thing with the LLMs, but that is no longer the case. I just remember like, you know, 2013, 2014 era when it was the capital war between Uber and Lyft, you know, and I remember Uber raised a billion dollars.

[CJ]

No one had ever heard of that.

[Martin Abrahamson]

Correct. It was just like so big. And then after that, I guess you had the SoftBank $100 billion fund, which was the elder, like, oh my God moment. This is an enormous amount of capital.

[CJ]

It was.

[Martin Abrahamson]

And Uber and you know, I think some Nvidia in there too, actually. So there were some of the good stuff.

[CJ]

Yeah.

[Martin Abrahamson]

But that sort of set, like the new, the new benchmark, and now you have these LLMs that are raising more than a hundred billion or, you know, close to it in a round, one round for one company. So I think the dynamics of the private market has changed a lot and it's now possible to be private for longer and actually access the amount of capital in the private markets whereas in the past it was simply not enough capital liquidity to do that.

[CJ]

Saudi Aramco in 2019 was the largest IPO to date. They raised $29 billion. OpenAI has broken that number twice in the private market, just 50 billion. And then I think like 105 or 100.

[Martin Abrahamson]

I think anthropic also exceeded that.

[CJ]

Yeah, and then you said 22 trillion. That's like, that's like 11 spacexes. These companies are some hungry hungry hippos when it comes to their capital intensity.

[Martin Abrahamson]

And then I think another thing that has changed is the possibility to do tender offer for employees and early investors which has also become more common. And I think back then that was the, that was less common. So I think you do have, you know, probably more tools at your disposal as a private company today than you would have had, you know, 10, 15 years ago. That being said, I am personally like a big bull on companies going public. I do think it's the right to do for companies that you know, have a big vision and are building generational companies. But I do think if you're a smaller company, the costs of being public is also quite high in terms of, you know, the disclosures, the compliance, the systems and everything you need to have in place, you know, aud becoming more expensive. So at some point that risk reward, trade off might not be the right thing for smaller companies.

[CJ]

In 2010 the average company that went public spent $150,000 on paper. There are also some benefits though around I think making, other than just making your stock liquid, the publicity that goes along with it, especially if you're selling to the enterprise.

[Martin Abrahamson]

And obviously this a little bit too of like companies that are asking us, hey, you know, if you'd like some certification of your, like your financial position. And it's easier for us because we've raised a fair amount of money. So you know, we can point them to like our, our Series E and our Series F fundraises as sort of evidence that we are well capitalized and we're not going to run out of money anytime soon. Whereas if you're a public company that's obviously much easier to do. Here's our last quarter. You can look at our cash balance, you can look at an auditor statement, etc. So some of those things I think for procurement teams at enterprises gives them a little bit better sleep at night.

[CJ]

Shout out to all the procurement teams working hard out there, always. Okay, me and you are back here at the New York stock exchange in 10 years are there going to be more publicly traded companies or less? And I ask that because 30 years ago there were 8,000 publicly traded companies. Today it's below 4,000 in the US. Is the number going to go up or down from 4,000? I think it'll go down okay, and

[Martin Abrahamson]

that is not because I think there would be fewer companies going public. I think there's still an opportunity for great companies to go public and they should. I think there will be more consolidation. I think it would be more private equity buyouts because it's less efficient for, you know, sub $2 billion public company to operate and there should probably be in private hands and I think there'll be more consolidation. I saw some other stat which was like that. The sum of the top 10 companies in the S&P 500 was 19% 10 years ago and today it's 38. So there's also a big move to concentrated positions at the top right. So there's more and more value in market cap accruing to the larger companies versus the longer tail of these indices.

[CJ]

Hey, thanks for listening. We'll be right back after a word from our sponsors.

[Podcast Host]

The CFO role has evolved faster than the tools built to support it. Most finance teams are still running infrastructure designed for a job that no longer exists. The reporting, the reconciling, the close that bleeds into the next month or quarter. That's not finance. That's overhead with a really crappy title. And the cost isn't just your time, it's everything your best people aren't doing while they're buried in it. Agentic finance shouldn't multiply your output. It should eliminate the work that was never worth doing in the first place. That's why I run mostly media on Brex, an intelligent finance platform with AI powered agents that capture expenses, automatically enforce policy before the spend happens, and close your books in minutes instead of weeks. 35,000 companies like OpenAI, Coinbase, Anthropic and Doordash already run on Brex. It's time to get Brex AF. Learn more at brex.com metrics I've seen a lot of FP&A tools and Aleph is one of the few that gave me that aha moment. Within minutes I remember watching their founder shout out to Albert, connect my netsuite data and build me a full pnl Live in minutes. Aleph is now trusted by hundreds of leading companies. I've had the CFOs from Turo, Eight, Sleep, Zapier and more on the pod and every one of them is a huge advocate. I also just published my second annual CFO Tech Stack Report and Aleph has been on the podium both years including a number one finish in the 50 to $100 million segment. This year, instead of being just another planning tool, they built a real enterprise grade data foundation for finance implemented at startup speed with AI native workflows woven into its DNA. All your systems erp, CRM, hrs, ats, product usage and more. Powering one clean governed data layer that finance can actually trust. With AI moving as fast as it is, they're pushing even further. Mcp, custom AI chatbots, AI powered variance analysis and the list keeps growing. Try it with your own data at get a left.com run that is G E T A L E P H.com run tell them CJ sent you. Here's a growth tax that nobody talks about Every new pricing model you ship creates a nightmare for your finance team. Team usage based pricing. Now you're tracking usage against commitments product bundles. Now you're untangling what to recognize and when for every line item. Mid cycle upgrades. Good luck Manually reallocating revenue. The pricing strategies that drive growth are the same ones that break your finance process. Right Rev turns that irony into a competitive advantage. Your product team can ship new pricing without asking finance for permission, and your sales team can close deals without worrying about downstream chaos. But I've seen too many companies where sales are celebrating a huge quarter while finance is still trying to figure out how to recognize half of it. It's actually me. So here's a good place to start. Right Rev built a free [e-mail] it scores your RevRec process, shows what's exposing you to risk, and tells you exactly where to focus before it bites you in the rear. Probably explains why your last close took so long. Check it out at calculator.wrightrev.com alright, without the Boston accent. Calculator.wrightrev.com we're warmed up here.

