ServiceNow's 51% Face-Ripper: Great Story, Terrible Entry Point

Six weeks ago ServiceNow was the poster child for "AI is eating SaaS" panic. Today it's the poster child for "actually, AI is SaaS's best friend." Same company, same quarter, wildly different narrative — and the stock has ripped 51% off its July lows to prove it. That's the kind of round trip that should make you suspicious of anyone claiming they saw it coming, including, frankly, the stock itself.

Let's get the timeline straight, because it matters. ServiceNow reported Q2 on July 22 and got smacked for it — shares fell about 6.5% to $95.46 despite genuinely strong numbers (subscription revenue up 24.5% YoY, AI annual contract value crossing $1 billion, guidance raised). Classic sell-the-news: great quarter, priced-for-perfection stock, investors took the opportunity to exit. By August 10 the stock had already clawed back to $124.88. Then Salesforce dropped a strong print on August 27, the market decided "AI isn't eating SaaS" was the new consensus religion, ServiceNow got name-checked as a prime beneficiary across every CNBC and Bloomberg TV hit for three straight days, and the stock proceeded to melt up another 10%+ in a single session. As of the latest read, we're sitting at $144.71 — up 31% on the month and RSI pinned at 72.7. That is not a normal chart. That is a stock in full FOMO mode.

Here's the thing about the fundamentals underneath all this: they're genuinely excellent, and I'm not going to pretend otherwise. Subscription growth accelerating to 24.5% instead of decelerating (which is what everyone feared after Q1's "SaaSpocalypse" scare) is a real signal, not a fluke. AI ACV blew through $1 billion and management raised the full-year target 50% to $1.5 billion. Agentic AI deployments are up roughly ninefold in nine months. Renewal rate sits at 98%. Total RPO is $29 billion. This is a company executing at a very high level, with 76% gross margins and $5 billion in trailing free cash flow. If you wanted a poster child for "enterprise AI monetization is real, not vaporware," ServiceNow is a reasonable pick.

But — and this is the part the momentum crowd conveniently skips — good company and good stock price are two different conversations. At $144.71, ServiceNow is sitting almost exactly at the analyst consensus target of ~$142, with a forward P/E near 29x and a trailing P/E that was north of 74x just weeks ago. The deep-value window that briefly opened at $95 in July has slammed shut. You're no longer buying a beaten-down growth compounder; you're buying a stock that just ran 31% in a month and is bumping its head against the exact resistance zone ($142-150) where it distributed heavily back in Q1. Overhead supply from trapped longs from the 52-week high near $195 doesn't just disappear because the vibes turned bullish.

There's also a quieter tell buried in the numbers that the AI-euphoria headlines aren't dwelling on: total debt jumped from $2.4 billion to $8.45 billion in a single quarter while cash fell to $2.5 billion. That's a meaningful capital-structure shift, presumably funding M&A or AI infrastructure, and it deserves more scrutiny than it's getting while everyone's busy high-fiving the AI ACV number. Add in 300 layoffs across two California offices effective September 28, framed as an "AI-shift, flat headcount" move, and you've got a company simultaneously spending big and cutting bodies — not necessarily contradictory, but not exactly a signal of comfortable margin trajectory either.

So where does that leave you heading into the next real catalysts — the now[dev] developer conference at month's end and Q3 earnings on October 28? My take: respect the trend repair, don't fight the tape, but don't chase it either. A stock at RSI 73, sitting on top of its own consensus target, up 51% off the lows in under two months, is a stock to trim into or wait out — not a stock to buy at the top of the candle. First support sits around $130, deeper at $120. If the AI-monetization story is as durable as the quarter suggests, there will be a better entry than "everyone already agrees this is great." There always is.