Hype, or excessive exuberance, is one of the most reliable forces in markets, even if it wears a different costume every few years.

Overhyped IPOs aren’t new, obviously, but what’s interesting is what happens after the party ends and the stock has to earn its keep.

The market has a funny way of throwing growing businesses into the penalty box simply because they were priced ambitiously, and sometimes, that’s the beginning of a long and deserved re-rating downward, but sometimes it’s the setup for something more interesting, which is a great product with a great runway that got stuffed into an unreasonable valuation at the worst possible time, and then discarded when sentiment flipped.

If you recall our discussion of Airbnb from last year, you may remember that we expressed similar surprise that the stock, six years later, still trades below its IPO price while operating profit has 6x’d in the meantime and share count has begun to shrink due to buybacks…

Figma’s post-IPO condemnation by markets hasn’t been as extreme in its disconnect from fundamentals or as prolonged as Airbnb’s, but with Figma down nearly 80% since its IPO, despite otherwise flexing incredibly impressive growth, I couldn’t help but wonder whether intrinsic value was being neglected in favor of fear & panic.

— Shawn

Figma is a useful case study right now if you’re trying to make sense of what’s happening across software more broadly. We’re living through a moment when investors are asking whether “SaaS-mageddon” is a generational opportunity or a value trap, because LLMs are lowering the barrier to software creation, which has the potential to compress the moats of businesses that once looked like some of the highest-quality enterprises in the history of capitalism.

Adobe has been at the center of this debate and has been discussed at length in our Intrinsic Value Community, because Adobe sits at the intersection of creative tooling for human professionals and AI-generated content — it’s also one of our larger Portfolio positions.

But to understand the offensive side of that story, it helps to study one of Adobe’s most meaningful emerging competitors, which is Figma, a company that had one of the hottest IPOs in recent memory in August of last year, and then proceeded to fall about 80% from that peak, as mentioned in the intro, leaving it at roughly a $12 billion market cap.

That’s around 60% of what Adobe offered to pay in their attempted acquisition of the company that was blocked by regulators (and yes, Figma walked away with a billion-dollar termination fee — 3x as much funding as they’d ever raised before, while Adobe got nothing.

If you’ve never used Figma, the simplest description is probably still the best one: It’s like the Google Docs equivalent for design. Designers can work on the same file in real time, from anywhere, with far fewer workflow bottlenecks than the old world of siloed local files.

I’ve actually used Figma to communicate design requests and additions to our company’s site, theinvestorspodcast.com, with colleagues, and honestly, it’s a very slick interface. If you’ve never worked in design (not claiming I have extensive experience by any means), it’s easy to overlook how useful real-time collaboration at the file level is, when, not that long ago, institutional design work was a much less glamorous, time-intensive, and fragmented process.

The closest comparison I can make is to imagine trying to write a book with a team of other writers, who are all making edits and additions saved locally to their computer, with no single unified version existing across everyone’s local devices, as extensive time is spent trying to track what the “latest” version is and keep that distributed to everyone. It sounds like an absolute nightmare, and yet, it’s not that far off from how things used to be, and may still be, in pockets of the commercial design world.

On top of that, you also have product managers and engineers trying to guide the design vision, or bring the vision to life, who want to see the latest versions of the file. When no such unified cloud version of the design exists, well, that only adds to the complications, and things become even more challenging when you consider that, unlike with writing, where almost everyone has access to, say, Microsoft Word and is familiar with it, design files are instead saved as iterations in much more complex pieces of software like Photoshop.

Two questions emerge: Do you pay for a costly Photoshop license just so engineers can download and access design drafts in a platform they’re not trained to use (and wouldn’t otherwise use)? And again, how do you stay on top of the latest file version as the number of stakeholders interested in design updates extends well beyond the actual design team?

This is the world we lived in before the Cloud, and even after technologies like Google Docs were rolled out, the design world largely remained Luddites, to Dylan Field’s chagrin.

Who’s Dylan Field? He’s the Silicon Valley wunderkind who co-founded Figma and runs the company as its CEO to this day.

Field was working as an intern at the social magazine, Flipboard, in the early 2010s. He kept getting derailed by file-sync issues in Adobe Fireworks, spurring frustration over why design work didn’t feel as smooth and collaborative as Google Docs, which he grew up using in school.