[CJ]

I think it's time to talk about AI adoption in finance because it's all a rage lately. Your team's at the center of creating an AI product for people to use. I also know that the finance team in particular is building some pretty cool stuff. What's something that your team has built recently?

[Martin Abrahamson]

Built a lot. So you know Vercel is an AI company and we also have our own tools to build AI features, AI products and to host them. So we, we are, we're really dogfooding a lot of this stuff ourselves. Vercel on Vercel, Vercel on Vercel. So I mean, for starters, like the thing we built first was a KPI dashboard, right? So rather than needing to use some looker or something like that, we built a really nice user interface with the exact metrics and the way we like to look at the business. Pulling data directly from Snowflake on my phone, it's like, you know, protected via SSO and that kind of stuff.

[CJ]

You built something that you can access mobile too?

[Martin Abrahamson]

Yeah. Wow.

[CJ]

Okay.

[Martin Abrahamson]

And now always at my fingertips, I can look at, you know, things that I care about. You know, like signups, new ARR, enterprise focus, you know, net dollar retention, everything is sort of available. Sometimes you hit and click refresh, changing

[CJ]

the button colors and stuff too.

[Martin Abrahamson]

It's amazing. So that that was something that we built first. And the second thing is as part of the accounting close process, we built a revenue flux by plan.

[CJ]

So glad you brought this up because I'm like a super dork. I went down a deep, deep YouTube rabbit hole and I found an old interview that you did and you said something about C606 module that you built.

[Martin Abrahamson]

That's nuts. It's actually pretty straightforward. Is it?

[CJ]

Take me through the flux analysis that your team has built.

[Martin Abrahamson]

We basically look at four main data sources. So we look at Netsuite, we look at our Snowflake database, we look at Slack and we look at notion to basically come together with a cohesive story of like, what actually changed from one month to the next.

[CJ]

Slack.

[Martin Abrahamson]

Slack.

[Podcast Host]

Okay.

[CJ]

I've never heard anybody bring that into a flux analysis. This is the next level.

[Martin Abrahamson]

So you're trying to explain, you know, why our enterprise business grew from X to Y, for example. And we have channels in our Slack that would be, you know, we have a list of all the deals won and Y. So if you're trying to look at why something went from X to Y, you need to have the context and information for that. So the agent will search all this data and, you know, explain exactly why the flux has changed. We assess it and sometimes like, oh, well, it doesn't. Didn't get that piece quite right. So it's not like we like click a button and like don't assess it, but it gives us a really good first draft. Wow.

[CJ]

Okay, so you've got the dashboard, also mobile optimized, you've got the flux analysis. I went down a deep, dark YouTube rabbit hole. I heard. I don't want to be apocryphal here that you also built a RevReck tool.

[Martin Abrahamson]

I actually built this one myself first, and then the RevReck team kind of laughed at me and they built a better one. But the way that it works is you take a dump of all of the committed contracts that were signed in the period and you upload it to a thing that we built and then it goes through and assesses each contract and even highlight for each one of things that might impact revenue recognition. Right. So if we get contract one, super vanilla, straightforward, and it would just be like, this one is easy to review, kind of like a green easy, and then goes to the next one and be like, this one's super complicated. Like they might have some unusual write or like a delayed start or something like that. So if it highlight that in red and then make sure that the team reviewing it mean, like, when you review this, make sure that you get this piece right. So it will draft the actual revenue recognition for you and highlight which contracts should be reviewed by human because there are some intricacies we should pay closer attention to. So again, it's not like we just had agents do all the work and the team is like, relaxing. It's that we can be much, much more efficient and we can have a much leaner team in revenue accounting as a result of these tools. They can spend all their time focusing on the stuff that is very challenging and difficult and requires human interpretation. Whereas if there is a contract that is very simple and straightforward to recognize, you know, we can utilize AI for that.

[CJ]

People talk about having a human in a loop. It sounds like you're actually using the tool to pull humans into the loop when needed.

[Martin Abrahamson]

Exactly. Like we have the agents highlight for us. Hey, you should pay extra attention to this one.

[CJ]

Yeah.

[Martin Abrahamson]

Because it has some unusual terms. All right.

[CJ]

Just to play devil's advocate here, Rev Rec is very important. Right. A lot of. A lot of things hinge on this. Why is this something that companies can feel comfortable building versus buying? Like, even if you can build, build it, do you think there are scenarios where you should buy something?

[Martin Abrahamson]

I mean, I think depends a little bit on the internal expertise that you have. But in this day and age, like, almost anyone can build these AI tools. Right. Like, I'm not and my team is not either, like, experts in, you know, coding or, like, dealing with command line interfaces, but we've made it so, like, these amazing tools out there, whether it's Codex or cloud code or, you know, Vercel's V0, that makes it much easier for People that are functional experts, like the revenue accounting team as an example, to actually build stuff to help their, their day to day.

[CJ]

That's pretty wild. So the whole finance and accounting team, everybody knows how to use V0? Yes.

