Question: Why NOW rallied during the August 27-28 software re-rate

Verdict: thesis-confirming

What we're asking

Was $NOW's 12.59% gain in the week through August 28 a company-specific development, a generic squeeze, or evidence that the enterprise-software side of the AI-capex-digestion thesis is repairing?

What we found

The tape and event clocks line up. Desk price truth shows $NOW closing at $125.80 on August 26, $138.43 on August 27, and $144.71 on August 28: the move began immediately after Salesforce's August 26 report and continued the next session. Salesforce reported 14% constant-currency cRPO growth, raised its full-year revenue guide, and said Agentforce plus Data 360 ARR was approaching $3.9 billion. Its expanded Anthropic relationship made the same point in product form: incumbent software can be the governed workflow and data layer through which model capability reaches an enterprise.

The independent market read corroborates the sector attribution. Axios reported a 5.2% one-day rise in the software-and-services ETF, naming ServiceNow among its constituents, as investors moved from "AI replaces software" toward "AI increases the value of software workflows." This was therefore not a new $NOW earnings event. It was a cohort re-rate with a clean read-through to $NOW.

$NOW had company-specific support before the move. Its July quarter showed 24.5% subscription-revenue growth, 21% cRPO growth, ServiceNow AI above $1 billion of ACV, and a raised full-year subscription-revenue outlook. The late-August move repaired a narrative and multiple discount around an already strong operating trend; it did not create that trend.

Verdict + reasoning

Structured outcome: thesis-confirming. The governing perspective has argued for rotation rather than a capex reckoning and has been waiting for the broken enterprise-software cohort to produce a fundamentals-led repair. The Salesforce print, the broad software reaction, and $NOW's follow-through are the clearest cohort-level confirmation of that repair so far.

One narrower hypothesis weakens: the July idea that enterprise budgets were being persistently crowded out of software and into AI hardware cannot remain the default explanation for $NOW. It stays a risk to test, not the base case. The move also does not justify calling every software rally durable: the next evidence is follow-through in cRPO and AI ACV, plus whether the sector keeps the re-rate after the initial earnings/positioning shock.

Sources

• ServiceNow Q2 2026 results
• Salesforce Q2 FY2027 results
• Salesforce and Anthropic expanded partnership
• Axios software-sector reaction
• Desk price truth: research/market-engine/data/stocks/NOW/ohlc/30d.json