Figma is finished. Most designers haven't noticed yet.

Figma's stock closed at around $21 last week. Its August 2025 high was $143.

That is an 85% drop in nine months. From a company that just reported Q4 2025 revenue up 40% year-over-year, full-year revenue of $1.06 billion, and 2026 guidance of 30% growth. Operationally, Figma is doing fine. Financially, the public market is voting that it is not.

There is one specific argument the chart is making, and I think it's right: the era of design-as-a-separate-file-you-hand-off is over. Figma is the most beautiful product ever built for that era. The era is dying. The product cannot save it.

I stopped using Figma for production work in 2024. So did most of the designers I respect. The stock chart is the lagging indicator. This article is about why the chart is right, what the designers who saw it coming did instead, and what's left for Figma after the file format dies.

The chart

Here is the timeline, condensed:

July 2025 — Figma IPOs at $25-$28. Opens at $85, closes at $115.50. One of the year's strongest IPOs.
August 2025 — All-time high of $143.
March 18, 2026 — Google releases an updated version of Stitch, its "vibe design" tool that generates UI from prompts. Figma drops 12% in two days.
April 8, 2026 — Anthropic announces Mythos. Stock falls 14% over three days.
April 15, 2026 — Anthropic launches Claude Design. Stock drops another 7%.
End of April 2026 — Down 16% on the month. Trading in the high teens.
Today, May 2026 — Around $21. An 85% retracement from the August 2025 peak.

This is not a normal post-IPO unwinding. The S&P 500 was up 10.4% in April 2026. Figma's drop is idiosyncratic — a structural reaction to specific competitor announcements, not market sentiment.

Each leg down lines up with a release of a tool that does what Figma does, without the file. Stitch generates UIs from prompts. Claude Design generates polished prototypes and slides. Lovable, Bolt, and v0 generate working applications. None of them ask the user to start with a blank canvas.

The chart is reading: every announcement that compresses the design-to-deployed-product gap erases another billion dollars off Figma's market cap. The market is pricing the elimination of an entire workflow.

Stylised stock chart of FIG from August 2025 peak at $143 down to $21 in May 2026, annotated at four event points: Stitch update, Mythos announcement, Claude Design launch, and the April 2026 close.

What Figma was actually selling

Figma's product is excellent. That is not the argument. The argument is what Figma was actually selling, underneath the product.

Figma sold the artifact. A .fig file that captured a design, pixel-perfect, in a collaborative URL. Around that artifact, an entire industry grew up — design system files, component libraries, dev-handoff inspect tools, Dribbble shots, Behance portfolios, "design ops" job descriptions. Every one of these existed because the artifact was the bottleneck between idea and product.

The artifact was the bottleneck because translating a design into code was expensive. You needed an engineer, an interpretation pass, a hand-off ritual, a back-and-forth review cycle. The artifact justified itself by being the contract between two specialists.

In 2026 that translation is no longer expensive. v0 takes a screenshot and ships React. Cursor takes a Figma file and writes the components. Claude Code takes a written description and produces a deployed Next.js app. Lovable takes nothing — it just asks you what you want and ships it.

When the translation cost goes to zero, the contract is no longer needed. The artifact becomes a step you skip.

Figma's product is unchanged. Its role in the workflow is being deleted.

The Dribbble years

I owe my career, in part, to Dribbble and Behance. I posted there for years. I won small things. I got hired off them.

I also have to be honest: ninety percent of what I posted there was theatre. Polished mockups of products that did not exist, would never ship, were never used by a single human. A "redesign of Spotify" by someone who had never read Spotify's analytics. A "rebrand of Air India" by someone who had never been inside an Air India office. A "case study" of a "fintech app" that was a thirty-screen Figma file with no API behind it.

We — and I include myself here — built an entire performance economy on the artifact. Figma was the stage. Dribbble was the audience. Behance was the portfolio. LinkedIn was the applause.

The hiring industry took it seriously. Companies hired off Dribbble shots. UX programs taught students to build "case study" portfolios that were Dribbble shots with paragraphs. A whole generation of designers learned to optimize for the appearance of solving a problem, not the act.

When AI tools made the artifact-as-contract obsolete, they also exposed the rot in the performance layer. A Stitch prompt produces a "Spotify redesign" in fifteen seconds. The skill that used to produce that artifact in fifteen hours is no longer scarce.