[Martin Abrahamson]

That is sort of a requirement. It's part of our interviewing process and we encourage and celebrate everyone to. To use it. So we have like a monthly org meeting. So that's basically the entire team, everyone that reports up to my organization. We meet for an hour every month and like half the meeting is people giving demos of stuff that they build in AI to sort of, you know, show all the people, you know, hey, I built this. Here's how I did it. So next time you have a problem in your own field, like you can maybe try to automate that.

[Podcast Host]

How has AI changed how you hire

[CJ]

and evaluate talent on the finance team? You said it's part of your interview process. So like walk me through, because I've taken case studies that were pretty hard before. Yeah, I was in Excel and I had to also have a PowerPoint deck that went along with it. How are you testing for capabilities?

[Martin Abrahamson]

Two years ago, you know, a typical case on the finance team was like, build me a model or do some kind of analysis project something. Today that's pretty straightforward. To do the mechanical aspects in Excel, you can use a Claude plugin, or you can ask Gemini or really any model to build you something. So if you were to give that exact same case study today, you're really only testing their ability to use an LLM to spit out some information. So the case studies today are basically twofold. We try to assess their judgment. Right. So the case studies are much more about how would you assess X versus why. And then we tend to throw a curveball in there and be like, here's some new piece of information during the case study. How does this change your conclusion to what you just decided? And again, there's nothing with like a right answer. Everything is sort of judgment based. So you want to assess for judgment. And then part two is you have to build something using a Vercel AI tool. So you have to build like, for example, like a dashboard. You have to build an agent. You have to build something to show that you'd be excited about working at Vercel and automating a lot of your job away.

[CJ]

Speeding tickets versus parking tickets. I've heard you throw this around before. I got to know what that means.

[Martin Abrahamson]

It's one of my favorite analogies. It's basically bias for action. Right. So we deal with lots of uncertainty every single day. Vercel is a very fast moving company. We may not at all times have perfect direction for what every single person should do. What we don't want is people sitting still and waiting. That's a parking ticket. You sat still for too long.

[CJ]

You didn't move.

[Martin Abrahamson]

You exceeded the limit. You were too slow. A speeding ticket is you move too fast, right? In a chaotic world, you went and did something that perhaps you hadn't fully thought through or perhaps you hadn't like kept everyone in the loop. Obviously we always strive to communicate and do these other things, but for the most part we would rather have someone move too quickly on something, buy us for action than just sitting around and waiting.

[CJ]

Hey, thanks for listening. We'll be right back after a word from our sponsors.

[Podcast Host]

I got news for you. The ERP category is finally getting disrupted. And if you haven't heard of Rilla yet, please pay attention. It's the AI native ERP built specifically to replace netsuite, and it's already won over hundreds of finance teams. Their mission is to make these zero day close a reality. And they're actually doing it. We're talking teams closing the books at 1:35pm on the first day of the month. Companies like Windservice, Mercore and hundreds of others run their entire finance stack on really revenue recognition, close management, multi entity, native Stripe and Salesforce integrations. Whoo. Everything a scaling company needs. They've got 5.050 wow 50 stars on G2. They're backed by A16Z and Sequoia. Heard of them and CPA led implementations that get you live in 45 days. That is simply unheard of in the airplane space. If your books aren't running as fast as your business, check out Rillet. Book a [e-mail] CJ that is R I L L E T.com CJ that's me. Maybe the first thought you had when you started a tech company wasn't about scenario modeling or Runway management or the fine print of an S1.

[Martin Abrahamson]

I love a good S1.

[Podcast Host]

That's not where the magic happens. It's more fun to build something new, Design your quarter zip or network with VCs. But what if you got your financials right early, long before your investors started asking for it? There's no limit to how far you can go and how many headaches you could stave off. Now that's where EY comes in. EY has been a part of Silicon Valley since, well, it was just a valley and it's helped some of the most successful names in tech go from startup to exit to mega cap. You build the next big thing. And with the teams across strategy, tax, audit and transactions, EY will help you build your next big thing right. Learn [e-mail] TechStartups that's ey.com TechStartups SaaS and AI spend is heating up, and most finance and procurement teams are already searching for shade. Uber burned through its entire 2026 AI budget in four months. ServiceNow did the exact same thing. Your vendors are counting on you to sweat through your next renewal and pay more than you should. Cool down your SaaS and AI spend this summer with SpendHound. Through June 12, qualifying companies that sign up get $500 just for getting started. Teams using Spentown reduce software and AI spend by up to 30%, using real pricing benchmarks from over 1,200 companies across 10,000 AI and SaaS vendors. So when your next renewal hits, you're not taking the heat. You know exactly what fair pricing looks like and what you should negotiate for. Spendhoud also includes on demand procurement experts who build negotiation strategies, review contracts, and ghostwrite emails on your behalf. That means you can stop getting burned by the tools you need and cut the ones that you don't before that auto renewal torches your budget. Rated number one on G2 in SaaS spend management, SpendHound is free forever. The team's up to 1,000 employees and only $10,000 per year for enterprise teams with $150,000 in savings guaranteed. Don't let your vendors enjoy your summer budget more than you. You do sign up by June 12th and get $500. How's that for a hot deal? Go to spendhound.com cj that's spendhound.com cj

[CJ]

asking for forgiveness instead of permission. Yeah, I, I try to instill that in a lot of the finance teams that I've run before. And some people are like, well it's the finance and accounting team. Shouldn't you always be, you know, going by the rules there? It's like, well, if you're going to be setting the pace for the rest of the org, I think you do have to take it upon yourself to show that you're willing to move fast.

[Martin Abrahamson]

Yes. This goes back to another sort of belief that both I and the rest of the company have, which is there are actually very few one way door decisions.

[CJ]

Oh, say more about this. I love this.

