Dreamforce 2026
Positioning Suggestions
→ CRM: the stock re-rated ~13% after the Aug 26 beat-and-raise (post-market $232.32 vs. $205.62 pre-print close) and has since run further to ~$259 as of Sept 8 -- another +11.5% in the two weeks before Dreamforce even opens. The confirmatory-catalyst case from the Aug 26 report is now weaker: more of any plausible good news is already priced, which raises the bar for Dreamforce to extend the re-rate and raises 'sell the news' risk if the conference is roadmap-only. Treat Dreamforce as a low-conviction window unless it produces a genuinely new ARR figure or named displacement.
→ CRM: if Dreamforce surfaces a genuine new ARR figure or a named displacement win that shows Agentforce is expanding wallet share rather than relabeling existing seat spend, that directly weakens the 'SaaSpocalypse' bear case (AI-agent seat cannibalization) -- treat as support for raising conviction toward base/bull on the next research-update pass, not a same-day trade.
→ GOOGL / AMZN: any incremental Salesforce/Google Cloud or Salesforce/AWS-Bedrock consumption commitment disclosed at Dreamforce is confirmatory of an already-priced hyperscaler-consumption thesis for names this size -- avoid sizing trades off conference headlines alone.
→ NOW / MSFT / HUBS / ORCL / ADBE: a real new named displacement loss (beyond McAfee) or a hard ARR number from Salesforce is the first falsifiable negative data point for these names' own competitive framing this cycle. ORCL and ADBE report five days earlier (Sept 10); read their prints first as the enterprise-software sentiment backdrop Dreamforce lands into, then treat any fresh Salesforce competitive claim as an input to the next research-update trigger for the affected name -- not a same-day short. ServiceNow remains the closest watch, compounding into the same agentic budget line from a faster subscription-growth base.
→ Across the list: Dreamforce keynote tone is always upbeat by construction. Do not treat enthusiasm alone as information -- score only concrete numeric disclosures, named customers/contract values, and shipped (vs. demoed) product capability against the Aug 26 baseline.
Predictions
8Salesforce will disclose a combined Agentforce + Data 360 ARR figure or growth rate at Dreamforce that is a fresh number beyond the 'nearly $3.9B, triple-digit YoY' baseline set on the Aug 26 Q2 FY2027 call, not a restatement of it
If a genuinely higher/updated figure appears -> supportive of raising conviction toward base/bull on the next research-update pass; if it's a restatement -> the 'is Agentforce real incremental spend' question stays open into the next earnings print.
Salesforce will name at least one net-new enterprise customer win or competitive displacement (vs. ServiceNow, Microsoft Dynamics, or HubSpot) with a disclosed contract value or seat count, beyond the already-announced McAfee ITSM win
If confirmed with a real contract value -> directly de-risks the 'Agentforce is a marketing story, not real displacement' bear case, and is a genuine incremental negative for the named competitor's own thesis.
Salesforce will announce a capability genuinely new beyond the already-GA'd Agent Fabric (April 2026 TrailblazerDX announcement, GA June 2026) -- e.g. hybrid reasoning or voice-native agents -- with a named GA or beta ship date, not a Dreamforce repackaging of Agent Fabric or a roadmap-only demo
Management will NOT give a specific numeric update on the disclosed softness in Marketing Cloud, Commerce Cloud, or Tableau bookings/renewals during the Dreamforce keynote window
Salesforce will disclose an expanded or new infrastructure/model commitment with Google Cloud or AWS (beyond the existing Gemini/Vertex and Bedrock integrations) at Dreamforce
Confirmatory only for GOOGL/AMZN at their scale -- do not treat as thesis-changing even if confirmed.
Rohan Kumar (President & Chief Platform and Engineering Officer since early August) will take a named, visible keynote or main-stage speaking role at Dreamforce 2026
A visible Kumar keynote role signals continuity/authority in the Agentforce engineering roadmap post-Tallapragada; absence keeps succession-execution risk on the monitoring list into the next earnings print.
The Fin acquisition ($3.6B, announced June 2026) will close before Dreamforce's Sept 15 opening
An early close would let Salesforce fold Fin into the Dreamforce agentic-support narrative with a real (not pending) asset; still-pending at Dreamforce keeps the integration story hypothetical for one more quarter.
CRM shares will move more than 5% (either direction) in the 3 trading days following the Dreamforce opening keynote (2026-09-15), measured from the post-Q2-earnings base ($232.32)
Lowered from the Aug 26 report's 0.25: the stock already added ~11.5% on top of that base before the conference even opened, so less of the plausible good-news move is left to happen inside the 3-day window.
Key Questions
- Does Salesforce disclose a harder combined Agentforce + Data 360 ARR figure or growth number beyond the 'nearly $3.9B, triple-digit YoY' baseline set on the Aug 26 call, or just restate it with conference polish?
- Does management name a net-new large-enterprise customer win or competitive displacement (vs. ServiceNow, Microsoft Dynamics, or HubSpot) with a disclosed contract value, beyond the already-announced McAfee ITSM win -- no new named displacement has surfaced in the two weeks since the print?
- Is there any numeric update -- turnaround or continued drag -- on the softness management has twice flagged in Marketing Cloud, Commerce Cloud, and Tableau bookings/renewals?
- Contentful closed Sept 1, 2026 and Fin remains pending (still guided to close 'during Q3 FY27,' i.e. by end October). Does Salesforce give integration specifics for Contentful inside Headless 360, and does it update Fin's expected close date at the conference?
- Rohan Kumar has held the combined President & Chief Platform and Engineering Officer role since early August (added engineering to the platform title he took June 8) -- does he take a visible keynote/stage role at Dreamforce, or does Benioff/Milano continue fronting technical announcements alone?
- Does Salesforce ship a capability that is genuinely new beyond Agent Fabric (announced April 2026 at TrailblazerDX, GA'd June 2026) -- e.g. hybrid reasoning or voice-native agents -- with a named GA/beta date, or repackage Agent Fabric under Dreamforce branding?
- Is the FY30 'Profitable Growth Framework' (50 by FY30, $60B+ FY30 revenue, set at last year's Investor Day) reaffirmed, updated, or left unaddressed?
- Does Salesforce disclose an expanded infrastructure or model commitment with Google Cloud or AWS (beyond existing Gemini/Vertex and Bedrock integrations) at the conference?
- Oracle and Adobe both report earnings Sept 10, five days before Dreamforce opens -- does either print move the market's read on enterprise-software AI monetization in a way that colors how Dreamforce announcements get received?