Growth and Agentforce

Growth and Agentforce

With revenue growth now in the high single digits — the deceleration traced in Switching Costs — management has staked its next five years on artificial intelligence reversing the trend. The evidence so far is real but early: an "Agentforce and Data 360" annual recurring revenue base that reached $2.9 billion at fiscal-2026 year-end — up from about $0.9 billion a year earlier, but still only about 7% of revenue. Whether that base scales fast enough to lift the whole company back to sustained double-digit growth is the open part of the durability case.

A decelerating engine, and a target to re-accelerate it

For most of the past decade Salesforce grew revenue at 18% to 25% a year. That rate roughly halved into fiscal 2026: revenue rose 9.6% to $41.5 billion [1], and sell-side consensus has fiscal 2027 near 11% and fiscal 2028 back to roughly 9.6%. The deceleration is not a leakage problem — attrition held near 8% and the contracted book kept growing (the retention floor documented in Switching Costs) — it is an expansion problem: new-logo and upsell growth no longer outrun the law of large numbers on a $41 billion base.

Source: FY2026 10-K Consolidated Statements of Operations for FY2021–FY2026 [2]; FY2027E–FY2028E are consensus estimates.

Against that backdrop, at its October 2025 Investor Day Salesforce set a new long-term goal: revenue of $60 billion or more by fiscal 2030, excluding the Informatica acquisition — a target it frames as "10%+ organic" compound annual growth from fiscal 2026 [3]. Getting from $41.5 billion to $60 billion over four years requires a roughly 9.7% CAGR, so the target is less a promised acceleration than a promise to hold the current pace for half a decade while the mix shifts toward AI. Management pairs it with a "Profitable Growth Framework" — subscription growth plus non-GAAP operating margin summing to 50 by fiscal 2030 [4].

Source: FY2026 revenue growth derived from the FY2026 10-K [5]; FY2030 target and implied CAGR from the Investor Day press release [6].

Where growth actually comes from now

The single-segment income statement hides an important split. Broken out by product family, the mature clouds that built Salesforce — Sales and Service — each grew about 8.4–8.5% in fiscal 2026, and Marketing & Commerce nearly stalled at 2.8%. The one product family growing at a genuinely different rate is Platform & Other, up 22.6% — the bucket that houses Slack, Data Cloud and the emerging Agentforce revenue [7].

Source: derived from FY2026 10-K revenue by service offering, FY2025 vs FY2026 [8].

In the MD&A management attributes fiscal-2026 momentum specifically to "Agentforce, Slack and Data 360" [9]. The read is that Salesforce's growth has become concentrated: roughly two-thirds of revenue (Sales, Service, Marketing) is compounding in the high-single digits at best, and the case for re-acceleration rests almost entirely on the newest, smallest layer scaling into something large.

The AI franchise: real, fast, and still small

That newest layer is now measurable. Salesforce reports an AI-and-data recurring-revenue metric — most recently "Agentforce & Data 360 ARR" — that it began disclosing as the products crossed the billion-dollar mark. At fiscal-2025 year-end (January 2025) it stood at about $900 million, growing 120% year-over-year, alongside 3,000 paying Agentforce customers just 90 days after launch [10]. By mid-fiscal-2026 the combined Data-and-AI figure reached $1.2 billion (again +120%), of which the narrower "agentic AI" slice was about $440 million, and Agentforce customers had grown past 12,000 [11]. By fiscal-2026 year-end (January 2026) the metric reached $2.9 billion, a figure disclosed not in the MD&A but in the proxy, because it was the payout trigger for management's new performance options [12].

Sources: $0.9B at Jan-2025 and +120% growth from the Q4 FY2025 call [13]; $1.2B at Jul-2025 (Q2 FY2026) from the Investor Day release [14]; $2.9B at Jan-2026 from the 2026 proxy [15].

Two facts matter here in tension. The growth rate is extraordinary — roughly a tripling in a year — and the base is trivial: $2.9 billion is about 7% of Salesforce's $41.5 billion of revenue, and the pure "agentic" component is nearer 1%. A product line tripling off a 7% base moves the consolidated growth rate only slowly; for AI to add even two points to company growth, it must contribute roughly $0.8 billion of net-new revenue a year and keep accelerating. The direction is unambiguous; the question is duration and slope.

