ServiceNow Just Told The Market Its AI Bet Is Real. The Tape Doesn't Believe It Yet.

Let's get the obvious out of the way: ServiceNow has had a rough year. Through late August it was down roughly 15% YTD and down about 24% over the trailing twelve months. If you bought into the late-2021 software euphoria, you're still underwater. But here's the thing the headline numbers quietly bury — the business underneath the stock isn't breaking, it's accelerating.

Q2 2026 revenue came in at $3.987 billion, up 24% year over year. Subscription revenue, the line that actually matters for a company of this scale, jumped 24.5%, or 23% on a constant-currency basis. Management handed back up its full-year subscription guide to a range of $15.76 to $15.78 billion. That's not a company begging for your attention. That's a company raising the bar while the rest of the enterprise-software complex is still picking shards out of its hair.

Now the part that should make skeptics uncomfortable: AI. Annual contract value from AI crossed $1 billion. Management lifted its full-year Now Assist target by roughly 50%, to $1.5 billion. Customers spending more than $1 million on it grew 130% year over year. Agentic AI deployments rose about ninefold in nine months. More than half of new business is now being priced on consumption rather than seats. Someone is actually buying this thing and spending real money on it.

Let me be the one to say it plainly: monetization is the entire AI story of 2026, and ServiceNow just posted a monetization report card that most AI pure-plays would kill for. The problem isn't the thesis. The problem is the price tag.

Here's where I actually have a view, and it's the crux of the whole thing. ServiceNow trades at a forward P/E in the high 20s to low 30s depending on which source you trust, with a trailing P/E sitting somewhere around 80x. Compare that to Salesforce at a forward P/E of 14.2x and you understand exactly why the market has been content to punish NOW while rewarding the older, slower giant. The street doesn't want to pay up for growth it can't fully verify.

And that's a fair fight, not a freebie. The bear case deserves its weight in the room:

The Q2 GAAP operating margin collapsed to 4.1%, down from 13.3% in Q1 and 16.8% a year ago. That's not a rounding error. Whether it's stock-based compensation or AI-infrastructure spend, management hasn't cleanly reconciled it, and margin is the variable that kills compounders when they get lazy with it.

Device jumped from $2.4 billion to $8.45 billion in a single quarter with cash roughly flat. That's roughly $6 billion of new debt that apparently just appeared. The most charitable read is that it funded the ~$7.75 billion Armis acquisition and the Veza deal, but a capital-structure shift that big deserves an explanation, not a rumor.

The stock is still recovering from what some are calling a "SaaSpocalypse" de-rating rather than earning a genuine rerating on its own merits. The analyst price targets span from $85 to $236 — a 109% spread that says the smart money is genuinely split, not converged.

So where does that leave us? Not at a bottom, and not at a top. At a decision.

Technically, the stock ran from an April low near $81 all the way up to the $144 area in late August — a brutal 51% rally in roughly two months — and has since cooled to the $134 zone as traders took profits. RSI reset from an overbought 72 down to a healthy 54. That's consolidation, not breakdown. The levels that matter are right here in front of us: $130 support on one side, with $140, $145, and the $150–170 supply zone overhead on the other. The easy money off that $81 low is already made.

And then there's the catalyst that changes everything, and it's not far away: Q3 earnings, roughly a month out. That's the print that either validates the AI-monetization thesis or exposes the margin weakness as a trend rather than a quarter. Management has guided Q3 revenue to $3.975–$3.980 billion, so the question isn't whether they'll hit it — it's what the margins look like when they do.

My read? ServiceNow is a best-in-class operating business carrying a best-in-class balance-sheet question and a valuation that only the most convinced buyers will defend at current prices. The fundamentals are an 8 out of 10. The technicals are a 5 — neutral, digesting the run. The growth story is an 8. The risk is a genuine 5.8, meaning the two yellow flags (margin collapse, the debt jump) are real enough to size against, not dismiss.

If you're a long-term buyer, this is a "watch and accumulate near $130" name, not a "chase into $145" name. If you're a trader, you're trading the $130-to-$150 box until October earnings hand you the next directional cue. And if you think the margin story and the debt are solved problems — fine, buy the conviction. But don't confuse a seven-figure revenue number with a seven-figure conviction, when the balance sheet is still trying to explain itself.

The thesis is intact. The patience is the edge. Everything else is noise.