Salesforce Stock in 2026: The Cash Engine Is Real, but Agentforce Still Has to Earn the Rerating

That 8.8% free-cash-flow yield is unusually high for a company that is still trying to convince investors it can build a second growth curve around autonomous AI workflows. It tells you that the market believes the cash engine. It does not yet believe the rerating story.

Source Evidence Snapshot

Source excerpt panel: Salesforce FY26 Q4 results release, rendered 2026-04-17 from the official FY26 highlights showing revenue of $41.5 billion, non-GAAP operating margin of 34.1%, operating cash flow of $15.0 billion, free cash flow of $14.4 billion, current RPO of $35.1 billion, and total RPO of $72.4 billion.

This is why CRM is not being sold because the core business broke. The official FY26 numbers show a company with strong revenue visibility, strong margins, and very large cash generation.

Source excerpt panel: Salesforce FY26 Q4 results release, rendered 2026-04-17 from the official AI highlights showing Agentforce ARR of $800 million, more than 29,000 deals closed, more than 19 trillion tokens processed to date, more than 2.4 billion agentic work units delivered, and combined Agentforce and Data 360 ARR above $2.9 billion.

This panel matters because it moves the AI story out of concept-land. Investors may still debate the final margin profile, but they can no longer say Agentforce is only branding.

Source excerpt panel: Salesforce FY26 Form 10-K, rendered 2026-04-17 from the risk-factor section stating that customers may renew for fewer subscriptions or lower-cost offerings, that consumption-based usage can disappoint, and that the markets and monetization strategies for Agentforce and Data 360 remain relatively new and uncertain.

This is the most important bear-case evidence in the whole stack. Salesforce itself is telling investors that the transition away from simpler subscription economics is not risk-free.

Source excerpt panel: Salesforce March 11, 2026 8-K, rendered 2026-04-17 from the sections describing $25 billion of accelerated share repurchase agreements, expected net proceeds of about $24.885 billion from the related offering, and a $6 billion five-year senior unsecured term-loan agreement.

The buyback is large enough to matter to the stock, but the capital-structure choice matters too. This is a shareholder-friendly action and a leverage decision at the same time.

Source excerpt panel: Public.com CRM stock page and Public.com CRM analyst target page, rendered 2026-04-17 from the public market snapshot showing market cap of $163.93 billion as of April 15, 2026, P/E ratio of 22.74x, 52-week range of $163.52 to $296.05, and analyst target of $283.03 based on 35 ratings.

This is the market's verdict right now. CRM is not being valued like a collapsing asset, but it is also not being valued like a software name that has already won the AI transition.

Source excerpt panel: BLS CPI release for March 2026 and FOMC minutes, March 17-18, 2026, rendered 2026-04-17 from the official releases showing the energy index up 10.9%, gasoline up 21.2%, and the Fed noting both AI-related business-model concerns and Middle East conflict pressure on energy prices and inflation expectations.

This is why a software rerating is harder than it looks. Even a company with strong internal numbers can struggle to earn a higher multiple when inflation and rates are pulling against the sector.

Reader Summary Cards

These cards are generated explanatory visuals based only on the verified facts above. They are not source captures.

The strongest bull case is still the cash engine

It is easy to get distracted by the AI narrative and miss the simpler point. Salesforce already prints enough cash to matter on any valuation framework.

Using Public.com's market cap of $163.93 billion as of April 15, 2026 and Salesforce's official FY26 free cash flow of $14.4 billion, CRM is trading at roughly an 8.8% free-cash-flow yield. Using the same market cap and official FY26 revenue of $41.5 billion, the stock is trading at roughly 4.0x trailing sales.

That is not the multiple of a market darling. It is the multiple of a company the market still treats with caution even after a very strong fiscal year.

The cRPO number sharpens the point. Current RPO of $35.1 billion is about 85% of the entire FY26 revenue base. That does not eliminate risk, but it does make the installed base look much more durable than the stock narrative often suggests.

Agentforce is now a real operating factor, not just an AI headline

The biggest reason to avoid a lazy bear case is that Agentforce has already crossed from story to measurable ARR.