In that spirit, he realized that the best tools are those that don’t require everyone to have the same paid license, the same software installed, or the same level of expertise just to participate. What Field noticed was that schools had discovered the power of collaborative work platforms, while the corporate world lagged behind.

As the story goes, Field routinely found himself sitting with designers and watching their screens light up with a nonstop stream of Dropbox notifications. That’s because teams saved files in shared folders, so every revision created a new notification, and every notification implied someone now had to figure out whether the “latest” file was actually the latest file, and whether feedback was embedded in an email thread, a Slack message, or a markup that didn’t sync cleanly back into the source design.

It’s funny now because it’s hard to imagine how anyone dealt with that reality, but that was the reality, and until someone builds a better alternative, people don’t complain about the friction they’ve normalized. It’s pretty cliché, but Figma reflects the quintessential story of an annoyed user looking around and saying, “How does [XYZ improvement] not exist yet?” and then building it themselves, as Dylan Field did. (Well, actually, Field didn’t so much build it as “Computer Jesus” did — more on that in a minute).

In most cases, these realizations lead to nothing, but every so often, they’re the inspiration behind a multi-billion-dollar business!

After raising capital from a Peter Thiel-backed fund while at the Ivy League school, Brown, it took about three years of painstaking work just to release the first version of Figma’s product. In hindsight, Field has called this a mistake born from a drive for perfectionism. Ideally, they would’ve published the product sooner and started iterating on external feedback / hiring faster at an earlier stage, but ultimately, the product was so good that the delayed rollout clearly didn’t derail the business in the long run.

Field, as alluded to, wasn’t completely alone, though. He had by his side Evan Wallace, who was nicknamed “CJ” — short for “Computer Jesus”. 

Wallace was a teaching assistant in a course that Field took. As they got to know each other, at some point, Field asked Wallace whether he, or I mean “Computer Jesus”, would like to start a company with him. 

This was actually before Field had the revelation that he wanted to change how design work is done. But Field was so impressed by Wallace that his grand idea was literally just to build any business he could with him, because working together with Wallace (CJ) would surely lead to greatness. Wallace didn’t disappoint, though he has since stepped away from Figma to pursue other interests (but still owns 6% of the company).

What they realized, leading to Figma, was that people choose tools not only because the tool is good but because it’s the most usable, especially when the work happens in a shared environment where your entire team’s coordination and productivity is at stake.

This is one of the reasons Figma’s position inside its niche became so dominant that regulators got involved in the Adobe acquisition attempt. What regulators were worried about wasn’t merely that Adobe + Figma would create a monopoly in collaborative design, but that Figma already looked close to a monopoly in that narrow vertical, and merging it into Adobe’s broader set of creative monopolies would have been even more problematic.

By the way, that’s a rare kind of compliment for a business to receive. It’s the sort you’d rather not receive in the form of a blocked acquisition, but it’s still a signal about the kind of market power Figma achieved in a relatively short window of time.

Over the years, Figma has boomed, while its closest competition faded. Competitor Sketch’s growth fell off, and Adobe actually sunsetted Adobe XD — their version of Figma — after recognizing they were so far behind (which is what then made them a motivated buyer of Figma).

Figma is, and has been, the best place to collaboratively design websites and apps. That’s what UI (user interface) refers to in the graph above. This is a narrow niche in the greater world of design, but an important one nevertheless. Going from $0 to $20b in value creation was driven by dominance of this niche. Now, for the company to go from ~$10b in market cap today to $100b+, they need to expand into more verticals and create a more comprehensive ecosystem to rival Adobe.

And that’s exactly what they’re doing.

Figma is trying to expand beyond a single flagship tool into a platform, where the endgame is less about selling one product and more about becoming the default operating system for a category of work. Correspondingly, Figma has been steadily layering in adjacent products.

FigJam, for example, gives designers & non-designers a space to sketch user flows and map out ideas with digital sticky notes, which takes the messy early-stage ideation process and keeps it inside the same ecosystem where the polished design work will eventually live. As such, FigJam digitalizes a lot of work that would otherwise occur on a team’s in-office whiteboard.

Then there’s Figma Slides, a collaborative presentation tool that looks a lot like Google Slides, honestly, but is valuable in that you can embed live prototypes of your work within other Figma tools directly into a deck.

In other words, it’s an outlet for designers to effectively demonstrate how an app/website will look and work to their non-designer counterparts without exiting the Figma ecosystem.

