AMPL · Amplitude, Inc. · Stock research

Current Thesis

The question this name carried into August — whether the Q2 print would validate the re-rate — was answered on 2026-08-05. The question now is how far a 98.4% three-month move runs before the next company-reported number arrives, and that number is roughly eleven weeks out. The 2026-08-21 close of $13.29 sits 0.1% under the 52-week high of $13.30. Two target raises landed after the initial post-print cluster: Morgan Stanley to $16 from $12 with an Overweight rating on 2026-08-13 — the first published target above the $13 ceiling set on 2026-08-06 — and BofA to $14 from $12 on 2026-08-20, inside a ten-name software raise framed on fading AI-disruption fear rather than on Amplitude specifically. The unfinished part of the story is unchanged: dollar-based net retention printed 103% (105% pro forma) against management's 115% anchor, and nothing dated addresses it before the Q3 report (~2026-11-04, est.).

Bullish and bearish views on Amplitude, Inc.

The model's bull view on Amplitude, Inc. (AMPL), in brief: Q2 2026 (2026-08-05): revenue $100.9M, +21% YoY, against $98.165M consensus — the first quarter above $100M. The bear view: Net retention of 103% (105% pro forma) is barely expansionary and sits 12 points below the 115% figure the re-rate is underwritten on. Both cases follow in full.

Bull Case

• Q2 2026 (2026-08-05): revenue $100.9M, +21% YoY, against $98.165M consensus — the first quarter above $100M.
• ARR $410M, +22% YoY, versus $374M and +17% YoY reported at Q1 on 2026-05-06. Statsig's disclosed ~$16M ARR entered the base that quarter.
• RPO $483M, +35% YoY, and 824 customers at or above $100k ARR versus 634 a year earlier (+30%) — contracted backlog and enterprise count both compounding faster than reported revenue.
• FY2026 guidance raised on every line on 2026-08-05: revenue to $407.2–411.2M from $397–403M, non-GAAP operating income to $6.3–9.3M from $2.5–6.5M, non-GAAP EPS to $0.06–0.08 from $0.03–0.06 against $0.05 consensus. Q3 revenue guided $105.6–108.0M versus $102.581M consensus. For scale, FY2025 revenue was $343.21M, +14.68% YoY (stockanalysis.com).
• Free cash flow $23.7M in Q2 at a 24% margin, versus $18.2M a year earlier; cash, equivalents and marketable securities $161.4M.
• Sell-side ceiling keeps moving: Raymond James initiated Strong Buy at $10 on 2026-07-01; the post-print cluster reset to $12.5–$13 on 2026-08-06; Morgan Stanley went to $16 on 2026-08-13; BofA to $14 on 2026-08-20. That is roughly a 60% move in the top published target inside seven weeks.
• The product leg predates the print: Agentic AI Analytics launched 2026-02-16 with MCP connectors piping behavioural data into Anthropic, OpenAI, Cursor, Figma, Notion and GitHub.

Bear Case

• Net retention of 103% (105% pro forma) is barely expansionary and sits 12 points below the 115% figure the re-rate is underwritten on. Pro forma NRR was 104% in Q2 2025 — one point of movement across a year.
• Price has caught up with the aggregate view. The mean published target is $13.77 across 12 analysts as of 2026-08-22 (stockanalysis.com), about 3.6% above the 2026-08-21 close, and Citi's Neutral rating with an $11 target (2026-08-07) sits below spot. Only Morgan Stanley's $16 leaves meaningful headroom.
• BofA's 2026-08-20 raise was a sector call across ten software names, not an Amplitude-specific upgrade — the marginal bid in the last week is at least partly group flow.
• GAAP economics are still wide: Q2 GAAP operating loss $(35.2)M and net loss $34.62M on $100.9M of revenue, GAAP EPS $(0.27) against non-GAAP $(0.01), with stock-based compensation bridging the two. Market capitalisation was $1.66B on 124.92M shares outstanding as of 2026-08-22.
• FY2026 non-GAAP operating income of $6.3–9.3M on $407.2–411.2M of revenue is a thin cushion; one quarter of integration or hiring spend removes the margin half of the story.
• RSI(14) at 82.4 with the shares up 98.4% over three months, into a calendar with no company-reported hard number until roughly 2026-11-04.