[Martin Abrahamson]

So you know, when people evaluate like should we do X or Y they tend to view it as absolute. Like if I pick this door I'm in in this path forever. I cannot like turn around and go through the other door. And there are actually exceptionally few one way doors out there. It might seem like a big decision, you know, like signing up for a real estate lease or something like that. If you don't like it, you could probably find someone to sublease it. Like it will cost you some money and it's obviously not ideal. And you know, it's definitely something you want to spend time making sure you get right. But you don't want to be paralyzed by making decisions that you think are one way doors but are actually not. So try something out. And this is, you know, let's ship a new product, let's learn and improve it as opposed to let's just wait and see.

[CJ]

I want to talk about running a company in hypergrowth here and how you possibly plan for this. Right. Like your job is capital allocation, it's resource allocation. Traditional planning assumes some sort of stability. You don't have that in a good way.

[Martin Abrahamson]

Yeah.

[CJ]

How do you wrap your arms around that?

[Martin Abrahamson]

Every year we're basically off plan almost immediately, which is a good problem to have. And if my board is listening to this, they will laugh because we kept moving the plan around last year.

[CJ]

February.

[Martin Abrahamson]

Yeah. You know, first month out of the gate, we did better than we had expected. That has sort of been true. But it is very, very different. Difficult when you're growing at high percentage growth rates to accurately forecast what it will be. So in the past we've operated just with a single model that we kept updating, but it became a little bit more like a what's the weather forecast for tomorrow? As opposed to what is actually helpful for a longer term horizon. So this year we're operating actually with three different models and they're, they're more like growth charts for your children. Like what percentile are they on? And then each plan has an expected end result and then depending on which sort of line you're on, and the top one is like extremely ambitious plan. We track which plan we're tracking towards based on this monthly plan and then we're spending accordingly. Right. So you might say like, you know, you sort of start out at the average plan and you can be like, okay, we're tracking a little bit above, but we're not quite at like the extreme plan yet. So we're not ready to sort of fully fund the initiatives that we had in that plan yet. Or fingers crossed, knock on Wood. But if we end up going in the other direction or we're tracking a little bit behind, you can adjust accordingly. So we kind of know the bookend and we're adjusting where we are inside of that.

[CJ]

And so it sounds like it's a number of standard deviations and it's am I riding that railway or this one?

[Martin Abrahamson]

Correct. Can kind of hop between. Yeah, yeah.

[CJ]

I often joke with people that this is the podcast where we describe charts but or in audio format and people like it. But how do you communicate that, though, internally and also with your board? Like, does everybody know that we're on this trajectory right now? We're on path A. We thought we were going to be on path B. Luckily, we're on this one. How do you talk about that?

[Martin Abrahamson]

Yeah, I'm being extremely honest with my board. It's exceptionally difficult to forecast this business 12 months out. And while also from the outside, I'm sure, you know, other CFOs in hypergrowth company feel the same way of. On the outside, it might seem like every single day is sort of like, amazing up until the right. And there's this consistency of it, but it is not. It is almost like a, you know, on a rowboat or something like that, where there's like a big push and you move forward and then like suddenly, like two weeks where like, signups were a little slower, and then suddenly there's like another big push. Like, it is absolutely not linear. It is also not perfectly exponential. So you're dealing with this sort of like, up and down, back and forth, generally positive trending. But when you live this stuff day to day, you find yourself sort of oscillating between, like, communicating how amazing the company is doing and then sitting in sort of an internal yellow flag meeting on like, oh, something is not quite right here. And then the next week you'd be like, oh, never mind. It sort of like, I assume that

[CJ]

you have to broker the budget with your department leaders depending on which trajectory you're on. How do you discuss this with them? Hey, you know, we've hit these targets. That allows us to unlock X, Y

[Martin Abrahamson]

or Z. Yeah, I mean, we focus a lot of, like, efficient growth. Right. So we have view of the CFO as his, like, mean person that, like, says no to everything. And then like, making an example of someone like our CMO being this, like, crazy cowboy that just want to spend money. Right. Like, in reality, that's not how it is. We are kind of sitting at the same side of the table looking at the business and Being like, hey, you know, we're seeing this. If you want to invest this, what are your expectations for actually driving this return? If you want to like access incremental funding for, you know, another event or like additional marketing spend, do you think you can deliver? Why? And like you have a discussion, does that actually reasonable or not? And sometimes you're like, yeah, we feel good about this. And other times you're like, yeah, you know what, I think we should like hold back. And they tend to come to more of like a agreement than a like one side being like, absolutely not. Over my dead body. I found people tend to like, want to make good decisions.

[CJ]

They do. They want to be good stewards of capital. What's interesting is when you're at an organization in hyper growth, what you encounter a lot of times is people have had varying experiences with having a budget in the past. Some people may be as sophisticated as like, oh, I'm a CMO and I know that you're going to allocate rent something I can't even control to my budget. Okay. Then other times it's an engineering leader who's never even managed a travel budget. I'm going to make the assumption that you have conversations and you often have to meet the leader where they are in their sophistication with the budget and help bring them along the way.

[Martin Abrahamson]

We do. And one of the things that have been a better job at this year is we have a new FP and a leader who's, who's great. But in addition to sort of where you are relative to budget, we have slotted in like, this is where we expect you to be relative to budget for the next few months. So sometimes, you know, there's a big engineering off site next month, right? So you might look good on your T and E budget right now, but we know you're going to have a large expense next month that has not hit your budget yet. So we're trying to sort of give people sort of visibility into also what's coming to help them make these decisions. And there's also learning here, right? Like if we move from a world where, you know, we didn't have super detailed budgets for like what software you want to buy. And then you started asking engineering leaders like, what are you going to buy this year? And then they come to you three months later and be like, oh, I want to buy this tool. But you didn't tell us that three months ago. But the world is changing, right? It's like, yes, that was true, but it wasn't an oversight. It is just now we have a new product, we have new needs. You also need to bake in some flexibility to work with them on. Like, okay, well, we have some budget that we can allocate. Are there perhaps some other tools that you had thought you needed three months ago that maybe you no longer need or maybe you can build yourself? So it is constantly evolving.