The last year, however, has been a blitz of rapid launches. Following a cadence of releasing new products every 1-2 years, Figma suddenly dropped four new products in 2025: Sites, Make, Buzz, and Draw, with AI baked into each one’s experience.

Based on the names, you can mostly figure out what the products do:

That AI point is where Figma gets interesting, and also where the analysis gets especially difficult, because AI is both an accelerant and a source of new competition.

The good thing about having a monopoly is that you’re on top. The bad thing is that everyone wants to take your spot. As Jeff Bezos would say, “your margin is my opportunity.” The question investors in Figma must understand is to what extent LLMs help to reinforce the company’s advantages versus opening the door to competitors that’ll eat into its market share and pricing power?

Of the products released in the last year, Figma Make is the most impressive to me. The idea is that, in plain English, you can prompt an LLM to code something on your behalf with zero prior coding knowledge needed.

And the thing is, it’s really, really good. The capabilities may not seem so magical to folks who’ve been fixated on the cutting edge of AI, but to more casual observers like myself, well, it truly does feel like magic.

For example, with just a sentence or two of input, I tested Figma Make by asking it to develop a meditation app on my behalf.

In a few minutes, it built an app with calming sounds, breathing techniques, timers, and most impressively, it figured out things I’d probably want without much guidance. I’ve always wanted to build an app, and I’m glad I waited, because even just a few years ago, I would’ve needed to invest dozens and dozens of hours into learning the necessary coding skills and doing the work.

Instead, Figma Make was able to reason and test out different iterations and determine what needs to be added to improve the user experience. To me, it’s an eye-opening shift from the era of AI where LLMs can assist with tasks but need very explicit guidance and oversight, to an era where LLMs can stylistically “think” for themselves and make judgments on their own about the most optimal way to build a product.

I don’t want to overhype the output, but you can check out the app for yourself here. For something that literally took two minutes to generate, it’s not bad!

What humbles me is that this is only scratching the surface, because if I can do that on a first try, just imagine what someone who actually knows what they’re doing can accomplish in an hour, a day, or a week!

Figma Make launched in July 2025 and has already been used by 30% of Figma’s customers with contracts above $100,000 in ARR, which is a pretty compelling early adoption signal among Figma’s largest customers, especially because enterprise customers tend to be the slowest to change behavior.

As panic has swirled around companies that provide software-as-a-service (SaaS), with fear consuming investors that LLMs and AI agents will disrupt SaaS business models, companies like Salesforce, Adobe, and Figma have faced blistering selling pressure.

Historically, enterprise software was sold per seat (per user), and the implicit assumption was that software scales with very little incremental cost, so revenue growth can cause a beautiful expansion in profit margins.

You can see this perfectly with Adobe’s operating margins since 2014, as revenues have compounded, margins rose from ~10% to more than 36%:

The concern is that AI changes this dynamic: a single user can become dramatically more productive, which means companies might be able to pay for fewer seats.

At the same time, because LLMs are so compute-intensive, they come with substantial marginal costs, which means that as software companies embed more and more AI-based tools into their platforms, variable costs will structurally increase, and they’ll benefit less from the kind of operating leverage that Adobe has flexed (shown above), where margins can scale rapidly.

If that bears any truth, then companies that were typically seen as having among the highest quality, most consistent earnings will have to be re-rated (i.e., lower P/E multiples) to reflect this shift.

To what extent this nightmarish reality unfolds for SaaS investors remains to be seen, but Figma is very much leaning into incorporating AI into its ecosystem. Clearly, they’ve done so with Figma Make, but aslo, they made an acquisition recently of an AI company known as Weavy, which has since been rebranded to Figma Weave.

It’s expected that Weave will be monetized on a usage basis rather than a per-seat basis, because that lines up better with the value being delivered and the costs being incurred. Charging based on usage is something of a compromise, reflecting that, if organizations can trim headcounts and get more done with fewer workers, SaaS companies need to effectively charge significantly more per user to keep their businesses intact.

To roughly simplify, if a company went from paying for 5 seats providing licenses to an enterprise software tool, now, in theory, they might only pay for 1 person to have access, but that single user may chew up 5x as much as computing power as they would’ve in the past, as AI tools amplify their productivity but also the variable costs powering their work.

