Salesforce Stock 2027: Agentforce, Claudeforce, cRPO & Valuation

Data status: August 29, 2026. Salesforce stock just had the kind of week that forces investors to reopen a thesis they thought they understood. CRM jumped 22.6% on Thursday after fiscal Q2 2027 earnings and added another 1.57% on Friday to close at $256.00. The easy headline is that AI has rescued software. The more useful conclusion is subtler: Salesforce finally showed evidence that its core demand engine is reaccelerating, but the spectacular $5.90 adjusted EPS figure was far less clean than it looked.

I think this distinction matters because Salesforce has spent the last year trapped between two narratives. In one, generative AI destroys the value of traditional software interfaces and turns established SaaS companies into expensive databases. In the other, AI makes those databases, workflows and permissions more valuable because agents need trusted systems in which they can actually take action. The latest quarter does not settle that debate, but it gives the second narrative much stronger evidence.

The market celebrated $5.90 of EPS. I would start somewhere else.

Salesforce reported $11.3 billion of revenue in fiscal Q2, up 11% year over year and in constant currency. Subscription and support revenue reached $10.8 billion, also up 12% reported. Current remaining performance obligations, or cRPO, rose 14% to $33.5 billion. Operating cash flow increased 71% to $1.3 billion, and free cash flow rose 81% to $1.1 billion.

Those numbers are solid. The headline EPS number was spectacular: non-GAAP diluted EPS of $5.90, up 103% year over year. But investors should not treat $5.90 as if it were a clean measure of recurring quarterly earning power.

Salesforce disclosed $2.613 billion of net gains on strategic investments in the quarter. The company states that those gains increased non-GAAP diluted EPS by $2.53. Subtracting that impact mechanically from reported non-GAAP EPS gives roughly $3.37. That is not an alternative accounting measure and should not be presented as official “core EPS,” but it is a useful way to see how much of the headline beat came from an item that can fluctuate dramatically from quarter to quarter.

This is exactly the kind of detail that gets lost after a 20% stock move. I do not think it makes the quarter weak. I think it tells us where to look for the real signal.

The real signal is cRPO, net new annual order value, Agentforce adoption and management’s confidence in second-half organic reacceleration.

cRPO is the number that made the quarter believable

Salesforce’s cRPO grew 14% year over year in reported and constant-currency terms. Management also guided to approximately 14% cRPO growth in Q3, excluding contributions from the pending Contentful and Fin acquisitions.

For a subscription software company, cRPO is useful because it captures contracted revenue expected to be recognized over the next twelve months. It is not perfect, but it gives investors a cleaner view of forward demand than a single quarter’s recognized revenue. When cRPO accelerates, it suggests the sales engine is improving before all of that improvement appears on the income statement.

Salesforce also said net new annual order value growth was the strongest in four years and that the company remains on track for second-half organic revenue reacceleration. That phrase is important because reported revenue growth includes acquisition contributions. Informatica added $456 million to Q2 revenue, and the full-year guide includes slightly above three percentage points of contribution from Informatica.

Investors should therefore separate two questions. Is Salesforce growing faster? Yes, on a reported basis. Is the core business accelerating organically? Management says the leading indicators point that way, but the next several quarters still need to prove it.

That is the right standard. Salesforce has made large acquisitions before, and acquisitions can make a mature software company look faster without improving the underlying engine. I would rather see a slightly lower headline growth rate supported by improving bookings and retention than a higher growth rate purchased through deal activity.

Agentforce is finally large enough to matter

The AI story also became more concrete. Salesforce said Agentforce and Data 360 annual recurring revenue reached nearly $3.9 billion, up more than 210% year over year. Agentforce ARR exceeded $1.5 billion, up more than 240%, although Salesforce broadened the definition in Q2 FY27 to include AI offerings, Slackbot and Headless 360.

That definitional change is worth noting because fast-growing AI metrics can become less useful when the basket itself changes. Still, the absolute scale is no longer trivial. Salesforce also reported 7.0 billion Agentic Work Units delivered to date across Agentforce and Slack, including 3.2 billion in Q2, up 97% quarter over quarter.