[CJ]

A lot of this is the art of asking the right questions because people will come to you with what they think is the problem or the need. But there's often something like three derivatives behind that of what they're actually asking for.

[Martin Abrahamson]

All the people are like trying to pretend they're in finance and they have like these ideas for how to like do something. It's like, oh, that's not quite how revenue recognition works. Or like, yes, that's, you know, we don't do like cash expenses. Right. It's, it's accruals. So like you have to sort of allocate this bonus one hit.

[CJ]

I do want to talk about your business model, right. I'm a huge business model dork and Vercel is famously both PLG and consumption based. Right. Two of my favorite things to talk about also can be complicated in a way when you're embracing consumption based revenue. How does that change how you measure and communicate performance?

[Martin Abrahamson]

Yeah, so a lot to sort of unpack there, right? Because we have the, in the PLG business, it's month to month, right. It's customers with a credit card. That could be anyone listening to this podcast. We have to rely on data models to sort of estimate what is a this cohort of people, how is their forecast going to look over time. And the best leading indicator for like what this cohort will generate over time is literally the size of the cohort. So you look at, okay, in this period, we sign up this many people and then based on month one usage, you know, we have a bunch of different data models. We try to segment users. You might be like a hobbyist or something like that, that, you know, relatively low spend. So we think, okay, you're only going to use like the base rate like to $20 per month currently. So that will sort of what we project.

[CJ]

I am on the $20 plan.

[Martin Abrahamson]

Great, thank you for, for your support. And then we might have other people that might have, you know, a team that has signed up and they're heavy users of infrastructure and we will have a different expected growth curve for that population. There hasn't really been a consumption PLG self serve business in the Public market yet?

[CJ]

No, I was trying to think of one today and I couldn't think of a great example.

[Martin Abrahamson]

Self serve businesses on like Slack and Dropbox are sort of like the first two that comes to mind. But they were all seat based, right? They were all subscription based. They didn't have the consumption piece. So yeah, that's one thing that I think will be unique with the Vercel story.

[CJ]

Definitely a benchmarking conundrum that you probably run into to, because I'm the type of person when I was trying to speak to investors is I would always look in the public markets and I would create my own cohorts of who I thought we were. Like I worked at a company, it was a vertical software for the automotive industry. Okay. So we can probably look at Samsara. We're not nearly as big as Samsara, but there's some things we can take there. Oh, but it has a marketplace element to it. So how do you blend all these different things to come up with something that looks and feels like a company like yours? And I think what you're doing with your business model is brilliant because you're able to take advantage of the high velocity sales, the people who are maybe consuming at the lower end and may bring it from like a prosumer angle into their company. And so you have multiple angles for expansion.

[Martin Abrahamson]

Correct. And we see this, you know, when people either graduate people using it on like the personal side and now at work they also want to use Vercel. We know that this PLG motion is heavily influencing the enterprise sale as well.

[CJ]

So what does Martin look at when he shows up at work? He opens his dashboard, the one that you created. What's the first metric you're looking at to gauge the health of the business?

[Martin Abrahamson]

So if we look at basic consumption, so we have a daily consumption dashboard which basically looks at how much infrastructure all of our customers are using. When you click in on that, you start looking at, well, where is it coming from? Like is it coming from new users or existing users? Is it coming from, you know, committed customers? Like what I think a normal company would call enterprise. Right. Where they've made a larger upfront commitment that they're spending down versus is it coming from the self serve tier? And then you look at new signups and you look at the amount of new commits for, for larger customers.

[CJ]

I was interviewing Brad Flooring, the SVP of finance at Snowflake, and he had built their original forecasting model back in the day with Mike Scarpelli. And I'D ask him the question, like, must be great. When you look at RPO each month, he's like, no, no. RPO is actually not a good metric to gauge our business's health. Is RPO important for you?

[Martin Abrahamson]

It is not. I really need to get my talk track together on why RPO is not best for Vercel because I feel like I'll be saying this a lot.

[CJ]

It's good for the analysts, though. They want it.

[Martin Abrahamson]

It's actually not so good for us either for the analysts either to assess our business because, you know, the majority of our revenue is coming from customers paying us with a credit card.

[CJ]

Yeah.

[Martin Abrahamson]

Majority of revenue is PLG revenue, which basically has no RPO.

[CJ]

There's not much RPO on my $20.

[Martin Abrahamson]

Correct. You know, every month we have to earn it, which is sort of a good spot to be in as well. We have to continue to innovate, we have to continue to earn it. But if you look at like our current rpo, which you do have to disclose in our financial statements, it does not give you a whole lot of insights into what next year will look like. As I think about like, you know, sort of like deferred revenue and backlog type businesses, Airbus, Boeing and like big shipyards would have this like, oh, you know, like many, many billions of backlog. And you knew that some percentage of that backlog would turn into revenue for

[CJ]

the year those planes will eventually become revenue.

[Martin Abrahamson]

Yeah. This is not how business works at all. We have to earn it every single month. Rpo. If you just look at our current RPO and our current crpo, it's a small percentage of what we expect revenue to be this year.

[CJ]

And I'm assuming there's cool quirks and seasonality too, when you're consumption based, even based on the days of the week or what month you're entering and like holiday season versus, I don't know, mid summer.