So if the offset is actually that clean, then SaaS companies may be able to charge based on usage without much of an impact on their top and bottom lines, but nobody quite knows how this will break out yet.

What if one person can do the work of 5 people in the past, and consume as much computing power as 3 people would have? In that case, charging based on usage would be better than charging per seat, but there’d be a substantial decline in overall revenue.

Weave, then, will use a separate system for AI credits and hints at a future where enterprise design subscriptions include consumption-based revenue streams. The catch is that, for now, Figma isn’t actually charging for AI credits. They want people to try the tools, embrace them, and build habits before pricing gets tightened, so at the moment, Figma is shouldering the inference costs, hence why I say it’s too early to have much insight into how this different pricing model will shape the business’s profitability.

The thing is, as much fun as it is to speculate about SaaS in a world of AI, Figma has, irrespective of AI, recently made significant changes to its billing model.

The intention is to deepen relationships with large enterprises, but by introducing more friction and formality into their billing processes, the very loop that helped Figma spread organically is being disrupted.

Early last year, Figma shifted from a “user-driven” model, where existing users could order new paid seats and administrators would review charges retroactively, to a model that requires upfront admin approval before a new license is provided.

Younger companies might appreciate the flexibility of the old system, but mature enterprises don’t want unexpected costs, and they do want visibility and control over spending. The tradeoff is that admin gating introduces friction into organic growth. Figma has also moved away from selling individual products separately and instead introduced a handful of seat packages, more similar to Adobe’s suite packaging. And because a bunch of new tools were released, Figma raised the price for a full seat by 30%.

Unsurprisingly, some users complained about this, especially those who liked full access but didn’t love paying more for tools they don’t use. That’s the bundle problem in a nutshell. Bundles are incredible deals for some customers and a persistent irritation for others, and you don’t really know which group dominates until you watch churn and net retention after the dust settles.

On that note, impressively, for customers spending over $10,000 annually, Figma reported net dollar retention (NDR) of 131%, meaning that, in aggregate, those same customers increased spending by 31% year over year, even after accounting for cancellations & downgrades.

The complication, though, is that this NDR figure largely reflects performance before the billing changes, which makes it stale information. We don’t actually have much data yet on how damaging the move will be to growth rates, churn, and NDR.

It’s also important to know that Figma is already pretty top-heavy.

In its S-1 IPO filing, Figma reported having 450,000 paying customers as of March 2025, and with roughly $1 billion in ARR, that implies around $2,000 per year per customer, though the definition of “customer” can include entire organizations or multiple departments within the same company being counted separately. So their exposure to a few large organizations could be even greater than it seems.

The top 0.2% of customers generated about 37% of ARR, and the top 2.5% comprised more than 60% of ARR!

Figma is also a surprisingly international business for being so young, with revenues split roughly 50/50 between the U.S. and international markets, though it monetizes U.S. users more meaningfully. Only about 15% of monthly active users are based in the U.S., yet almost half of Figma’s revenue comes from the U.S.

And deal momentum is picking up, with the number of customers signing multi-year deals increasing 27% quarter over quarter.

This is a business with viral word-of-mouth adoption, low churn, expanding use cases, and a mix of self-serve and enterprise sales, which makes it easier to see why the market was excited at IPO. And the question is, why did the stock decline 80% if the business looks so strong?

To me, Figma looks close to a monopoly in its niche, and normally, investors love monopolies, but monopolies in narrow niches can run into TAM ceilings faster than the market expects, while moving into adjacent categories tends to bring you into collision with well-funded incumbents like, well, Adobe.

In other words, the market loved seeing a fast-growing near-monopoly, but as waves of insiders have cashed out post-IPO, and the sentiment around SaaS companies has shifted, few marginal buyers have stepped up to balance supply and demand.

Figma is now firmly in the stage of trying to move into new verticals, which means intensive capital investments into areas with more competition and less certain returns, hence the four new product launches in 2025.

Point being, Adobe has a broad portfolio, deep enterprise relationships, and a lot of muscle memory in creative tooling, and while Figma very clearly won the collaborative design battle, scaling into a wider suite means fighting on more fronts that Figma hasn’t yet proven it can win.

Now we’re arriving at the part of the story that makes valuation, in the traditional sense, unusually difficult right now. Figma isn’t profitable yet, and while part of that may be “normal” for a company investing heavily in growth, a large part of the ugliness in the reported financials is tied to what happens when you go public and stock-based compensation accounting catches up with economic reality.

