Ten-Year Outlook

Ten-Year Outlook

This report set out to test whether the cash Salesforce generates a decade from now will be reliably larger than the roughly $14 billion it produces today. The evidence supports a yes: year-ten revenue is very likely higher than today's $41.5 billion, and free cash flow with a little more room, because a ~92% revenue-retention floor, $72.4 billion of contracted revenue, and a base no single customer dominates make a sustained decade-long revenue decline the minority outcome. The case is most sensitive to one variable — whether AI agents erode the per-seat base faster than usage pricing recaptures it. The engine meant to replace displaced seats — pure Agentforce ARR of about $800M, roughly 2% of revenue once $1.1B of acquired Informatica is stripped out — is still a rounding error against the ~95% of $41.5B revenue that is per-seat, even as Salesforce's own 10-K newly names AI-native rivals that can 'displace established user interfaces' and Benioff proves the mechanism by cutting his own support heads. [1][2][3]

The gap the durability case must eventually close is wide. Pure Agentforce ARR of about $800 million is roughly 2% of revenue, while the per-seat base it would have to defend is about 95% of $41.5 billion — some $39.4 billion, a gap of roughly fifty times. The counter sits in the same figures: that $800 million grew 169% year over year [4], 362 customers "refilled the tank" on Agentforce usage in Q3 FY2026 against three a year earlier [5], and the company closed 16 multi-year "Agentic Enterprise License Agreements" in the quarter [6] — so a base compounding at triple digits could reach materiality in two to three years. The one number that would settle the direction is not published: net dollar retention on the Agentforce and Data 360 line is undisclosed, so whether usage revenue is replacing lost seats or merely adding to them cannot yet be read from outside. In the scenario table below, only the Displacement path — where recapture treads water against seat erosion — matches today's ~$170 share price (What the Price Implies).

Sources: FY2026 Form 10-K, MD and A key metrics (revenue, RPO) [7]; Q3 FY2026 earnings call (attrition ~8%) [8].

The base under the question

Whether revenue in year ten clears today's level is, at bottom, a question about net dollar retention held over a long horizon, and three facts already established in this report set the floor.

First, the retention floor. Revenue attrition has sat near 8% for four straight years, through a demand slowdown, a restructuring, price increases and the AI transition (Switching Costs), and Q3 FY2026 confirmed it ended "at approximately 8%" again [9]. Roughly 92% of contract value is retained each year before a single new seat is sold. Second, the contracted book. Remaining performance obligation stood at $72.4 billion at January 2026, up 14%, with $35.1 billion of it recognizable inside twelve months [10] — about 1.7 times a year's revenue already under signature. Third, diversification: no customer is more than 10% of revenue [11], so the decline path runs through tens of thousands of independent renewal decisions, not one.

The honest counterweight sits in the same arithmetic. A 92% gross-retention rate compounds against you if net expansion ever stops: with no new logos and no upsell, 0.92 to the tenth power leaves 43% of today's revenue. Revenue has grown every year only because expansion and new business have added far more than churn subtracts. The mechanism most likely to flip net retention below 100% — customers automating work and buying fewer seats — is real and newly disclosed by the company itself: the FY2026 10-K warns that new AI offerings "may… transform workforce needs and negatively impact demand," and calls the offsetting consumption model "relatively new and uncertain" [12]. That is the risk the recapture engine described in this chapter's opener must eventually offset (Growth and Agentforce).

Three ways the decade plays out

The scenarios below carry one variable forward — whether Agentforce-style usage revenue outruns, matches, or trails the seat erosion its own technology enables. Each path starts from FY2026 revenue and runs ten years; free cash flow is scaled at roughly the ~35% FCF margin the business earns today (Cash Quality), less the incremental interest on the March-2026 debt in the weaker cases.

Source: illustrative paths derived from FY2026 revenue of $41.5 billion [13] compounded at the CAGRs in the table below; not company projections.