Setup & Price Structure

The narrative is accelerating, dated by the 2026-08-13 Morgan Stanley raise to $16 (the first target printed above the post-earnings $13 cluster), the 2026-08-20 BofA sector raise, and a 52-week high made on 2026-08-21. Attention is still widening rather than thinning — but the crowding markers are stacked at the same time, and they are worth stating as observables rather than as a verdict.

Observable positioning evidence, as of 2026-08-21/22:

• The close of $13.29 is 0.1% under the 52-week high of $13.30; three-month price change of +98.4%; RSI(14) 82.4.
• Spot sits above every published target except Morgan Stanley's $16, and 3.6% below the $13.77 mean.
• Structure below: the 2026-08-07 close of $11.26 marks the immediate post-print shelf, and the prior 52-week high of $12.60 is the level the August breakout cleared. The July consolidation, before the print, ran $8.30–$9.65.
• Insider flow on 2026-08-17 was mechanical rather than discretionary: four officers reported Class A shares withheld at $13.03 to cover tax on an RSU vest dated 2026-08-15 — Chief Commercial Officer Nathaniel Glenn Crook 109,368, CFO Casey Andrew 72,814, CTO Curtis Liu 53,878, CEO Spenser Skates 52,369. These are withholdings, not open-market sales, though the vest itself puts freshly delivered stock into insider hands near the highs. Separate CTO sales run through a Rule 10b5-1 plan adopted 2025-11-26.
• No earnings date sits inside the next 30 days; the print already happened on 2026-08-05.

Catalyst Calendar (next 30 days)

• 2026-09-09, 3:55 p.m. ET — Management participates in the Citi Global TMT Conference (announced by press release 2026-08-10). First scheduled public commentary since the print; the venue where a net-retention trajectory update, or a reframing of the 115% target as a long-term ambition, would surface.
• 2026-09-15, 3:30 p.m. CT — Piper Sandler Growth Frontiers Conference (same 2026-08-10 announcement). Piper carries an Overweight rating with a $13 target set 2026-08-06, now below spot.
• ~2026-11-04 (est.) — Q3 2026 results. Outside the 30-day window and the only dated hard number: Q1 landed 2026-05-06, Q2 landed 2026-08-05.
• No FDA-style binary, no lock-up, no index event scheduled in the window. Analyst revisions can arrive on any date and are not schedulable.

What Would Change Our Mind

The structural break is the loss of the August post-print shelf. Price cleared the prior $12.60 52-week high and left the 2026-08-07 close of $11.26 as the last consolidation before the Morgan Stanley leg; a weekly close below $11.20 gives that entire leg back and puts the shares inside the range they occupied before the sell-side ceiling moved.

On the fundamental side, the trigger is retention. If the 2026-09-09 and 2026-09-15 conference appearances come and go with no update to the 103% net-retention figure — or with the 115% target restated as a multi-year ambition rather than a trajectory — then the leg above $12.60 rests entirely on multiple expansion in a $1.66B-capitalisation software name whose next verifiable number is in November.

A third condition: the theme flipping to saturated. The observable version of that is target revisions stopping while price stalls in the $12.60–$13.77 band through September, with no new published target above $16 and no rating changes out of the Neutral cohort. Conversely, a Q3 print showing net retention above 105% core with the FY revenue guide of $407.2–411.2M held or raised would extend the leg rather than break it.

Correlation Notes

• Group beta dominates at this size. BofA's 2026-08-20 note lifted targets across ten software names on the same thesis (AI-disruption fear fading), so a reversal in that sector call pulls AMPL with it regardless of company execution.
• Direct comparables for the retention read are Braze, Sprout Social and Klaviyo — mid-cap application software with NRR in the low 100s. Their autumn prints land before Amplitude's Q3 and set the read-through on seat and usage expansion.
• The AI-analytics leg correlates to MCP-ecosystem newsflow (Anthropic, OpenAI, Cursor, Figma, Notion, GitHub integrations shipped 2026-02-16), which is headline-driven and unscheduled.
• With a GAAP operating loss of $(35.2)M in Q2 against non-GAAP $(1.5)M, the shares carry the standard duration sensitivity of unprofitable-on-GAAP SaaS: rate moves and IGV/WCLD drawdowns transmit faster than fundamentals.