The central question has always been whether AI becomes a new revenue layer or merely a feature Salesforce must provide to defend existing seat-based pricing. These numbers suggest the company is beginning to create a monetizable layer, but I would still watch pricing architecture carefully. Agentic software often shifts economics from per-seat subscriptions toward usage, outcomes or consumption. That can expand revenue if customers deploy agents widely, but it can also introduce more variability than traditional SaaS contracts.

This is where Salesforce differs from a company such as Palantir. Palantir’s AI story has been tied to rapid organic growth and unusually high operating leverage, while Salesforce is trying to reaccelerate a much larger installed base. The opportunity is enormous, but the transformation is more complicated because Salesforce must change monetization without destabilizing the legacy franchise.

Claudeforce changes the strategic argument

One day after earnings, Salesforce and Anthropic announced an expanded partnership called Claudeforce. The initial product, Salesforce in Claude, launches with 37 prebuilt sales skills that let Claude reason over live Salesforce context, automate pipeline updates and take governed actions. Salesforce says the product is available to selected pilot customers now, with an open beta planned for September 2026.

I think this partnership matters more strategically than most feature announcements because it addresses the deepest fear surrounding enterprise software.

The bearish AI thesis says users will increasingly work inside general-purpose AI interfaces instead of navigating traditional SaaS applications. If Claude can answer questions, create reports, update records and coordinate workflows directly, why should an employee open five separate applications?

Salesforce’s answer is essentially: perhaps they will not. But the agent still needs Salesforce underneath.

That is a clever repositioning. Instead of defending the user interface as the source of value, Salesforce is trying to make the data model, permissions, workflows, business rules and auditability the source of value. The interface can become Claude, Slack or something else. Salesforce still wants to be the governed system in which enterprise action is executed.

If that works, AI does not disintermediate Salesforce. It can make Salesforce more deeply embedded because autonomous agents need trusted context and deterministic permissions. Probabilistic models are good at reasoning; enterprises still need systems that know who is allowed to change a price, issue a refund, move an opportunity stage or access sensitive customer data.

This is the most persuasive version of the Salesforce bull case I have seen in some time.

But partnerships with frontier models also weaken one part of the moat

There is a counterargument. If Salesforce becomes increasingly headless, the visible relationship with the user may shift toward Claude, Microsoft Copilot or another AI interface. That can reduce the strategic importance of Salesforce’s own front-end experience.

In a best-case scenario, Salesforce becomes the indispensable transaction and governance layer behind many interfaces. In a worse scenario, it becomes a high-quality data and workflow utility while the AI platform captures the user relationship, pricing power and innovation premium.

That is why I would not automatically celebrate every deep AI partnership. Integration can increase usage and reduce churn, but it can also move bargaining power to the layer that owns the interaction.

The next few years will show whether Salesforce’s “AIforce” architecture makes the company more central or more invisible. Those are not the same outcome for shareholders.

The margin story has improved enormously

One reason the market is willing to give Salesforce more time is that profitability is no longer the embarrassment it once was. Q2 GAAP operating margin was 20.5%, while non-GAAP operating margin reached 34.1%. Full-year FY27 guidance calls for a 20.1% GAAP operating margin and 34.3% non-GAAP margin.

This represents a very different company from the growth-at-any-cost Salesforce of the prior decade. Activist pressure, cost discipline, buybacks and a more mature sales organization have forced management to prove that revenue can translate into shareholder economics.

However, investors should continue distinguishing GAAP and non-GAAP results. Stock-based compensation remains meaningful, acquisition-related amortization matters, and the strategic-investment portfolio can create large swings. The company’s own reconciliation shows why the headline non-GAAP number should not be read as cash in the bank.

Free cash flow is therefore one of my preferred anchors. Salesforce generated $1.1 billion of free cash flow in Q2, up 81% year over year, and maintains full-year free-cash-flow growth guidance of approximately 4% to 5%. The lower full-year growth rate compared with the quarterly jump reminds us that seasonality matters.

The balance sheet is becoming more acquisition-heavy

Salesforce has also changed its balance-sheet profile. At July 31, noncurrent debt stood at $39.3 billion, up substantially from $10.4 billion at the January fiscal year-end. Goodwill was $59.25 billion. The company continues integrating Informatica and has separate pending agreements to acquire Contentful and Fin.

