Why did ServiceNow's stock drop after Now Assist launch?

The honest framing first: the stock didn't drop AT launch — Now Assist went GA in September 2023 with $NOW around $580, and the stock ran roughly 80%+ over the next 12 months, clearing $1,100 by mid-2024. The 'drops' people remember were two intra-cycle pullbacks of 10-15%, not a launch-day reaction.

The first came around Q3 FY24 earnings on Pro Plus pricing transition friction and a public-sector spend pause; the second came in Q1 FY25 when Now Assist attach math came in below the bull-case analyst models. Four narrative shifts compressed the multiple over that window: pricing-transition friction, attach-rate disappointment, federal cRPO drag, and Microsoft Copilot bundling pressure.

By Q4 FY25 / Q1 FY26 the AI-narrative reset and Now Assist deal-count inflection re-rated the multiple back toward the highs. *Not investment advice — historical post-mortem analysis.*

The Timeline

• Sept 2023 — Now Assist GA + Pro Plus launch — $NOW ~$580; initial pop on AI-credibility, Knowledge keynote halo, NVIDIA partnership signal
• Q1 FY24 (April 2024 print) — first attach optimism — stock ran into earnings on bull-case Pro Plus modeling; cRPO beat held the narrative; $NOW pushed toward $815
• Q3 FY24 (Oct 2024 print) — first pullback — 10-12% intra-quarter drawdown on Pro Plus pricing-transition friction commentary + federal spend pause callout; multiple compressed ~3-4 turns of EV/Sales
• Q1 FY25 (April 2025 print) — second pullback — 8-12% drawdown when Now Assist attach commentary came in below the 25% bull-case; analyst models reset to 12-15% attach
• Q4 FY25 (Jan 2026 print) — narrative reset — AI deal-count disclosure, Pro Plus attach acceleration, named flagship Now Assist references; stock began re-rating
• Q1 FY26 (April 2026 print) — re-rate confirmed — McDermott $30B FY30 narrative landed; Now Assist deal-count inflection visible in cRPO; multiple recovered most of the prior 12-month compression

Reason 1: Pro Plus Pricing Transition Friction

• Mid-market sticker shock — the Pro Plus uplift over Pro SKU (typically reported in the 30-60% range depending on workload) created renewal-cycle deferral as customers re-scoped seat counts before signing
• AE quota friction — sales reps carrying both renewal quota and Pro Plus uplift quota faced a structural conflict; field commentary in Q3 FY24 surfaced the GTM-realignment pain
• Named customer downgrades / pauses — analyst channel checks (Goldman, Morgan Stanley post-Q3 FY24) flagged a handful of large-enterprise renewals slipping a quarter as procurement re-evaluated SKU mix
• Pro Plus attach slower than analyst models — sell-side bull cases had modeled 20-25% Pro Plus attach inside the install base by end of FY24; actual prints tracked closer to 10-15%, forcing model resets
• The 'pricing transition tax' pattern — every major SaaS pricing-tier launch (Salesforce Unlimited+, Workday Skills Cloud, Adobe GenStudio) has shown a 2-3 quarter cRPO drag as the install base absorbs the new SKU; ServiceNow was not exempt

Reason 2: Now Assist Attach Below Bull Case

• Analysts modeled 25%+ attach in 12 months — the September 2023 sell-side consensus (Goldman, Morgan Stanley, JPM initiation notes) baked in aggressive AI attach math built on early customer-pilot enthusiasm
• Actual attach tracked 12-15% — by Q1 FY25 prints, management commentary and channel checks suggested attach was running roughly half the bull-case rate; the gap, not the absolute number, was what triggered the re-rating
• Named slow-adopt accounts — large financial-services and healthcare verticals, which dominate ServiceNow's install base, showed measured AI adoption due to data-governance and audit-control review cycles
• The 'Pro Plus IS the AI uplift' reality — investors gradually understood that the ARR lift wasn't coming from Now Assist as a separate SKU — it was coming from customers being moved up to Pro Plus, which bundled the AI capabilities; this compressed the standalone AI-narrative premium
• Pilot-to-production conversion lag — the typical enterprise AI pilot cycle ran 6-9 months in 2024, longer than the 3-6 months sell-side had assumed; deal-count disclosure didn't catch up until Q4 FY25