The official FY26 results release said Agentforce ARR reached $800 million, more than 29,000 deals closed in the quarter, and more than 2.4 billion agentic work units had been delivered to date. The same release said combined Agentforce and Data 360 ARR exceeded $2.9 billion.

That does not mean Salesforce has already solved the transition. It does mean the market has to work harder if it wants to dismiss the product cycle as hype.

The pricing structure matters here. Salesforce's public Agentforce pricing page shows examples where a standard action costs 20 Flex Credits, while voice-heavy workflows consume higher credit totals. That supports management's argument that Agentforce is being monetized around work completed rather than around simple seat expansion.

The bull case is that this new work-unit pricing lets Salesforce defend value even if traditional seat growth slows. The market has not fully accepted that argument yet, but the ingredients are now visible.

Why the market still refuses to pay a full AI multiple

This is where the 10-K matters more than the product demos.

Salesforce explicitly says that customers may renew for fewer subscriptions, renew for shorter contract lengths, or switch to lower-cost offerings. It also says consumption-based usage can come in below expectations, and that Agentforce and Data 360 monetization strategies remain relatively new and uncertain.

Those are not vague macro risks. They are the exact risks that can keep a stock cheap even when one quarter looks excellent.

In other words, the market is not saying Salesforce cannot build AI products. It is saying it still wants proof that the new AI revenue pool can become large before the old software multiple disappears.

That is a much narrower and harder debate than the old "Is Salesforce still relevant?" question. Salesforce has already answered the relevance question. It has not fully answered the monetization and durability question.

The buyback helps, but it does not make the thesis easy

The March 11, 2026 8-K is one of the most important documents in the whole setup because it cuts both ways.

On the bullish side, a $25 billion accelerated share repurchase is very large against a company with a market cap of about $163.93 billion. That is roughly 15% of the public market value. If management is right about the durability of cash generation, that kind of buyback can materially improve per-share economics.

On the more cautious side, the same filing shows that Salesforce paired the ASR with fresh debt funding and a $6 billion five-year credit agreement. That does not mean the company is overlevered. It does mean investors should not describe the buyback as costless support.

The practical reading is that management clearly believes the stock is undervalued relative to the internal cash engine. The market, meanwhile, wants to see that balance-sheet flexibility remains intact if the AI investment cycle stays expensive for longer.

The macro tape is still a headwind for software multiples

Even if you like the company-specific story, you still need to respect the backdrop.

The March 2026 CPI release showed an energy shock, with the energy index up 10.9% and gasoline up 21.2% in the month. The March 17-18, 2026 FOMC minutes also said concerns about AI-related business-model disruptions were weighing on some equity prices, while the Middle East conflict pushed up energy prices and near-term inflation expectations.

That matters because Salesforce does not need a rescue. It needs a rerating. Reratings are much harder when inflation and discount-rate pressure are working against long-duration software assets.

This is another reason the stock has stayed controversial. Salesforce can show strong execution and still remain trapped in a valuation holding pattern if the macro regime is fighting the sector.

What matters from here

The next few quarters should answer three questions.

First, can Agentforce and Data 360 continue to scale fast enough to become a clearly material share of the company? Investors do not need perfection, but they do need more evidence that the current ARR base is the beginning of a larger revenue pool rather than just a promising side business.

Second, can Salesforce keep operating margins in the mid-30s while it expands AI infrastructure, go-to-market pressure, and integration work around Informatica? If the company can do that, the market's fear that AI monetization will simply dilute the software model should fade.

Third, does the ASR improve per-share math without creating new balance-sheet anxiety? If the buyback looks smart in hindsight and Agentforce keeps growing, the valuation gap should narrow. If leverage rises while growth proof stalls, the stock can stay stuck.

Bottom line

Salesforce does not look like a broken software company in 2026. The official numbers are too strong for that argument.

The better way to frame CRM is this: it is a real cash-flow story with a real AI growth option, but the market still wants more proof before paying a full premium for that option.

That is why the right rating today looks closer to Outperform than table-pounding Strong Buy. The core business is good enough. The new AI layer is real enough. The rerating, however, still has to be earned.