Figma, Inc. FIG

Key data

The business

Figma is one browser tab where a whole product team designs the thing before it gets built. A designer draws the screens, a product manager leaves comments, an engineer opens Dev Mode to read the exact spacing and colors as code, and none of it is a file emailed around, just one living link everyone edits at once. That shared link is the moat: the company's design system, its reusable buttons and components, lives inside Figma, so leaving means relocating every team's active work at the same time. Around the core design tool the company has bolted on whiteboards, slides, marketing assets, published websites, and two AI products, Make and Weave, that turn a written prompt into a working prototype. Most of the money still comes from teams paying for each person on the core tool. Dylan Field, the founder, still runs it, with about 1,900 employees.

The numbers

The story the sequence tells is acceleration, not a fade. Revenue growth ran 41% for full-year 2025 and then quickened to 48% in the June quarter, its fastest print as a public company.

Ignore the September 2025 loss. It is a one-time stock charge from the IPO, when the company booked more than a billion dollars of restricted-stock expense in a single quarter. Underneath, sales rose every quarter, gross margin held in the low-80s% and reached about 84% in June, and the paper loss has shrunk each quarter since. The question coming into this print was whether growth would stay near 40% with margins steady. It did better, holding 48% at the widest gross margin the company has posted, so that leg held.

The 2023 profit is not real earnings power. It is the $1B breakup fee Adobe paid when regulators blocked its takeover. Strip that and the picture is a company that has roughly doubled revenue in two years, a 45% compound rate, now running past a $1.4B annual pace. The number that matters against the losses: Figma generated $246M of real cash in 2025, close to a fifth of sales, while reporting a $1.25B loss. The whole gap is stock pay, which ran above 100% of revenue on a trailing basis because of that IPO grant. On the earnings the company asks you to judge it by, it has come in ahead of expectations every quarter since listing. What this memo believes that the tape does not: the multiple has already de-rated for a slowdown that has not arrived, and the June acceleration is the print the market has not yet re-priced. The next revenue growth rate against this 48% settles it.

Management

The record is a founder holding a giant paper grant and executives selling into the first open windows. Dylan Field's 2025 pay reads as $864M, but essentially all of it is a one-time $862M restricted-stock award tied to going public, not cash out the door, and the CTO got a $174M version of the same. Since the IPO, insiders have sold about $50M across 33 transactions and bought nothing, the largest being the CFO's $9.6M and the CTO's two sales totaling $14.7M, several clustered on July 29 right after the June quarter opened the window. That is post-lockup selling into strength more than a warning, but the direction is one-way. There are no buybacks and no dividend, the company just listed, and shares outstanding are up about 5% in two years. Whether the July sales were pre-scheduled or discretionary is not separable from this data.

How it fails or surprises you

Seats meet AI (downside). Figma charges for each person who edits. If its own AI lets a team ship the same work with fewer designers, paid seats can stall while usage climbs, and revenue growth quietly decouples from adoption. The print that shows it first is net dollar retention and paid-seat growth over the next two quarters. Watch for the two lines to split.

The stock-pay gap (downside). The cleanest bear fact is in the memo's own numbers: a $1.25B reported loss against $246M of cash generated, a gap that is almost entirely stock. The IPO grant pushed trailing stock pay above revenue itself. If it does not fall back toward that cash line over the next few prints, the "one-time" story is wrong and the dilution is permanent.

The prompt-to-software door (right tail). Make and Weave could turn Figma from a tool designers pay for into the front door where anyone builds software by describing it, moving the buyer from a few designers per company to every builder. The market pays nothing for this today. The first print would be AI revenue broken out as its own line, or new-customer growth stepping up outside the core design seat.

Closing thoughts

This is an uncertainty a named print resolves, not an exposure you only survive. The stock has already fallen to a third of its IPO price and about nine times sales, which prices a real slowdown, and the June quarter instead accelerated. If the next one or two prints hold growth in the 40s% at this margin and stock pay starts falling toward the cash line, the de-rating was the opportunity. If growth rolls toward the 30s% as AI thins design seats, nine times sales of a company that loses money on paper is not cheap, and the multiple compresses further. On the evidence in hand the fatter tail is the upside one, because the number that would prove the bear right, decelerating seats, is the one thing the latest print moved the wrong way for the bears.

The bet is still that Figma stays where product teams design and pulls the people around them in, and that AI adds seats rather than erasing them. What breaks it is the pair to watch every quarter: paid-seat growth and net dollar retention against the revenue line. The day usage keeps rising while paid seats flatten, the read is wrong, whatever the headline growth rate says that quarter.