Reason 3: Public Sector Spend Pause

• Q3 FY24 federal pause callout — management explicitly flagged a slowdown in U.S. Federal contract velocity tied to the FY25 continuing-resolution / budget-uncertainty cycle
• DoD and civilian agency commentary — analyst notes referenced specific delays in DoD and large-civilian-agency expansion deals that had been baked into Q3 / Q4 FY24 sell-side cRPO models
• Impact on cRPO — public sector represents a meaningful slice of ServiceNow's enterprise mix; even a one-quarter pause materially shifts cRPO growth optics for a company priced at premium SaaS multiples
• The federal-cycle whiplash — federal spend rebounded in subsequent quarters as appropriations cleared, but the optics of the pause hit the multiple harder than the actual revenue impact warranted
• Sector-wide drag, not ServiceNow-specific — Salesforce, Workday, and Palantir all flagged similar federal-spend choppiness in the same window, but ServiceNow's premium multiple amplified the reaction

Reason 4: Microsoft Copilot Bundling Pressure

• The 'Microsoft eats workflow AI' narrative — Copilot Studio + Power Automate positioning created sell-side concern that Microsoft would bundle workflow-automation AI into existing E5 contracts at zero marginal price
• Named buyer scenarios — analysts referenced Microsoft-house enterprises (large manufacturing, retail) where Copilot + Power Platform was being pitched as a Now Platform alternative for ITSM-adjacent workflows
• Multiple-compression on competitive risk — even without material revenue impact, the *perception* of Microsoft competitive overhang compressed ServiceNow's EV/Sales by 2-4 turns during the worst of the narrative
• The 'platform of platforms' counter — McDermott's response — that ServiceNow was the workflow layer above the productivity suite, not a competitor to it — took 2-3 quarters to land with sell-side
• Salesforce Agentforce piling on — the late-2024 Agentforce launch added a second competitive-narrative front, compounding the 'workflow AI is contested' story even though early Agentforce traction was modest

What Re-Rated The Stock In Q1 FY26

• Now Assist deal-count inflection — Q4 FY25 disclosure of named, large-deal Now Assist attach references gave sell-side the proof points they'd been waiting for; the deal-count, not the attach percentage, was what shifted the narrative
• Pro Plus attach acceleration — by Q1 FY26, Pro Plus attach was running closer to the 20%+ range; the install-base migration math finally caught up to the original bull case, just on a delayed curve
• Named flagship Now Assist wins — large-enterprise customer references (financial services, telco, federal) were disclosed on earnings calls and at Knowledge events, anchoring the AI-thesis with concrete logos
• McDermott $30B FY30 narrative landing — repeated, consistent CEO framing of the FY30 number on every earnings call gave long-only investors a multi-year anchor; the narrative-discipline itself re-rated the multiple
• AI sector-wide multiple expansion — Q4 FY25 / Q1 FY26 saw broad-based SaaS-AI re-rating as enterprise AI revenue began showing up in actual prints across the sector; ServiceNow rode the tide alongside the fundamental improvement

What Investors Should Take Away

• AI launches don't compress stocks; pricing transitions do — the Now Assist launch itself was a tailwind; the Pro Plus pricing-tier transition was the actual multiple-compression event
• Named precedent: Salesforce Einstein 1 — the Einstein 1 Platform pricing transition in 2023-24 produced an analogous 12-18 month attach-disappointment + multiple-compression cycle before recovery
• The 24-month 'first AI-launch hangover' pattern — across SaaS, the pattern is consistent: launch pop → 12-18 months of attach-math disappointment as analyst bull cases reset to reality → re-rate once deal-count disclosure catches up
• **Watch the *gap* between modeled and actual attach, not the absolute rate** — sell-side models drive the multiple; the delta to consensus is what moves the stock, not the underlying number
• cRPO is the leading indicator — for ServiceNow specifically, cRPO growth deceleration leads the stock by 1-2 quarters; pricing-transition friction shows up there before it shows up in revenue
• Pricing transitions are the underwriting risk every SaaS investor underprices — every major SKU re-tier (Adobe, Salesforce, Workday, ServiceNow) has produced 2-3 quarters of cRPO drag and 10-15% intra-cycle drawdowns; this is structural, not company-specific