The market is not just repricing Figma. It's repricing the whole industry that grew up around the artifact.

A 4×4 grid of Dribbble-style mockup thumbnails fading from full color in 2018 to near-greyscale in 2026.

What I did instead

In 2024 I made a specific choice: I stopped designing in Figma for any work that would actually ship. I designed in code. Cursor, then Claude Code. The case studies on this site — every one of them since 2024 — was designed as code, in the running product, with version control, with a deploy URL, with users on it within the same week.

The reasoning was simple math. Figma added a step. The step was expensive in calendar days, not just hours. Three rounds of handoff feedback take three weeks. Three rounds of "I built it, ship it, watch the user, iterate" take three days.

The artifact lost the math.

This wasn't a clever take. It was, by 2024, the obvious move for any designer who could read a stack trace. The designers who couldn't were the ones who stayed in Figma, and they're the ones who are now reading Reddit threads asking "is the design industry dying" while the chart answers them every day.

It is not the design industry that is dying. It is the artifact-economy version of it. Designers who can think — who can hold a user's mental model in their head, choose the right system, decide what to build — are more valuable than ever. The translation step is what's gone. The translation step was 70% of what most designers were paid to do.

The counterargument — and why it's weaker than it sounds

The strongest argument for Figma is this: AI tools generate fragmented artifacts. Figma is the source of truth. When v0 produces a component and Lovable produces an app and Stitch produces a screen, there is still a question of which one is right, what the design system looks like, who owns the spec. Figma, the argument goes, is where these competing artifacts get reconciled into a canonical design.

This is true and it is the next ten years of Figma's business. It's the bull case for the stock at $21.

I'd push back two ways.

First — "source of truth for design" is a real job, but it's a smaller job than "design tool for the entire industry." It's a job that lives at companies with enough designers to need governance — Google, Atlassian, Shopify. It's not the job that justifies a 200,000-customer SaaS at the prices Figma charges. The TAM contracts.

Second — the source of truth, in the long run, is the running product. Not a Figma file describing what the running product should look like. Tools like Storybook, design tokens in code, AI-driven design system maintenance — these are eating the "source of truth" job too. The reconciliation happens in code, not in a parallel design tool.

Figma at $21 is priced for a future where it's the canonical design-governance tool for enterprise. That future exists. It's just smaller than the future Figma was priced for at $143.

What this means for designers who are still in Figma full-time

Two specific moves, in order:

Open Cursor or Claude Code this week. Pick the smallest design problem you have at work — a card component, a single screen, a marketing page. Build it directly in code with AI assistance. Do not start in Figma. See how it feels. The first time will be uncomfortable. The third time will be obvious.
Stop building Dribbble portfolios. Build a thing that runs. The hire-able portfolio in 2026 is a list of working URLs, not a grid of polished JPEGs. Even one shipped side-project — a redesigned login screen, a small Chrome extension, a calculator — beats fifty mockups.

If you do these two things and they don't change your sense of what design is in 2026, fine — go back to Figma, that's a valid choice. Most of the designers who try will not go back.

The close

The Figma stock chart is not telling us that AI is replacing designers. It's telling us that the file is being replaced. The designer is fine. The thing they hand off is dying.

Most designers haven't noticed yet because the file is the air they breathe. It will become obvious to them the way the chart is becoming obvious to public-market investors — slowly, in a series of fourteen-percent drops, until one Tuesday they realize their main work artifact has been redundant for two years and they're the last to know.

Figma will survive. It will be a smaller, governance-focused product for companies that can afford a design-ops team. The designers who survive are the ones who already know that.

The ones who haven't noticed yet — open Cursor.

Sources used for this article: Why Figma Stock Lost 28% Last Month — Motley Fool, April 2026, Why Figma Fell 16% in April — Motley Fool, May 2026, Figma stock drops 12% after Google releases 'vibe design' product Stitch — CNBC, March 2026, FIG Stock Slips As Anthropic AI Design Threat Rattles Traders — Timothy Sykes, May 2026, Figma shares climb on earnings beat, but analysts note that AI risk remains — CNBC, Feb 2026, Figma Stock: AI Growth Is Real, But So Are Structural Risks — Seeking Alpha, Vibe Design Tools 2026 — NxCode. All numbers verified May 8, 2026.