Built by Claude Fable

Investor Sentiment -- 7/10

Two things are happening at once, and both are contrarian-positive. Targets are being cut hard -- $286 to $260 to $222 -- and coverage intensity is collapsing, from 15 targets over the year to 5 in the quarter to a single one in the last month. Sell-side attention is draining away from a $172B mega-cap. The most recent target implies only ~6% upside versus the ~24% implied by the quarterly average, meaning the marginal analyst has essentially capitulated to the current price.

That is the opposite of the ADSK setup used as the calibration anchor, where a Strong Buy consensus sat alongside targets roughly 60% above spot -- the bull case already being the street's base case, which earned a 4.

This is the main thing capping the score. A 76%-bullish rating distribution is not a hated stock, and there is room for downgrades if the print disappoints. Ratings are, however, the stickiest and least informative sentiment measure -- analysts cut targets long before they cut ratings, which is exactly the pattern visible above.

Not one published model assumes reacceleration. The street has settled on a flat ~9–10% revenue CAGR through FY30, in line with management's own FY30 framing. That is the definition of a reset expectation base: the bar management must clear to surprise is their own guidance, and the bar to disappoint is the same. It also means the 2H organic reacceleration management has promised is, if delivered, not in anybody's numbers.

Cheapness alone is not a contrarian thesis; it is just a low multiple. The contrarian position requires a specific, checkable claim where management and the street disagree -- and here there is one, stated on the record, escalating in conviction for three consecutive quarters, and quantified.

This is the sharpest fact on the page: the sell-side is carrying a FY30 revenue number roughly $2.5B BELOW management's own publicly raised target. That is not a stock where the bull case is consensus. It is a stock where the company's own guidance is treated as not credible.

The claim has escalated for three quarters while the street has not moved

The street's rebuttal, stated plainly on the call

The single most useful passage in six quarters of transcripts is Keith Weiss framing the disbelief directly to management:

That is the whole bear case in one question: strong AI KPIs, bookings that do not yet corroborate them. And it is a fair challenge -- cRPO landed exactly in line for two consecutive quarters, which makes backlog coincident rather than leading.

The proof points management offers that no model carries

Management is not arguing from sentiment. They keep putting forward operational evidence, none of which appears in a published estimate:

Why the street's scepticism is not unreasonable

A contrarian call is only worth making if you understand why the other side is there. Three reasons the disbelief is rational, and they should be held in view:

1. Management undercut its own quarter. The Q1 revenue beat, per the CFO, "was driven by Informatica's on-prem business and professional services timing" -- an admission that the upside was acquisition and timing, not demand.
2. The indicator cannot be audited. Net new AOV is disclosed only as a direction, never a number. Investors are asked to take a four-quarter trend on trust.
3. The tone has turned defensive. Benioff on the FY26Q4 call: "I did read a tweet that RPO does not matter, but evidently, we have it if it does matter," and "this obviously is not a rational market... This is not our first SaaS Pocalypse." Managements confident in an imminent inflection rarely litigate their own KPIs.

The trade this creates. You are being offered 14.9x on a business whose management has staked its credibility on a dated, falsifiable inflection that no analyst will underwrite, and whose own FY30 target sits above consensus. If they are right, the estimates move and the multiple re-rates simultaneously. If they are wrong, the credibility premium that supports even 14.9x is gone. That asymmetry -- not the low multiple -- is the contrarian case, and it resolves on 2026-08-26.

Six consecutive beats sounds bullish and is largely a mirage for sentiment purposes. Revenue surprises never exceeded ±1% in any of the six -- demand has landed exactly where the street modelled it every single time. The beats came from margin and below-the-line items, and the two largest (+24.0%, +24.9%) coincide with the strategic-investment marks detailed on the financials page. The market has worked this out, which is precisely why the multiple is 14.9x despite the streak.

This is the offset to the de-rating argument, and it is why the score is 7 and not 8. The valuation cushion is real, but a good deal of it has been consumed by the last month's rally. Buying a reset multiple is attractive; buying it after a 21% run, five days before a zero-cushion print, is materially less so.

Not 8, because 76% of ratings remain bullish with only 2 sells, leaving genuine downgrade capacity, and because the stock has already rallied 21% off its 50-day into a binary event -- the washed-out entry point was a month ago, not today.

Not 6, because the de-rating is not cosmetic and, more importantly, because there is a genuine management-versus-street divergence rather than merely a low multiple. Targets have fallen $286 to $222 with coverage thinning to a single recent analyst; the multiple is 14.9x forward with an 8.8% FCF yield; and the sell-side is carrying a FY30 revenue number below the company's own raised $63B target while management attaches words like "certain" and "absolute confidence" to a dated inflection. Compared with the ADSK anchor at 4/10 -- Strong Buy consensus, targets 60% above spot, the bull case already being the base case -- this is the inverse setup: cheap, thinly covered, and with the company's own guidance treated as not credible.

What would move this to 8: the divergence resolving in management's favour on 2026-08-26 without the ratings distribution having caught up. What takes it to 5: the inflection slipping again, which would convert a credibility discount into a credibility problem and remove the floor under even a 14.9x multiple.