FIG: A Live Test of the Perceptual Roller Coaster

In an earlier piece I argued that asset prices are not driven by fundamentals — they are driven by Expected Growth, a sociological construct in which fundamentals are inputs but psychology, narrative, and herd behavior do the actual work of price formation. The most actionable corollary of that framework is the perceptual margin of safety: situations where pessimistic perception has not yet caught up with strong fundamentals. (See Perceptions Are Fundamental — Investing Is a Sociological Game.)

Figma is one of those situations right now.

A short history of the narratives the market has attached to Figma

FIG is a textbook case of the sociological cycle in action because the market has told three distinct stories about this company in three years, each one moving well ahead of the underlying business.

Story one — "The $20B design platform." In September 2022, Adobe announced its intention to acquire Figma for roughly $20 billion. Overnight the company became, in the collective imagination, the asset Adobe was willing to pay a generation-defining premium to own.

Story two — "Worth more standing alone." In December 2023 the EU and UK forced the deal off. The collapse was widely understood to be Adobe's antitrust problem, not Figma's, and the market never repriced FIG downward for it. When Figma IPO'd in July 2025 at $33, the stock opened at $115.50 the same day — a 250% pop, the largest first-day move for any US IPO over $1 billion in three decades — and briefly traded at a fully-diluted valuation near $68 billion. The verdict was explicit: the market believed Figma standing alone was worth substantially more than what Adobe had agreed to pay. Story two was a re-rating, not a recovery.

Story three — "AI is killing Figma." By late January 2026 the stock was near $21, market cap roughly $10.5 billion, down approximately 85% from its post-IPO peak. The dominant explanation crystallised: design tools were the category generative AI would eat first; anyone with a chat box could draft a comp; the moat was melting; Figma's incumbency in professional design hands could not survive the AI shift.

Story three is the one the consensus still partially holds. It is internally coherent and emotionally satisfying, which is exactly the kind of narrative that lingers after the underlying facts have moved. The forty-day stock return going into last week was −3.4%. The market had not rewritten the model.

What the earnings actually showed

On May 14 Figma reported a quarter that the dominant narrative struggles to absorb.

Revenue of $333.4M came in $20M above consensus, a year-over-year growth rate of 46% — the second consecutive quarter of YoY acceleration, walking 38% → 40% → 46% across three prints. Net dollar retention rose to 139%, a two-year high, meaning existing customers expanded their spend by 39% over the prior twelve months. The company added 1,357 net new paid customers above the $10,000-ARR threshold — its best quarter on that metric ever. Management raised the full-year revenue guide by $55M after only one quarter, an unusually large mid-year adjustment that signals their own internal conservatism is being overrun by what they are seeing on the ground. AI feature monetization through credit consumption, enforced from March 18, had not yet flowed through a full quarter; management explicitly framed Q2 as the first one that will.

Two consecutive quarters of acceleration through a major pricing-tailwind anniversary is the pattern that separates a real inflection from a single-quarter anomaly.

The analyst gallery as live evidence of perceptual lag

The same afternoon, RBC Capital cut its price target on FIG to $28 from $31. JP Morgan cut to $42 from $45. Barron's headlined the print "Figma Stock Rises. Why the Software Maker's Earnings Defy the AI Narrative" — the framing itself confirms the consensus has not yet rewritten the model.

This is the perceptual margin of safety being observable in real time. The kill metric for the "AI commoditizes Figma" thesis was two consecutive quarters of revenue re-acceleration with NDR holding north of 130 and visible AI monetization. All three legs are present in the print, and yet the published sell-side perception is moving the wrong way. The framework predicts exactly this: facts arrive faster than the social process that converts them into consensus. Narrative is sticky because it is social.

What the product story now looks like

A separate question is whether the inflection is built on something durable. The product evidence — readable directly from Figma's own release cadence — supports the structural read.

The shape of that cadence over six years is more informative than any single moment in it. Through 2023 product velocity accelerated steadily, culminating in the major design-to-code push around Dev Mode and variables — the company's first serious attempt to escape the "design tool" category into the broader product-development workflow. 2024 read on the surface as a relative pullback in surface expansion, but reading it as deceleration misses the substance: it was a platform-hardening year between two waves, the moment the company turned earlier launches into something enterprises could standardize on. 2025 then produced the largest acceleration of the entire arc, with the product widening to Make, Sites, Buzz, Draw, and Weave — moving the company decisively past the "collaborative interface design" identity. 2026 so far carries an even higher run-rate, but the texture is different: fewer brand-new categories, far more depth — Make connectors; hooks into Claude Code, Codex, GitHub Copilot, ChatGPT, Microsoft 365 Copilot, and Notion; MCP workflows; reusable Make skills; FigJam repositioned as a whiteboard for coding agents.

The shape that emerges is unambiguous. The center of gravity has shifted from "collaborative interface design" to a coordination layer trying to own the workflow from idea to whiteboard to design system to prototype to code to published site or marketing asset, with enterprise governance running across all of it. Customer behavior confirms it: the share of large customers using Make weekly roughly doubled in two quarters, meaning the product evolution is being consumed, not merely shipped.

The seat-based SaaS narrative is being replaced — by Figma, in the open — with a hybrid seat-plus-consumption model in which AI is a revenue stream rather than a threat.