Management has removed any doubt about how seriously it takes the number by tying its own pay to it. The fiscal-2026 long-term incentive plan introduced performance stock options for every named executive, earned only on "Agentforce & Data 360 ARR" — and only valuable if the stock also clears a $280.62 strike [16]. The ARR came in at $2.9 billion against a $2.0 billion "maximum" goal, so that half of the award vested at 200% of target [17]. The $280.62 strike, set well above the roughly $171 the shares fetch in mid-2026, is worth holding onto: management has already hit its AI-adoption bogey, but the equity that reward rides on is underwater until the stock rises about 64%.

Consumption pricing: the monetization mechanism, and its two edges

Agentforce and Data 360 are sold differently from Salesforce's classic per-seat subscriptions — through a consumption-based model, where customers pay for usage rather than a fixed number of licenses [18]. That design is a direct answer to the loudest bear argument against enterprise software: if AI agents let customers do more with fewer employees, a business priced per human seat should shrink. On the Q3 fiscal-2026 call an analyst put exactly that concern to management — the "very big mismatch" between investors expecting generative AI to be "injurious to the SaaS application layer" and the inflection Salesforce claims to see [19]. Consumption pricing is how Salesforce intends to grow revenue "even in a stable or declining headcount situation": value is billed on agent activity, not employee count [20].

The same mechanism cuts the other way, and Salesforce says so plainly. The company has "limited experience with determining optimal pricing" for consumption contracts, and describes the markets and monetization strategies for Agentforce and Data 360 as "relatively new and uncertain" [21]. More pointedly, its most recent 10-Q warns that "AI-driven consumption models may make it more difficult to optimize our pricing, predict attrition rates, and accurately forecast revenue" [22]. The switching-cost moat that made Salesforce's revenue so predictable was built on multi-year seat subscriptions; a shift toward usage-based billing trades some of that predictability for a larger addressable wallet. That is a defensible trade, but it does give up some predictability.

Re-acceleration signals

Because reported growth has not yet inflected — the fiscal-2026 subscription guide was about 9% in constant currency, plus roughly 80 basis points from Informatica — the case rests on leading indicators [23]. Three are worth watching. First, current remaining performance obligations: organic cRPO was guided to about 9% constant-currency growth for the fourth quarter, with management stating it remains "on track to reaccelerate revenue in twelve to eighteen months" [24]. Second, consumption re-ups: management said 362 customers "refilled the tank" on Agentforce usage in a single quarter, up from three a year earlier — early evidence that usage-based revenue can recur and expand [25]. Third, large structured deals: Salesforce closed 16 multi-year "Agentic Enterprise License Agreements" in the quarter, with roughly 100 more in the pipeline, each a multimillion-dollar commitment [26].

On the reader's underlying question — can you say with high confidence that revenue a decade out will be higher than today — the direction and the rate should be separated. Direction is close to settled: with attrition near 8%, a $72 billion contracted book, and Agentforce driving usage re-ups rather than one-time sales, revenue would have to fall through a retention floor that has held for four years to end lower in ten years. Outright decline is a low-probability outcome. The rate is the genuine uncertainty. The $60 billion fiscal-2030 target is a bet that AI converts a maturing, high-single-digit core into a sustained 10%-plus compounder; today that bet is supported by a franchise growing 100%-plus but still only about 7% of revenue, and contradicted by reported growth that has been stuck near 9% for three years.

The most honest read is that the base is durable and the slope is not yet proven. What would move it toward the bull case is sustained organic cRPO re-acceleration above roughly 11–12% in constant currency, or the Agentforce-and-Data-360 line crossing into the mid-teens as a share of revenue — either would show AI adding points to the top line rather than merely growing fast in the corner of it. The clearest counter-evidence is already on the page: management has been forecasting re-acceleration for several quarters while the reported number has not moved, and by its own admission consumption pricing makes that number harder to predict, not easier.