[Martin Abrahamson]

Yeah. You know, this is actually a funny one, which is traditionally, yes, there is more vercel usage during the week than weekends. And historically there's been less vercel usage during holidays. Last year that changed. So over both Thanksgiving and Christmas, we saw an increase in individuals using it. And it was sort of like the first time that we saw a mass increase of hobbyists starting to use AI tools. So people had time off for the holidays for the first time. They're like, hey, you know what? I've heard so much about quad code or V0 or codex or whatever it is, I want to start playing with it. So People use downtime to actually work with these tools and then signing up to become customers. 2025 calendar was like the first time they saw like a inversion in that trend, which is typically like less volume, fewer deployments and stuff over a holiday break.

[CJ]

If you keep growing at this pace, you're going to be almost a macro indicator for productivity.

[Martin Abrahamson]

Pretty funny to do actually.

[CJ]

When I think about go to market motions. It is neat that someone like me could try your product and then say, you know what, I got a lot of usage out of this. Say I'm a cpo, the product team should start using this. It's hard for Slack to do that, right? Like me and my wife aren't using Slack. And then I can say, let me bring this into the Org. You have a path from consumer or prosumer into business use cases.

[Martin Abrahamson]

So a few things that have actually changed that's interesting with the business. When I joined Vercel almost three years ago, most of the people on the credit card plan were teams, right? They were individuals that were collaborating because back then it was hard to build an application or really anything alone.

[CJ]

So they were already in multiplayer mode.

[Martin Abrahamson]

They were already in multiplayer mode because, you know, they had like a designer and a front end person and a backend person and like needed to work together on this project. And today that has shift very much to the solo developer which might have one area expertise. They might even not be a developer. They might be, you know, someone like you and me that are not developers by training, that are building stuff using AI. And this is also true as it leads into the enterprise because people are now making decisions in the enterprise of what tools should we use, what infrastructure should we use. In the past, you know, they were relying on input from engineers and now they can actually test out these tools themselves. They can get a feel for what the product is, whereas that was not really possible in the past.

[CJ]

I ask CFOs this all the time, how do you think about TAM? We were talking to Figma CFO a couple of weeks ago and he was saying that 2/3 of their users are not designers, right? And so he tries to track this in the cohort data. But to me, like, it breaks my brain to think how many people outside of like the core ICP could potentially be monetizable users.

[Martin Abrahamson]

Yeah, our ICP has changed completely. Three, four years ago, like the core Versailles ICP was someone with like a GitHub account that was, you know, had every single week filled and lots of, lots of pull requests and commits. We have those customers too still, of course. But now we have many customers that you know, have almost inactive GitHub accounts and they might have brand new GitHub accounts. So that's been like a big, big change. So I do think our sort of addressable number of whether it's measured in people or dollars have meaningfully expanded both because we can attract users that are not very technical and they can still use and benefit from all these tools. And it's also much easier to build software, right? So in the past you have to spend your money to purchase a productivity tool, but today, you know, you can build one. Like if you have, hey, I want to build like a packing list for my business travel so that I don't forget stuff when I come to New York, for example. And then I can easily build that using, you know, V0 or some other tool and I can deploy it on Vercel. And in the past I probably have to pay a couple of dollars for some like to do list app or something to do the same. And now I just built it my own. So you have an opportunity to capture some of that software spend shout out

[CJ]

to LaGuardia Airport and the new shops they have in there. Because I've forgotten about twice on business trips recently.

[Martin Abrahamson]

I'll build you a tool for this.

[CJ]

The use cases can't even keep track. I want to talk about creating enterprise value. What's your mindset on capital allocation?

[Martin Abrahamson]

How, how do you know if you

[CJ]

were over or underinvesting in this massive opportunity that you have?

[Martin Abrahamson]

Yeah, so this is a topic that's really challenging today, right, because the market is moving so fast that as we talked about earlier, even like the, the best plans are stale pretty quickly. And the same is true for like capital allocation. You have to make decisions with perhaps less complete data than like ever before. And what we instead do instead is like we try to measure quickly and if something is not working out, we iterate. An example of this would be like in marketing, like we've made historically investments on big events. This year we're going one step further and we're actually doing events all around the world. Events historically has been US centric. The bet for this year is, you know, we're taking sort of Vercella ship on the road, more regional events. And the expectation coming out of this is like, we should see an increase in pipeline generation globally. So this year is the task for like, does having events in like, you know, Sydney and London, et cetera, help drive pipeline in those regions? And then we will Reassess.

[CJ]

Take me into that. Then you're talking with the marketing team and saying, I understand we want to allocate this money that was previously for one maybe just major user conference. We're going to go into these smaller regional areas and try to promote. Are we going to be able to come back and look at the metrics within these areas to see if it works?

[Martin Abrahamson]

Yeah. So we sort of have a what you need to believe type model. Like we're going to spend this amount of money Here would be our expectation of like pipeline generation. Now, sort of marketing famously is hard to allocate. Exactly. Like, what was the impact of attribution? Attribution always challenging. So we have to look at it somewhat like holistically of like, did we actually see sort of evidence of a boost in the region as a result of this? Now, obviously you can track it very closely. If it is someone from an enterprise company attended the conference and had an executive briefing with like our CTO and then became a customer. That's obviously very, very easy. You also have the benefit of like, if you're doing an event in London, people might talk about it and that might help influence and we'll never capture that attribution.

[Podcast Host]

No.

[CJ]

But I think a rising tide lifts all ships and you can kind of feel it in the rhythm of the business. Yeah.

[Martin Abrahamson]

So there is sort of the combination of some science and some art and some belief.

[CJ]

Well, you've argued that growth compounds value in a way that margin improvement can't. I thought that was a really succinct statement. There's a lot in there. Can you unpack it for me?

