Salesforce Shares Fall 32% in 2026: Agentforce and the £20B Buyback

As Salesforce shares fall 32% in 2026 and approach the Q1 FY2027 results (for the quarter ended 30/04/2026) under real pressure: the market wants to see if Agentforce is already paying its way, whether the £20bn buyback is defensive or offensive, and if the AI thesis of the CRM giant can withstand competition. Here is a practical read — numbers, signals and what to expect — for those following the sector from the UK.

TL;DR — what matters before the 27/05/2026 release:

• Salesforce shares fall around 32% in 2026 (ticker CRM on NYSE), with the stock hovering near US$180 ahead of results.
• Wall Street consensus: revenue of US$11.05bn in Q1 FY2027 (+12.5% YoY) and adjusted EPS of US$3.12 (+21% YoY).
• Agentforce ARR closed Q4 FY2026 at US$800m (+169% YoY); the market wants acceleration in Q1.
• £20bn buyback, financed by debt in 8 tranches, is the largest repurchase programme in the company's history.
• Options price a move of 8.7% in either direction after the release.

What happened to Salesforce shares in 2026

I have followed the enterprise software sector for over a decade, and what happened to Salesforce shares in 2026 is not just market noise. It is a structural repricing.

The CRM stock started the year around US$265 and touched US$164 in mid-May. A drop of nearly 35% from the 52-week high — placing Salesforce among the worst performers in the Dow Jones Industrial Average for the year. The partial recovery to around US$180 seen in the week before results was more of a technical bounce than a change in narrative.

The pain does not come from poor operational performance. Salesforce closed fiscal 2026 with US$11.2bn in Q4 revenue (+12% YoY) and US$14.4bn in free cash flow for the year, with an FCF margin of 34.7%. That is operational excellence. However, the market does not pay a premium software multiple for 12% growth — and that is exactly what was repriced.

The psychological effect on investors matters: a stock that was near its all-time high in 2024 has spent 18 months correcting. Those who bought at the top are in their third consecutive quarter of mark-to-market losses. This fatigue shows up in sector ETF flows and helps explain why bounces are short-lived.

The numbers the market expects in Q1 FY2027

When Salesforce reports Q1 FY2027 on 27/05/2026, the consensus of 42 analysts points to:

Metric	Estimate	YoY
Q1 FY2027 Revenue	US$11.05–11.06bn	+12.5%
Adjusted EPS	US$3.12–3.13	+21%
FY2027 Revenue (full year)	US$46.11bn	n/a
Expected price move	±8.7%	(options)

The company has beaten estimates for four consecutive quarters. The risk here is not the quarter's number itself — it is the guidance. Every update to guidance, every comment from CFO Robin Washington on the pace of Agentforce adoption, every cRPO (current remaining performance obligations) figure becomes a trigger for one of those 8% candles that options are pricing in.

The dispersion of targets is unusual for a mega-cap: Bank of America cut to Underperform with a target of US$160, sitting alongside a consensus Buy with an average target of US$274. This shows the thesis is still in dispute — it is not a contrarian view on the stock, but disagreement on when and how much AI will turn into revenue.

Agentforce: the AI thesis that could save (or sink) the quarter

Here lies the whole story. Agentforce is Salesforce's AI agent platform, and the number that matters is its ARR (annual recurring revenue).

Q4 FY2026 results (most recent reference, released in February 2026):

• US$800m Agentforce ARR, +169% YoY
• 29,000 deals closed involving Agentforce in the quarter (+50% QoQ)
• More than 60% of Agentforce + Data 360 bookings came from expansion in the existing base
• Combined with Data 360, ARR exceeded US$2.9bn (+200% YoY), including US$1.1bn from Informatica Cloud
• The platform has processed 20 trillion tokens and converted them into 2.4 billion units of agentic work

Why is this decisive? Because the US$800m Agentforce ARR still represents less than 2% of total annual revenue. If this figure accelerated to US$1.1–1.3bn in Q1, the market re-reads the thesis. If it stagnated in the US$850–900m range, it validates the AI anxiety narrative that has weighed on Salesforce shares all year.

Monetisation of agents is still mixed — part is consumption-based (per conversation/per action executed), part is a seat add-on. Real production adoption is what separates demo from revenue. And that is exactly what Salesforce spent the quarter trying to sell to investors.

For those wanting to understand the parallel in the rest of the market, it is worth reading Google Antigravity 2.0: What Changed in the Agentic IDE and the case of Atlassian in 2026: Layoffs, AI and the Bet on Agents — SaaS companies are all running the same agent marathon, and the market is not being patient with any of them.

The £20bn buyback and the signal behind it

On 16 March 2026, Salesforce announced the largest repurchase programme in its history: £20bn in accelerated share repurchase (ASR).

