Salesforce reports second-quarter fiscal 2027 results Wednesday, August 26, after the close, with the call at 5:00 p.m. ET.
Two facts about the same company, both current:
Its artificial intelligence business β Agentforce, Data 360 and Informatica combined β reached $3.4 billion in annual recurring revenue last quarter, up about 200% year over year. Agentforce alone crossed $1.2 billion in ARR, up 205%.
And the stock is the worst performer in the Dow Jones Industrial Average this year, down roughly 56% from its all-time high.
Those are not two views of the market. They are two views of the same disagreement, and Wednesday evening is where it gets settled β or doesn't.
Tuesday's session set the stage. The S&P 500 rose 24.42 points, or 0.32%, to 7,677.25. The Dow Jones Industrial Average added 160.01 points, or 0.30%, to 53,577.17. The Nasdaq Composite gained 171.11 points, or 0.66%, to 26,151.30, leading as chipmakers clawed back part of last week's losses. The VIX fell to 15.45 from 15.85. The 10-year Treasury yield closed at 4.64% and the 30-year at 5.17%, both lower on reports the Treasury may tap its cash balance for expanded long-dated buybacks. Oil broke down hard: Brent settled at $88.58, off 3.9%, and WTI at $82.36, down 3.1%, both one-week lows, after traders judged the new U.S. sanctions on Iran less disruptive to supply than feared.
Salesforce did not participate. It closed Tuesday at $205.69, down 1.61%, on 10.0 million shares β roughly a fifth of a percent above its session low, on a day the index it belongs to went up.
The Number That Actually Matters
It is not revenue and it is not earnings per share.
The Street expects non-GAAP EPS of $3.27 on revenue of $11.33 billion, growth of about 10.4%. Salesforce guided to $11.27β$11.35 billion and $3.25β$3.27. Consensus is sitting inside management's own range, at the top of it. The quarter is effectively pre-announced.
The live number is cRPO β current remaining performance obligation, the contracted revenue expected to be recognized in the next twelve months. Management guided to growth of approximately 14% (13% in constant currency). That line is the bookings signal. It is where a genuine AI-driven acceleration would show up first, months before it reaches the revenue statement.
For most of 2026, it hasn't. That is the whole argument.
The Gap Nobody Has Closed
Look at what Salesforce reported in the first quarter, on May 27:
Revenue $11.13 billion, up 13.3%.
GAAP diluted EPS $2.42, up 52%. Non-GAAP $3.88 against a $3.13 consensus β a 24% beat.
Operating cash flow $6.7 billion, up 3%; free cash flow $6.6 billion, up 4%.
28.6 trillion AI tokens processed, up 152% sequentially. Agentic Work Units up 111% to 3.8 billion.
Half of all Agentforce and Data 360 bookings came from customers who already buy Salesforce, expanding their spend.
Then read the other column. Informatica contributed $444 million of that revenue. Strip it out and organic growth was roughly 8.5%, not 13.3%. Free cash flow growth guidance for the full year was cut to about 4β5% to absorb a $25 billion debt issuance funding a $25 billion accelerated share repurchase. Noncurrent debt went from $10.4 billion to $39.3 billion in a single quarter.
The stock rose on the beat and gave it back within days.
Seven Weeks of Failed Upgrades
This is the pattern that makes Wednesday interesting.
July 21: Morgan Stanley's Adam Wood cut Salesforce to Equal Weight and took his target from $287 to $185, citing exactly the disconnect above β strong Agentforce metrics, insufficient progress in contracted revenue.
August 13: JPMorgan's Samik Chatterjee resumed coverage at Overweight with a $250 target. The stock touched $201.37 that day and closed the week at $196.21 β below where it traded before the note.
August 18: Citi's Tyler Radke raised his target from $187 to $204 while keeping a Neutral rating, on channel checks showing improving Agentforce deployment. The stock popped as much as 3.8% and faded again.
Guggenheim went to Buy at $228. Monness Crespi lifted to $222. Bank of America's Tal Liani reinstated at Underperform with a $160 target back in May, calling the shift an "AI-driven structural reset" from high-growth platform to mature cash generator.
The consensus price target now sits around $243. Twelve months ago it was $348. Every bull call this summer produced a pop and then gave it back, which is the market's way of saying it has stopped accepting narrative in place of a number.
The AI-disruption question underneath all of it is genuinely unresolved, and it is not confined to one company β the argument over who actually wins the AI race is being fought at national scale right now. If agents can do the work that seat licenses used to bill for, Salesforce is either the primary beneficiary or the primary casualty. There is no comfortable middle position, and the stock is priced somewhere between the two.
What the Stock Already Reflects
Salesforce closed Monday at $209.06. Its 52-week range is $146.32 to $269.11. Market capitalization is roughly $169 billion on 819 million shares. The trailing P/E is about 24; the forward P/E is roughly 15.
Fifteen times forward earnings, for a company guiding to 11% revenue growth, a 34.3% non-GAAP operating margin, and a fiscal 2030 revenue target it raised from $60 billion to $63 billion.
Here is the detail that says the most. Management's own full-year fiscal 2027 guidance is $14.06 to $14.12 in non-GAAP EPS on $45.9 to $46.2 billion of revenue. Wall Street's annual consensus is $13.17.
Analysts are modeling roughly a dollar per share below what the company has publicly guided to. That is rare. It means the Street is no longer taking management's forecast at face value β and it means the bar for Wednesday is set by skeptics, not by the company.
Wednesday Is Crowded
Salesforce is not even the biggest event on its own day.
8:30 a.m. ET: July PCE, the Fed's preferred inflation gauge, arrives with personal income, personal spending, durable goods and the second estimate of second-quarter GDP. Consensus looks for core PCE up 0.2% month over month and around 3.3% year over year, with headline near 3.6β3.7% β against a 2% target. Personal spending is forecast up just 0.1%, down from 0.3%, which matters for every consumer-facing name; the softness showing up in retail data has already been running hotter than the headlines suggest.
After the close: Nvidia reports, having guided to $91.0 billion Β±2% against a consensus near $91.8 billion β a company whose beats have been shrinking in size for several quarters running. CrowdStrike reports the same evening, with the Street looking for revenue up 23% to $1.44 billion and EPS up 26% to $0.29.
Thursday: Jackson Hole opens, with Chair Kevin Warsh delivering his first symposium keynote as chair on Friday.
A software company reporting a decelerating-but-in-line quarter into that lineup does not get much benefit of the doubt.
What Lands Next
Four things to watch when the release hits Wednesday evening:
cRPO growth versus the ~14% guide. Above it, the second-half acceleration story survives. Below it, Adam Wood's July thesis gets confirmed in public.
Agentforce ARR and whether the growth rate is still triple digits without acquisition help.
Organic revenue growth ex-Informatica β 8.5% last quarter, and the number that tells you what the core business is doing.
Free cash flow guidance, now carrying $39.3 billion of debt behind a $25 billion buyback.
Salesforce will very likely hit the quarter it guided to three months ago. It usually does. The question the market has been asking all year is a different one: whether the fastest-growing part of this company is big enough, soon enough, to matter to the part that is slowing down.
Found this helpful? Share it with others.