Sources: scenario CAGRs applied to FY2026 revenue [14]; FY2030 $60B+ / 10%-organic-CAGR target from the October 2025 Investor Day [15]; FY2028E from the consensus estimates feed.

Three of the four paths end above today's revenue; only the severe-decline tail ends below it, and it requires the displacement mechanism to run unchecked for a decade against a book that renews at 92% and $35 billion of revenue already contracted for the next year. That asymmetry is the substance of the answer: a sustained decade-long decline is the minority outcome, not the base case. Management's own bet sits at the top of the range — a $60 billion-plus FY2030 target implying a 10%-plus organic CAGR, and a "50 by FY30" framework tying subscription growth plus margin to a fixed sum [16] — while Q3 FY2026 gave the first mechanical evidence it might be more than a slide: 362 customers "refilled the tank" on Agentforce usage in one quarter, up from three a year earlier [17], with the company reiterating it is "on track to reaccelerate revenue… in twelve to eighteen months" [18]. Free cash flow is the higher-confidence line of the two: it has risen every year for a decade, capex runs near 1.4% of revenue [19], and even the displacement path leaves cash roughly flat rather than falling, because operating leverage and the capex-light model cushion a slower top line — the offset is the incremental interest on the $25 billion of senior notes issued in March 2026 to fund the accelerated buyback [20] (Capital Allocation).

Reconciling the scenarios with the price

The value in the setup is that the market is priced closer to the bear path than the base. At about $170 the enterprise is worth roughly $170 billion, near 11.8 times FY2026 free cash flow, which discounts to something close to 0% perpetual FCF growth (What the Price Implies). Hold today's ~11.8x multiple and the Fade scenario alone — mid-single-digit growth, no re-acceleration — implies FCF near $25 billion and an enterprise value approaching $290 billion a decade out. Only the Displacement path, where usage revenue merely treads water against seat erosion, roughly justifies the current price. The market-value hit has outrun any deterioration visible in the contracted book or the retention rate, and the price now embeds an outcome the durability evidence puts in the minority.

That is a reconciliation, not a recommendation. The same facts support a fair-value reading if you believe agentic automation inverts rather than decelerates seat-based software — in which case a maturing franchise priced for no growth is priced correctly, and the recapture line never scales. The evidence does not settle that; it locates where the disagreement lives.

What would move the read

The read is falsifiable, and the signals that would raise or lower it are specific, checkable line items rather than sentiment. The most informative disclosure still missing — net revenue retention on the Agentforce and Data 360 line — is not yet published; its absence is why the recapture question stays open.

Sources: attrition and RPO, FY2026 10-K and Q3 FY2026 call [21][22]; Agentforce ARR mix, 2026 proxy [23]; consumption-model uncertainty, FY2026 10-K risk factors [24].

Catalysts, base rates, and universe fit

The next scheduled read is the Q2 FY2027 print on 2 September 2026, where consensus looks for about $11.3 billion of revenue and $3.27 in non-GAAP EPS. The base rates say the swing at that print is not the earnings number. Salesforce has beaten EPS estimates for eight straight quarters, the last three by double digits, yet the shares de-rated more than 50% over the same window — the market is trading the company on current RPO and the re-acceleration timeline, not on margin beats.

Source: reported results versus consensus, per the earnings estimates feed.

Sentiment is constructive but split: about 40 of 52 covering analysts rate the stock buy or better, the mean price target is roughly $245 against a $170 quote, and the range runs from $160 to $475 — a spread that is itself the bull-bear disagreement priced into one tape. Reported short interest is not published in this dataset, so crowding cannot be sized here.

Salesforce is a deeply traded US mega-cap with listed, long-dated options — LEAPS currently extend to January 2028, well beyond a one-year horizon — so it is investable through the long-dated options a multi-year view requires. The term structure that would let an investor act on it — at-the-money implied volatility, open interest and bid-ask depth on the longest-dated strikes, against the realized volatility of the price series — is developed in Options and Volatility.