Cantor Fitzgerald raised its Salesforce price target to $300 from $250 on September 2, maintaining an Overweight rating after a company webinar in which executives detailed AI-adoption momentum. The note pointed to current remaining performance obligation growth of 14% against a 13% Street consensus, a 77% gross margin, and management’s raised full-year guidance as evidence Salesforce is positioned to “fully participate in the early days of AI adoption at large enterprises.” At a 0.39 PEG ratio, Cantor argued the valuation still lags the growth reacceleration the Q2 print confirmed.

Salesforce stock reached a new 52-week high of $269.11 this week before closing at $259.23 on September 4, down 1.97% on the session as an unexpectedly strong August jobs report (162,000 nonfarm payrolls added versus 53,000 expected) lifted Treasury yields and pressured equities broadly, including the Dow (-0.51%) and Nasdaq (-0.29%). The pullback looks macro-driven rather than company-specific — CRM is still up roughly 77% from its June 26 low of $146.32. Salesforce also declared its regular quarterly dividend of $0.44 per share, with an ex-dividend date of September 17, the same week as Dreamforce.

Salesforce completed its previously announced acquisition of Contentful on September 1, bringing a composable content platform used by more than 4,800 brands into the fold. Paired with the Qualified acquisition earlier this year, the deal extends Salesforce’s push to let Agentforce agents generate, manage and personalize content and digital experiences directly through Data 360 and Contentful’s composable APIs, rather than depend on separate CMS tooling — infrastructure that becomes more relevant as Winter ’27 pushes agent adoption from opt-in to default.

Salesforce started rolling out Winter ’27’s Agentforce auto-enable feature on existing orgs in the first week of September, after originally targeting an August start. Every org on Lightning Experience with Foundations on Enterprise, Performance, Unlimited or Developer Edition, plus Agentforce 1 Edition, will have the platform turned on by default with the setup toggle removed entirely; new orgs are enabled at creation. Salesforce says there is no additional cost and no billing change, but the shift converts Agentforce from an opt-in decision into an opt-out one just as the company heads into Dreamforce with adoption metrics as its central pitch.

Salesforce confirmed Dreamforce 2026 for the Moscone Center in San Francisco, built around the “Agentic Enterprise” framing, with more than 1,600 breakout sessions, 50-plus keynotes and 240-plus community roundtables; a free Salesforce+ virtual program runs one extra day through September 18. Benioff opens with a keynote alongside unnamed special guests on what it means for humans, agents and platforms to work together. Coverage ahead of the show has flagged the same gap this digest has tracked for weeks: strong ARR and usage-count metrics, but almost no customer-reported cost-to-run or ROI figures for agentic deployments in production.

Snowflake reported fiscal Q2 results on September 2 that beat on both lines, with revenue up 35% year over year and management raising its full-year product revenue forecast to $6.07 billion from $5.84 billion in May — the third straight quarter of accelerating growth. Shares jumped 22% in after-hours trading, capping the busiest stretch of enterprise-software earnings this cycle: Zoom, CrowdStrike, Salesforce and now Snowflake have all reported within nine days of each other, and every one beat estimates.

CrowdStrike’s Q2 revenue grew 26% year over year, and the company raised full-year guidance to $5.99–$6.01 billion; shares jumped more than 11% in extended trading. A day earlier, Zoom beat on both revenue and EPS but fell 3.7% in the regular session and another 4% after hours, because its Q3 revenue guide of $1.275–$1.28 billion sat at or below what analysts already modeled — against 4.9% year-over-year growth, the slowest of the earnings cluster. The split is now a pattern across four straight reports: the size of the beat barely matters next to the direction of the guide.

TD Cowen raised its HubSpot price target by $65 to $285 on September 1 while keeping a Hold rating — a signal that valuation support has improved after the stock’s post-Q2 selloff, without the firm turning outright bullish. HubSpot separately confirmed it will join the S&P MidCap 400 effective September 21 as part of the quarterly index rebalance, with its Analyst Day set for September 17 — the same week as Dreamforce — as the next test of its AI-pricing and go-to-market reset.

Workday shares are down 32.2% since the start of 2026 and trade around $139.60, roughly 44% below the $247.69 52-week high the stock set in September 2025. While Salesforce, ServiceNow, Snowflake and CrowdStrike have all posted double-digit post-earnings pops this cycle, Workday has not, leaving it the clearest counterexample to the sector’s AI-monetization narrative and a reminder that the “guidance over beats” pattern cuts both ways.

Nonfarm payrolls grew by 162,000 in August, triple the 53,000 economists expected, while unemployment held at 4.1%. The surprise strength pushed Treasury yields higher — the 2-year note hit its highest level since January 2025 — on rising odds of a Federal Reserve rate move, and equities pulled back Friday: the Dow fell 0.51%, the S&P 500 slid 0.38%, and the Nasdaq dropped 0.29%. The reaction is a reminder that even the best earnings season enterprise software has had in years is still trading inside a macro rate regime it doesn’t control.

Software-sector analysis from Bulloak Capital puts total market-cap destruction since early 2026 at roughly $2 trillion, the headline number behind the “SaaSpocalypse” framing Benioff has been pushing back on. This week’s results argue the label is an increasingly inaccurate description of the current moment: Salesforce, ServiceNow, Snowflake and CrowdStrike have all posted double-digit post-earnings rallies on AI-driven guidance raises, while Zoom and Workday remain well off their highs. The $2 trillion figure describes a wave that already broke — the current divergence between AI-native winners and legacy laggards is the more useful signal for where multiples go from here.