[Martin Abrahamson]

All right, where to begin? So on growth we're in a market is moving incredibly fast. Capturing more market share will give you the opportunity to do more things and to get a foothold with customers and become the de facto standard. I think that will compound and give you more benefits over time. We even see it with our own customers that may have started on one resell product and are now using multiple products because they had a good experience with the first product. As long as we continue to make improvements and ship new products that are delighting customers and adding value, I think there is a lot of value of having a relationship with that customer. If you optimize for margin too early, you're likely going to miss out on that growth curve. Now, I'm not saying don't have any gross margins, but find the optimal balance between the two and don't make short term decisions that will reduce your growth in favor of margin when you still have the potential to massively accelerate, how

[CJ]

do you know to keep accelerating into that? What are the signals?

[Martin Abrahamson]

Are we able to like maintain or accelerate our growth rate? Right.

[CJ]

Are you growing faster than you were before at a larger scale?

[Martin Abrahamson]

Correct. So that's like one obvious sign that like that'll hit you over the head. You're doing well. Yeah. And we obviously have some like guidance metrics as well of like do we have a healthy cash balance? Which you do. Relative to burn. Which we do. So we do look at sort of that from more holistic basis. But I do think we're in a market that the opportunity to sort of build a very large business in AI infrastructure is happening right now. And if we don't have like the biggest bucket and we possibly can to capture this wave, someone else will.

[CJ]

I like how you brought the competitive element into it as well. Like someone is going to take advantage of this opportunity. It's not just about our gross margin or profit margin. It's about in the macro context, do we have the right to win?

[Martin Abrahamson]

Correct. I'm going to say something slightly spicy. I think the SaaS industry or software industry has sort of gotten used to like oh, these like software companies should have 80 plus percent or 90% gross margins. And if you sort of take a step back and like what does like 80% gross margin mean in practice? It's like a, basically means a 5x markup on the cost to serve to a customer. And in the past it hasn't been that expensive to serve these customers with like good application tools, et cetera. Now that we have AI and more compute and personalized experiences, it is going to be more expensive to serve customers. And I think this is not going to just make it automatically more expensive for everyone to have these products. I think structurally there's going to be a margin reduction. Particularly see it for SaaS companies that are now being disrupted. Prices will have to come in, come into account and look at should I build something myself or should I buy something? If you're competing with the opportunity to build something that's 80% as good directly versus buying a off the shelf tool that might not even be perfect for you, you now have a new competitor which is build yourself with AI.

[CJ]

What's different about cash burning AI companies in 2026 versus the cash incinerating companies of 2021?

[Martin Abrahamson]

A lot of the value for businesses are in the terminal value. How much value will this deliver long term? And I think in 2021 I think we were perhaps A little relaxed on how much value these companies would generate in terminal value. Obviously low interest rate environment at the time and we assign a lot of value in that terminal growth rate driver for these businesses.

[CJ]

Bucket was chunky.

[Martin Abrahamson]

Very, very chunky. I think as you see it today, you don't have to be particularly creative to imagine a world where AI is creating an enormous amount of value for individuals, for businesses, for the world. I think medicine. So I think there is likely to be meaningfully more substance in that terminal value growth rate driver today than it was in 2021.

[CJ]

Every public company out there today in the investor presentation has this chart. Right? It's a P and L and it shows. This is what our margin profile looks like today as a percentage of revenue. This is what we think it is in three years. This is what we think long term. Do you think in the future the gross margin will be lower, but the profitability will be able to be the same or higher?

[Martin Abrahamson]

Like how does that possibly. Yeah, so you know, if you can be more efficient with your opex by perhaps having more efficient employees, you should be able to drive more revenue. Yeah, perhaps with slightly lower gross margin, but still very healthy EBITDA margins. Right. So in the past, what was gross margin really? A. Why did people care so much? It was effectively a metric or barometer if you wish for how much value is there in this business? Like what is the premium someone is willing to pay for this service? And I think some of that value is going to shift as a result of AI for the value of the service that is being offered will actually go to the LLMs and shift over in that direction and you're going to recapture some of that value and increased productivity.

[CJ]

The greatest lie the devil ever told was the assumption that if you have this gross margin, you could technically turn off the G and A and the sales and the R and D and they were all stick around and in perpetuity. We could serve just this customer base at that. But we all know the customers care about the roadmap. They care about what you're building down the line.

[Martin Abrahamson]

Take a step back and like very basic sort of finance. One, what is like the thing that ultimately drives value is like long term cash flows. Yeah, right. In the very, very long term it shouldn't theoretically matter how you got. Let's pick a cool number like $10 billion in free cash flow in a year. If that is coming from a business that does a trillion dollars in revenue at 50% gross margins with some opex and you're generating $10 billion of free cash flow, that's going to have some value. And now you've maximized for like insane amount of revenue. Perhaps it's somewhat disappointing gross margins. We have a more efficient OPEX structure and in other world we're like oh no, we were like really focused on high gross margins. And now you have a much smaller business by scale. More people are sort of making the trade off that I would rather build something myself versus using your service. You should be optimizing for gross margin dollars and then you should be optimizing for EBITDA or free cash flow dollars.

[CJ]

Do you think that there are companies out there who are growing really fast but they're potentially destroying value?

[Martin Abrahamson]

I mean I think if you're a gross margin negative company and you are effectively selling dollar bills for 90 cents even if you're capturing a lot of market share.

[CJ]

Doordash and Uber figured it out. I guess pets.com and webvan didn't.

[Martin Abrahamson]

Now if I believe of somehow you can like turn it around, it might be healthy to get it to scale. Long ago, right, they had a lot of fixed cost infrastructure that they needed to sort of amortize billion dollars in infrastructure. So in that case, yeah, maybe you do need to get to scale first before you can become profitable. But I think if you have high variable expenses and there are no scale benefits, probably more challenging to create a high value business as a result.

[CJ]

How much do you care about unit economics at the stage you're at?