If you look at Figma’s financials without context, it looks like the company IPO’d and margins collapsed immediately. The reason is that, for years, Figma wasn’t recognizing the expenses associated with employee stock awards in the same way that post-IPO reporting as a public company forces you to. Pre-IPO, maybe half of an employee’s comp package was stock, but only the cash portion was reflected as an expense in financial statements. But with the IPO and related liquidity events, huge chunks of accrued stock-based-comp vested, and the company had to recognize these expenses all at once.

So you could say 2024 and 2025 look artificially bad because they include past years’ costs, or you could say the private-company accounting was artificially favorable and didn’t reflect the real economic costs Figma faced.

Both are true, which is why pre-IPO and immediate post-IPO numbers aren’t all that useful if your goal is to value the business based on more normalized earnings looking forward.

Figma’s recognition of accrued SBC has actually exceeded revenues in the last two years!

Then there’s Dylan Field’s comp to consider. I do like Dylan, and he comes across as sincerely kind. People often point to him as an example of a tech CEO who achieved success without being a manic psychopath.

Field owns about 13% of the company but controls 73% of the voting power, and his compensation package included multi-year stock price targets that, in hindsight, were implemented in a way that seems to have rewarded short-term stock spikes rather than sustained value creation.

The problem is that the stock hit target levels that the board thought might take years to achieve after IPO, meaning Field unlocked his long-term bonuses almost immediately, and then the stock cratered, leaving shareholders with losses while the CEO had already “earned” years’ worth of performance incentives. Evidently, tying comp to stock price targets without proper time-weighting or consideration of per-share metrics like operating profit per share or earnings per share is problematic…Long-time readers of this publication should know that well.

So where does that leave us?

Daniel’s view was that, unless AI suddenly flips everything upside down, the business has the ingredients you’d normally want to see: viral adoption, expanding use cases, and robust growth, with revenues compounding at a 45% CAGR since 2023.

And he’s right to point out that the fundamentals don’t obviously justify an 80% stock decline on its own, which is why the situation is tempting.

But temptation and investability are different things, and for me, the challenge with Figma right now is that I don’t have enough clarity on what “normal” looks like. I can’t look at margins and confidently say where they settle once SBC normalizes. I can’t look at NDR and assume it holds after billing changes. I can’t look at the AI product suite and confidently forecast whether it expands the addressable market or compresses per-seat economics and injects variable costs.

And I can’t look at governance and compensation and feel good that incentives are cleanly aligned with sustained per-share value creation, even if I like the founder and even if I believe the product is exceptional.

In other words, I can tell you why Figma’s products are good, but I can’t tell you what the business is worth with the kind of confidence I need to allocate capital.

Sometimes the right move is to watch a great business for a bit longer. I need to see how things look after the reporting normalizes, pricing changes wash through retention metrics, and once AI monetization becomes more explicit.

So, for now, we’re not adding Figma to The Intrinsic Value Portfolio. I’d rather be a little late with better information than early with a narrative.

Updates on our Portfolio holdings below!

"You can't connect the dots looking forward; you can only connect them looking backwards. So you have to trust that the dots will somehow connect in your future.”

— Steve Jobs

📺 WATCH: CBS’s extended interview with Tim Cook

🎧 LISTEN: Die With Zero and Linde Stock Analysis

📖 READ: AI Hurtles Ahead, by Howard Marks

You can also read our archive of past Intrinsic Value breakdowns, in case you’ve missed any, here — we’ve covered companies ranging from Alphabet to FICO, Transdigm, Lululemon, PayPal, DoorDash, Crocs, LVMH, Uber, and more!

Enjoy reading this newsletter? Forward it to a friend.

Was this newsletter forwarded to you? Sign up here.

Join the waitlist for our Intrinsic Value Community of investors

Learn how to join us in Omaha for the 2026 Berkshire Hathaway shareholder meeting!

Shawn & Daniel use Fiscal.ai for every company they research — use their referral link to get started with a 15% discount!

Use the promo code STOCKS15 at checkout for 15% off our popular course “How To Get Started With Stocks.”

Read our full archive of Intrinsic Value Breakdowns here

Keep an eye on your inbox for our newsletters on Sundays. If you have any feedback for us, simply respond to this email or message [email protected].