The details matter:

• Financed by issuance of £20bn in senior notes, in 8 tranches with maturities between 2028 and 2066
• Expected receipt of around 80% of shares on 16 March itself, with final settlement in Q4 FY2027
• Net proceeds estimated at £19.9bn, all allocated to the ASR
• Direct result: lower float, turbocharged EPS even with revenue growing "only" 12%

There are two readings. The bull: the company generates US$14.4bn of FCF per year, so taking on investment-grade debt to buy discounted stock is a rational capital allocation play. The bear: financing a buyback with debt at a company that needs to invest heavily in AI is an admission that organic growth does not justify the current multiple.

For those trading the stock, there is a relevant technical effect: the buyback provides a mechanical floor while the ASR runs. This helps explain why CRM did not break the support at US$164 even during the worst trading days in April.

The real cost of funding

A little-discussed detail: issuing £20bn of debt in 2026, with the US interest rate curve still at elevated levels, compresses net interest margin. Salesforce historically operated with net cash. Now it moves to positive leverage. This imposes greater discipline on capital allocation — which could reduce appetite for large M&A in the short term.

Why the market is sceptical: AI anxiety and competition

As CTO of Agathas Web, I closely follow the SaaS race to plug AI everywhere. The doubt priced into Salesforce shares is simple: will the B2B customer pay for a ready-made agent, or will they build their own?

Three real pressures:

1. Microsoft with Copilot Studio competes for the same corporate automation dollar, with the advantage of already being in Office 365 and Teams at no additional integration cost
2. OpenAI, Anthropic and Google selling APIs directly to companies with strong in-house engineering cuts out the CRM middle layer
3. Vendor lock-in has become an objection: customers want portable agents, not locked into Salesforce's Customer 360

Add to that a macro cycle of tightening in corporate IT (CFOs reviewing SaaS contracts downwards since 2024) and AI Anxiety — the investor's fear that Salesforce's AI infrastructure will not deliver payback before competitors commoditise it — becomes plausible. It is not a manufactured narrative; it is a testable hypothesis, and Q1 FY2027 is the test.

What this means for UK companies

Here is the point that no stock analysis mentions, but that changes real decisions. Many of the clients we serve operate structures that orbit the Salesforce ecosystem (Sales Cloud, Service Cloud, Marketing Cloud) or open source alternatives.

Three practical effects of the fall in Salesforce shares in 2026:

• Pressure for commercial discounts: Salesforce needing to defend expansion of its existing base usually means more room to negotiate renewals and add-ons (especially Agentforce and Data 360, the lines the company wants to flag to investors)
• Window to rethink stack: teams that inherited an expensive CRM in a stuck project now have internal pretext to open an RFP — a topic we cover in Google I/O 2026: What Changes for UK Businesses
• Roadmap risk: if the Agentforce thesis does not take off, it is reasonable to expect portfolio reorganisation, SKU deprecation and roadmap changes — something to consider in long-term contracts

For those structuring digital customer service, it is worth checking how we handle integration with official platforms in the post WhatsApp Web Down: What to Do and How to Protect Yourself — the logic of not depending on a closed platform applies here too.

Scenarios after results: what to monitor

Instead of guessing a price, I prefer to leave a checklist of what to look for when the release comes out on 27/05/2026:

1. Agentforce ARR in Q1: acceleration above US$1bn changes the thesis; below US$900m validates BofA's pessimism
2. cRPO (current Remaining Performance Obligations): the best thermometer of future demand, not reported revenue — it was around US$30bn
3. Non-GAAP operating margin: historical target around 33%; expansion signals discipline, contraction raises an alert on AI costs
4. FY2027 guidance: an upward revision from the US$46.1bn range is the catalyst expected by the buy side
5. CEO Marc Benioff's comments on M&A and Informatica: is the acquisition route still aggressive or on pause?
6. Buyback execution: pace of the ASR and indication of an additional programme would support the stock even in a lukewarm quarter

An operational detail: the conference call is usually at 5pm Pacific (1am BST). It is in the Q&A that the phrase that moves 5% in after-hours trading typically comes out — worth listening to, not just reading the release.

How Agathas Web follows this ecosystem

At Agathas Web, founded in 2008, we have been through three major CRM hype cycles — Web 2.0, cloud, and now AI agents. In each, the right question was not "which stock to buy", but "what does this movement change in the client's budget 18 months from now?". The movement of Salesforce shares in 2026 changes it — and significantly.

For companies that want to seize the moment without locking into expensive closed AI SKUs, we help design pipelines with open and proprietary models, API integration, and data governance. If you want to talk about it, just reach out through our usual channels.

Conclusion

Salesforce shares fall 32% in 2026 not because of operational failure — the company remains highly profitable, with huge FCF and AI ARR growing at triple digits. What has fallen is the market's patience with the gap between agentic platform promise and recognised revenue. Q1 FY2027 is the quarter where that patience is either renewed or becomes a definitive exit. For those making stack or contract decisions in the UK, the window is one of active reading: what comes in the results affects price, but also affects roadmap. Worth watching.