I am not automatically negative on acquisitions. Informatica strengthens Salesforce’s data stack, which is strategically important if agents need access to fragmented enterprise information. Contentful expands content infrastructure, while Fin adds AI customer-service capabilities.

But acquisitions create a recurring analytical problem: they can support the strategic story while obscuring organic performance. Salesforce explicitly says the $200 million increase in FY27 revenue guidance includes $100 million of organic growth, $200 million from pending Contentful and Fin contributions, and a $100 million FX headwind.

I like that disclosure because it lets investors see the bridge. It also shows why the market should not value every dollar of reported growth equally.

What does $256 imply for Salesforce stock?

CRM closed Friday at $256.00 after trading as high as $263.53. The stock is now dramatically above the pre-earnings close of $205.62, meaning investors added roughly 24.5% in two trading sessions.

A large move like that can create the illusion that the valuation story has become obvious. It has not. The stock is simply being repriced from “AI may disrupt legacy SaaS” toward “AI may reaccelerate Salesforce.”

Those are very different narratives, and narrative transitions often produce violent multiple expansion before the full financial evidence arrives.

The bull case is that cRPO acceleration, Agentforce growth, Data 360 and Claudeforce mark the beginning of a multi-year organic reacceleration. If Salesforce can return sustainably to low-to-mid teens organic growth while maintaining 30%-plus non-GAAP operating margins, the stock can justify a materially higher valuation than it carried during the SaaS panic.

The bear case is that Q2 contains too much help from acquisitions, strategic investment gains and favorable enthusiasm around AI. If organic revenue remains stuck in the high single digits after the acquisition effects wash through, the stock could give back part of the rerating.

Three Salesforce scenarios

Why I am more constructive after the quarter

Before this report, the easiest bearish argument was that Salesforce looked like a mature SaaS incumbent trying to attach AI branding to a slowing business. Q2 makes that argument harder.

cRPO at 14%, the strongest NNAOV growth in four years, nearly $3.9 billion of Agentforce and Data 360 ARR, and the Anthropic partnership all point toward a platform that is finding a credible role in the AI stack. The role is not necessarily “build the smartest model.” It is “make models useful inside companies without breaking governance.” That is a valuable position.

It is also a position Salesforce is structurally suited to occupy. The company already sits on years of customer data, pipeline history, service records, marketing interactions, workflows and permissions. AI models without context can generate impressive language. Enterprise agents need context plus authority plus auditability.

Salesforce has those ingredients.

Why I am not chasing the stock blindly

The first reason is the earnings quality. A $2.53 non-GAAP EPS contribution from strategic investment gains is too large to ignore. Investors who annualize $5.90 as if it were recurring operating earnings are building a weak model.

The second reason is acquisition dependence. Informatica contributed $456 million of Q2 revenue, and future guidance includes expected contributions from Contentful and Fin. Organic reacceleration needs to become visible in reported numbers once those acquisition effects normalize.

The third reason is interface risk. If Claude, Copilot or another AI assistant becomes the primary place employees work, Salesforce must ensure that being the back-end action layer still carries enough pricing power.

The fourth reason is valuation reflexivity. A 22.6% one-day move can pull forward months of good news. Even if the thesis improves, the expected return from the new price may be lower than it was before the market recognized the improvement.

What I will watch over the next two quarters

First: organic revenue growth. Management says the second half should reaccelerate. I want to see it in the numbers, not only in bookings commentary.

Second: cRPO. Sustained growth around or above 14% would support the idea that demand is strengthening beneath reported revenue.

Third: Agentforce monetization. I care less about raw interaction counts than about ARR, renewal rates, customer expansion and whether usage economics improve gross profit.

Fourth: Claudeforce adoption. The open beta planned for September gives investors a near-term product milestone. The key question is whether the integration drives incremental Salesforce consumption or simply makes existing workflows easier.

Fifth: free cash flow. Salesforce has become a much more disciplined company. Cash generation should remain the proof that the transformation benefits shareholders rather than only product strategy.

