Amplitude's agentic pivot goes from vision to number

Revenue crosses $100M, agent share of insights jumps past 40%, and the market hands the stock a +137% three-month re-rating — even as the AI tax on gross margin starts to bite.

AMPL · Earnings Call · 2026-08-05

Price action: a market that has already decided

Before a single financial was read, the tape had telegraphed the answer. Amplitude's shares are up nearly 137% over the trailing 90 trading days — a single unbroken up-17w segment that stands in stark contrast to the name's broader life: still down ~76% since listing and sitting ~84% below its late-2021 peak. This is a market pricing in a genuine re-rating from legacy analytics SaaS to agentic AI platform, and the Q2 2026 print largely validates that swap. The loudest signal on the call is not a financial one at all. pull request is Amplitude's top keyword of the quarter — an almost embarrassingly company-unique topic for an earnings call, and a direct window into how this company now operates. Spenser Skates spent the opening of his prepared remarks on organizational transformation, not guidance:

This has all resulted in 3x the number of pull requests in 6 months. We have reduced our pull request cycle from 5 hours to 44 minutes. Bug reports are down 55%. 5% of our pull requests are submitted from designers and product managers with no engineering involvement.
Spenser Skates · 2026-08-05

The numbers: first $100M quarter in hand

The financial milestones confirmed the narrative. Total Revenue hit $100.9M, up 21% year over year; ARR reached $410M (+22% YoY, +$36M sequentially — split between ~$17M of assumed Statsig ARR and ~$19M of organic growth). Total RPO grew 35% YoY to $483M, with current RPO up 30%. The balance sheet machine is also finally whirring: record quarterly free cash flow of $23.7M (24% of revenue) and $69M returned in buybacks. “We crossed $100 million in quarterly revenue, ARR reached $410 million, growing over 22% with the addition of the ARR assumed from the Statsig, business, and free cash flow was a record quarterly high of $23.7 million.” — Andrew Casey · 2026-08-05 The "how" is the interesting part. The company consolidated on a new pricing and packaging model — 70% of closed ARR in Q2 was on the new model (up from 25% in Q1), now 28% of total ARR — and the data-in-versus-entitlement utilization metric that CFO Andrew Casey tracks sits at an all-time high, a leading indicator for upsell.

AI adoption: the promise kept

The most investable data point is the agent share of insights. agent analytics now accounts for over 40% of all insights versus humans — a sharp step up from the ~25% figure cited earlier in the year. This matters because the company deliberately withheld the number a quarter ago: “We are not sharing numbers on this particular call, but we will have an update next quarter on agent adoption relative to human usage of data analytics.” — Spenser Skates, CEO and Co-Founder · 2026-05-07 They kept the promise, and the update was directionally strong — with Global Agent handling 1.3M interactions per week, management says root-cause discovery rates are improving one percentage point per month. That step-up also reinforces why the company bought what it bought. On the Statsig partnership, the framing from the prior call still holds: “Vijay was looking for a home for the kind of continued support of the Statsig customer base. And after looking at a number of different places, him and I agreed that the best place that would be Amplitude.” — Spenser Skates, CEO and Co-Founder · 2026-05-07

The acquisition tax on gross margin

The explicit cost of this AI-native turn is on the gross-margin line. Gross Margin printed 71%, down roughly 4 points year over year, driven by accelerating inference costs and the Statsig integration (whose hosting environment runs in the low 50s, targeted at 70+).

We are embracing this change in cost structure as part of our transition to an AI native company. For now, we expect gross margins to stay in the low 70s. We will offset that with a commensurate reduction in operating expenses.
Spenser Skates · 2026-08-05

That tension — high inference cost against the promise of 20%+ long-term operating margins — is the crux of the model. Management's answer is to keep sales and marketing below 40% of revenue and G&A in the low teens, buying back the margin on the cost side.

Wade and the (paper) future of product development

The vision rests on future of product development — the Wade product, still in alpha, that promises "self improving products" that recommend what to build next from user signals. The demo was real (a docs-site search failure found, fixed, and measured automatically), and the TAM framing was explicit: Spenser called the winnable outcome "multiple billions in revenue, if not more." Meanwhile Statsig customer cross-sell remains the near-term lever — ~80 of 400 Statsig customers already overlap with Amplitude, but the bigger unlock runs the other way, bringing the experimentation platform to Amplitude's far larger base. What changed this quarter, in one line, is this: the agentic story went from vision to number. Revenue crossed $100M, agent share of insights jumped toward 40%+, and the market responded with a +137% three-month re-rating — even as Price-to-Revenue sits at ~2x, down ~91% from its 2021 peak of ~22x. Whether that re-rating is durable depends on whether gross margin can be clawed back as fast as inference costs grow. That is the equation to watch into Q3 and Q4.