[Martin Abrahamson]

We still look very closely at unit economics. Not necessarily because we need every customer to deliver X, but it's a really good way to sort of assess the health of the business, right? So like you track churn rates, you track sort of increases in spend over time like all the components and you track how expensive was it to acquire. All the components of unit economics are really good metrics to track. Hey, are things getting better? Are things getting worse so you can fix it? So for me it's like less about like the absolute dollar, even like a ratio. I look at the components and how they're trending because I want all the components to trend positively. And if we're starting to see some deterioration among the metrics, then we know where to sort of dig in and focus.

[CJ]

I think that's what's lost in a lot of the conversations about unit economics with companies that are maybe at a lower gross margin than we're used to seeing or aren't profitable yet. It's not that you're not tracking them, you're Just caring about the direction and the velocity and the pace of change.

[Martin Abrahamson]

Yeah. And going back, you look at all the food delivery businesses, they were not very profitable at small scale. And even a percentage basis they're not particularly profitable. But when you add in a ton of volume, even at lower percentage margins, you generate a ton of dollars. I saw on a New York Stock exchange floor today, I saw Uber did 3.6 billion rides in Q1 was literally just on the screen there. Each ride not that expensive and like their take on each ride not that when you do it at this massive scale, like I think you build a pretty good business.

[CJ]

Yeah, we've become an instacart family over time. I don't know if that's cool to admit or not, but I don't really like the grocery stores. I joke all the time that like delivering one watermelon on a bicycle, not great unit economics. You deliver 10 million watermelons on multiple bicycles. You can get to scale over time and make the math work, but you got to get to a pretty big number.

[Martin Abrahamson]

And this is I think some of the physical delivery or transportation business. This has really proven out that they can be very, very good businesses at scale and I think big instacart family here as well. I was looking at the bill the other day, I was like, wow, I eat a lot.

[CJ]

Martin, I'm going to take you into what we call our long ass lightning round. You're a successful guy, Vercel is doing well. But you got to give me one thing. You've messed up on the job before.

[Martin Abrahamson]

So one of them was actually my first job as an intern. A long, long time ago I was working at Royal bank of Scotland doing residential mortgage backed securities Pre financial crisis 2006. So I had a little bit to a little bit of a contribution to that. We were going to Ireland, to Dublin, we from London to see Allied Irish Bank. The day before they were like actually also got a meeting with the bank of Ireland. Can you put together like a briefing pack so we can show up for that meeting? Really prepared, internal use only. But like everyone needs to like read up on it before the meeting. And I was like great. So I went to the Central bank of Ireland, pulled all sorts of academic research and everything created this really, really good binder. Turns out bank of Ireland is actually like a business bank in Ireland. It is not at all the central bank. So the entire team showed up with like reading a briefing pack on something that was on a completely different topic that they were.

[CJ]

That happens. I mean it's all in the name.

[Martin Abrahamson]

Absolutely. So the central bank is actually called the Central bank of Ireland or something like that as opposed to the bank of Ireland. I think more recently at Vercel, one of the mistakes that I made is we're growing really fast. We're like setting up some systems like oh wait, but we're going to do like this thing like manually because right now we don't have the resources to like automate it. But what if we get to scale and this thing doubles? That's going to be insanely painful. All that'd be like a good problem to have if we get there. But I think I should have been much better early on and be like, no, we're going to fix these things right now before we get there. Because trust me, it's not a good problem to have when we get there. Because on top of needing to deal with the problem you could have solved a long time ago, like we do some manual building stuff today for like very custom enterprise customers. We have all these other problems we're dealing with. Right. So you're like compounding future problems that you could have solved today. So if I hear anyone say like, oh, we're making this like short term thing right now and that'll be a good problem to have. That's like my immediate like, no, we're solving this thing right now. We don't want to like basically punt the problem to the future.

[CJ]

Can you walk me through your finance software stack? What tools does your team use to get the job done?

[Martin Abrahamson]

Yeah, so obviously like we use NetSuite Zip for procurement. We use Flowcast for close ramp for corporate cards. We also use Brex actually for a different use case. We're big fans of both of them. Both are on Vercel and Good Partners. And then we use Avalara for sales tax. And then we build a bunch of stuff ourselves, like sort of like socks compliance. We're like build our own thing for like deal desks have their own dashboard that they're tracking deals. Procurement built their own tool for contracts coming up for renewal.

[CJ]

It's on V0. Last one I got for you. What's the craziest thing you've ever had someone try to expend?

[Martin Abrahamson]

I once had someone try to expense an oven. And the reason for this, which might looked ridiculous, it's still ridiculous in retrospect. But we were offering some employees UberEats credits when they had worked late and this person, you know, was like a health fanatic and like didn't want to eat takeout food so felt it was appropriate for this individual to instead expense an oven. This was not in fact an expense.

[CJ]

Was that a speeding ticket or a parking ticket?

[Martin Abrahamson]

That would be a speeding ticket of moving too fast.

[CJ]

Martin, thank you so much for making the time for me today.

[Martin Abrahamson]

Thank you for having me. This was awesome.

[CJ]

That's pretty cool right?

[Martin Abrahamson]

Very very cool.

[Podcast Host]

Run the Numbers is a mostly media production, yelling and intro by Fat Joe Artwork by Meg Delesandro.

[CJ]

Show is executive produced by Ben Hillman.

[Podcast Host]

Nothing said on this podcast is intended to be business or investment advice. It's the sole opinion of me. A guy who feeds his dog way too much ice cream and has a history of net operating losses. Lol. If you like this podcast us hit subscribe and give us five stars. It will take like two seconds and our algorithm overlords love it. Drink water, call your mom and have a great day.

[Martin Abrahamson]

Peace.