For a broader contrast between application-layer AI and the enormous infrastructure spending required beneath it, our analysis of Amazon’s AWS growth and AI capex is useful. The AI stack needs both layers to earn attractive returns eventually.

Is Salesforce stock a buy after the 22.6% surge?

My answer is more positive than it would have been before earnings, but less enthusiastic than the stock chart might imply.

The quarter gave investors three things Salesforce badly needed: evidence of forward demand acceleration, a credible AI monetization path, and a strategic answer to the fear that general-purpose AI interfaces will replace traditional SaaS applications. Those are real improvements.

At the same time, the $5.90 non-GAAP EPS headline overstates recurring earnings power because strategic-investment gains contributed $2.53 per share. Reported revenue includes material acquisition contributions. And the company still has to prove that second-half organic reacceleration survives beyond one strong bookings period.

So I would not frame Salesforce as a simple post-earnings momentum trade. I think it is becoming a more interesting long-term platform question.

If the future of enterprise software is a world in which employees talk to AI agents instead of clicking through menus, Salesforce may lose visual prominence while gaining infrastructural importance. The data, workflow and permission layer underneath those agents could become more valuable precisely because the interface becomes less deterministic.

That possibility is what changed for me this week.

The market saw a huge earnings beat and bought the stock. I see something more durable but less dramatic: a company that may finally have found a way to make the AI threat part of its moat.

FAQ

Why did Salesforce stock jump after Q2 FY27 earnings?

Investors reacted to 14% cRPO growth, stronger bookings indicators, raised full-year revenue guidance, rapid Agentforce and Data 360 growth, and the expanded Anthropic partnership. CRM rose 22.6% on August 27 and another 1.57% on August 28.

Did Salesforce really earn $5.90 per share from operations?

Salesforce reported $5.90 of non-GAAP diluted EPS, but the company disclosed that gains on strategic investments contributed $2.53 per share to that figure. Those gains can fluctuate and should not be assumed to recur.

How fast is Agentforce growing?

Salesforce said Agentforce ARR exceeded $1.5 billion and grew more than 240% year over year. Agentforce and Data 360 combined reached nearly $3.9 billion of ARR, up more than 210%.

What is Claudeforce?

Claudeforce is Salesforce and Anthropic’s expanded partnership. Its first product, Salesforce in Claude, offers 37 prebuilt sales skills that allow Claude to reason over Salesforce data and take governed actions. An open beta is planned for September 2026.

What is the main risk for Salesforce stock?

The biggest risks are that organic growth fails to reaccelerate, AI revenue merely replaces existing seat revenue, acquisitions mask weak underlying growth, or AI interface providers capture too much of the customer relationship and pricing power.

Sources and further reading:

• Salesforce Q2 FY27 earnings release, Aug. 26, 2026
• Salesforce and Anthropic: Claudeforce announcement
• CRM market price and historical data, Aug. 28, 2026
• The Kapital: Palantir and AI Software Valuation
• The Kapital: Amazon, AWS and the AI Capex Bill

This article is independent financial journalism and is provided for informational and educational purposes only. It is not investment advice or a recommendation to buy or sell securities.

September 2026 update: Claudeforce gives investors a measurable product test

Salesforce in Claude is now available to selected pilot customers, with an open beta expected in September. That turns the strategic argument in this article into something investors can begin measuring rather than merely debating.

The key question is not whether employees like using Claude. It is whether working through Claude increases the value of Salesforce’s underlying data, permissions and workflows. If users can interact through an external AI interface while Salesforce remains the governed system of record and action layer, the company can survive a major interface shift without losing its economic role.

Salesforce says the initial product includes prebuilt sales skills, while Claude is already the default model behind Slack AI, Slackbot, Salesforce in Claude, Headless 360 and Agentforce Coworker. Internally, Salesforce says Slackbot is generating an annualized 8.1 million hours of productivity gains. That number is interesting, but shareholder value still depends on external monetization.

I would track three Claudeforce metrics: beta-to-paid conversion, incremental Agentforce/Data 360 consumption and customer expansion after adoption. If those rise together, Salesforce can credibly argue that general-purpose AI interfaces are pulling more value through its platform rather than disintermediating it.

Source: Salesforce and Anthropic — Claudeforce.