Quarter-By-Quarter Stock Movement Table

Pullback Driver To Re-Rate Flow

FAQ

Did ServiceNow's stock actually drop when Now Assist launched? No. Now Assist went GA in September 2023 with $NOW around $580, and the stock ran roughly 80%+ over the next 12 months, clearing $1,100 by mid-2024. The "drops" people remember were two intra-cycle pullbacks of 10-15%, not a launch-day reaction.

The first came around Q3 FY24 earnings and the second in Q1 FY25.

What caused the two pullbacks? The first pullback was a 10-12% intra-quarter drawdown around the Q3 FY24 print on Pro Plus pricing-transition friction commentary plus a federal spend pause callout. The second was an 8-12% drawdown at the Q1 FY25 print when Now Assist attach commentary came in below the 25% bull-case, forcing analyst models to reset to 12-15% attach.

Four narrative shifts compressed the multiple over that window: pricing-transition friction, attach-rate disappointment, federal cRPO drag, and Microsoft Copilot bundling pressure.

How far below the bull case did Now Assist attach actually track? Sell-side consensus from Goldman, Morgan Stanley, and JPM had modeled 25%+ attach within 12 months, baked on early customer-pilot enthusiasm, but actual attach tracked roughly 12-15% by Q1 FY25, about half the bull-case rate.

The gap, not the absolute number, triggered the re-rating. Large financial-services and healthcare verticals, which dominate ServiceNow's install base, showed measured AI adoption due to data-governance and audit-control review cycles.

How did the public-sector spend pause affect the stock? At the Q3 FY24 print, management explicitly flagged a slowdown in U.S. Federal contract velocity tied to the FY25 continuing-resolution and budget-uncertainty cycle, with analyst notes referencing specific DoD and large-civilian-agency delays baked into sell-side cRPO models.

Federal spend rebounded in subsequent quarters as appropriations cleared, but the optics hit the multiple harder than the actual revenue impact warranted. Salesforce, Workday, and Palantir flagged similar federal choppiness in the same window.

Did the stock recover, and what drove the re-rating? Yes. By the Q4 FY25 print in January 2026, AI deal-count disclosure, Pro Plus attach acceleration, and named flagship Now Assist references began re-rating the multiple, and the Q1 FY26 print in April 2026 confirmed it as McDermott's $30B FY30 narrative landed and Now Assist deal-count inflection became visible in cRPO.

Investors also came to understand that the ARR lift wasn't from a separate Now Assist SKU but from customers being moved up to Pro Plus, which bundled the AI capabilities. The article notes this is historical post-mortem analysis, not investment advice.

Bottom Line

ServiceNow's stock did not drop on the Now Assist launch — it ran 80%+ in the year that followed. The 10-15% pullbacks investors remember came in Q3 FY24 and Q1 FY25, driven by Pro Plus pricing-transition friction, attach-rate disappointment versus aggressive sell-side models, public-sector spend pause, and Microsoft Copilot competitive-narrative pressure.

By Q4 FY25 / Q1 FY26 the deal-count inflection, Pro Plus attach acceleration, and McDermott's $30B narrative discipline re-rated the multiple back toward the highs. The lesson that generalizes: AI launches pop stocks; the *pricing transitions* that follow compress them for 2-3 quarters before the deal-count disclosure catches up.

*Not investment advice — historical post-mortem analysis.* *(see also: q1610, q1615, q1616, q1618)*