On August 13, JPMorgan raised its Salesforce price target to $250, arguing the market has overpriced the risk that AI agents cannibalize per-seat CRM revenue and underpriced Agentforce’s own contribution. Shares jumped as much as 4.2% intraday, with the stock ending the session in the high-$190s to low-$200s depending on the data source — the strongest single-day move for CRM in weeks, and a direct rebuttal to Morgan Stanley’s July downgrade thesis.

Evercore ISI’s Kirk Materne cut his Salesforce price target to $250 from $260 on August 13 while keeping an Outperform rating — a trim, not a downgrade, but a sign that even the bulls are tightening their numbers ahead of the print. Paired with JPMorgan’s same-day move, the net sell-side signal is a stock whose bull case is intact but whose margin for disappointment on August 26 has narrowed.

Salesforce reports fiscal second-quarter results after the close on August 26 — a print already carrying the weight of a new COO and platform chief, DoD IL5 authorization, and the VA Missionforce ramp. This week it picked up a fourth variable: whether Agentforce’s Agentic Work Unit billing model, running alongside two other pricing schemes, can show revenue durability even as agent efficiency improves. Current remaining performance obligation (cRPO) acceleration is still the number that decides which story — JPMorgan’s or Morgan Stanley’s — wins the quarter.

Salesforce’s newest Agentic Enterprise Index puts hard numbers on this year’s adoption curve: the average business now runs 13 agents, up from 5 in February 2025; deployment time is down 53%; agent-driven retail sales are up 4x; and customers report over $100 million in annualized savings with a 34% productivity lift. It’s the vendor’s own data, so it’s marketing as much as measurement — but the direction matches what Workday and ServiceNow are separately reporting about agentic ACV growth this same month.

Coverage this week sharpened the structural critique of Agentforce’s Agentic Work Unit pricing: because AWUs bill by task completion, a more efficient agent that finishes the same job in fewer steps can generate a smaller bill even as it delivers more value — the inverse of typical SaaS economics. Salesforce is running per-conversation, flex-credit, and per-user models simultaneously, which reads as healthy experimentation to bulls and as unresolved architecture to bears. August 26 is the first call where analysts get hard usage data to judge which view is right.

Reuters reported August 13 that private equity firm Silver Lake is in talks to acquire Workday, sending shares up as much as 21–25% intraday before trading was halted for volatility; the stock closed near $206.45, a roughly $51 billion market cap, erasing Workday’s year-to-date losses in a single session. No deal is confirmed and talks may not produce one, but it’s a striking read on how private equity values a “structurally challenged” seat-based HCM vendor relative to where public markets had priced it — the same discount thesis weighing on Salesforce.

ServiceNow disclosed its AI-related annual contract value has crossed $1 billion, with agentic deployments up ninefold over nine months and management saying it’s tracking ahead of its target for AI to reach 30% of ACV by 2030. The stock remains about 25% below its 52-week high despite the growth, and CEO Bill McDermott called the quarter “the foundation for a re-rating” — a framing every enterprise SaaS vendor reporting this month, Salesforce included, would like the market to accept about its own numbers.

Multiple market-research estimates now put the SaaS sector on pace to close 2026 somewhere between $375 billion and $465 billion depending on scope, with growth increasingly attributed to AI-enabled tooling, hybrid-cloud adoption, and industry-specific platforms rather than seat expansion alone. Large enterprises are expected to account for roughly 60% of spend this year. The wide range is itself a signal: the market is being re-segmented around AI capability faster than analysts can agree on how big it now is.

Gartner’s 2026 forecast is unchanged from earlier this summer: enterprise software spending rising 14.7% to more than $1.4 trillion, with 40% of enterprise applications expected to carry task-specific AI agents by year-end, up from under 5% a year ago. The forecast keeps holding — the open question, sharpened by this week’s Workday buyout talk and Salesforce’s AWU pricing debate, is which vendors capture that spend as durable revenue versus which get valued as acquisition targets for someone else to fix.