Salesforce: A Long-Term Owner's View

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?6/10

  A

  Bottom line first: Salesforce's ceiling is not low, but it is not an "unlimited imagination" story either. It is essentially the leader in a market that is already well formed yet still expanding at a double-digit pace, rather than creating a brand-new market from scratch. What Baillie Gifford truly prefers is a company that creates new demand and builds the pie from zero. Salesforce is closer to a mature platform that keeps expanding its share inside an existing pie and extends into adjacent categories. That needs to be stated honestly.

  Start with the size of the pie itself. According to Grand View Research, the global CRM market was about $73.4 billion in 2024 and is expected to grow to about $163.2 billion by 2030, with a CAGR of about 14.6% for 2025–2030. This is a "mature market that is still growing": demand is resilient, since enterprises will not stop managing customer relationships, sales processes, service, and marketing, but the category is already clear and the leading-player structure is stable. It is not an early wild-west phase. The report's characterization of the industry as "a good company in a good industry, but not an industry where management can win while lying down" is consistent with that.

  Salesforce's position in this pie is "making its own slice of an existing pie bigger." According to IDC estimates on a 2024 basis, Salesforce had about 20.7% share of the global CRM applications market, ranked No. 1 for many consecutive years, and generated more than $21.6 billion of CRM revenue in 2024, higher than Microsoft, Oracle, Adobe, and SAP combined. It is worth noting that its share slipped slightly from about 21.7% in the prior year to 20.7%, and its own CRM revenue growth of about 9.5% was slower than the overall market's about 12.8%. That is exactly what makes it look more like an incumbent leader defending its position than a company capturing incremental demand faster than the industry.

  So where is the "new market" story? The real source of a higher ceiling is the AI/agent layer, Agentforce, and the unified data layer, Data 360. The report treats this line as an extension of the moat rather than a disruptive new continent. I agree. On objective data, the FY2026 Q4 results disclosed that Agentforce ARR reached $800 million, up 169% year over year, with 29,000 cumulative deals. That growth is striking, but relative to roughly $41.5 billion of FY2026 full-year revenue, AI is still a minority contributor. It is more like adding a high-value new layer on top of the existing customer base, deepening the same large customers' wallets, rather than opening up a set of entirely new customers that previously did not exist. The report's judgment that "AI is still a minority contribution and needs time before it can redefine the overall growth curve" holds up.

  My ceiling conclusion: its SAM is large enough, growth is not slow, and the company can continue to expand within the same enterprise customers through AI add-ons, more seats, and more modules, a classic case of "going deeper into an existing pie." But it is not the kind of company that creates a completely new market and rewrites the demand curve by itself. Under Baillie Gifford's ten-year, five-bagger framework, the "market is large enough" item barely qualifies; the "creating a new market" item deserves an honest haircut. The ceiling is not a hard constraint, but growth is coming more from penetration and upgrades than from a category being built from zero.

  Jun 10, 2026
• QCan its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?4/10

  A

  Bottom line first: a revenue doubling over the next five years, from about $41.5 billion to $83.0 billion and requiring roughly 15% annualized growth, is almost unrealistic. Salesforce has entered maturity, growth has clearly slowed, and growth over the past year or two has also depended meaningfully on M&A consolidation. This is exactly why Salesforce cannot score highly against Baillie Gifford's hard test of "can revenue at least double over the next five years." It needs to be stated plainly.

  First, how fast is it running now? According to the FY2026 Q4 results, FY2026 full-year revenue was about $41.5 billion, up about 10% year over year, including about $399 million of contribution from Informatica consolidation. Now look at the company's own FY27 guidance: revenue of $45.8–$46.2 billion, up only about 10%–11% year over year, with management explicitly disclosing that about 3 percentage points come from Informatica consolidation. In other words, ex-M&A organic growth is only about 7%–8%. The report's summary that "growth in FY27 guidance still clearly depends on Informatica consolidation, and management itself uses 'reacceleration in the second half' to describe the future cadence" matches the first-hand guidance completely.

  At a realistic growth rate of 7%–10%, revenue in five years would roughly land in the $58.0 billion–$67.0 billion range, far short of the $83.0 billion required for a doubling. Doubling would require growth to reaccelerate to about 15% and then hold there for five straight years. That is inconsistent with the company's disclosed guidance and with the CRM industry's market growth of about 14.6% according to Grand View Research, especially because Salesforce's own growth has been slower than the industry. The report lists "organic growth in the core subscription and support business not falling into the low single digits for long" as a key assumption. Put another way, if even "not falling into the low single digits" has to be defended as an assumption, a five-year doubling is naturally out of reach.

  Will growth be driven by volume, price, or new business? All three contribute, but none is enough to support a doubling:

  • Volume, meaning seats and customer expansion: this remains the base. The FY2026 10-K shows subscription and support revenue at about 95% of total revenue, with cRPO of $35.1 billion (+16%) and total RPO of $72.4 billion (+14%) providing decent revenue visibility. But these metrics themselves are growing in the teens, and cRPO growth also includes 4 percentage points of Informatica contribution.

  • Price, meaning upgrades and AI add-ons: there is some pricing power. The report cites the official price list, from the SMB Starter Suite at $25 per user per month to Agentforce 1 Service at $550 per user per month and the new Agentforce add-on starting at $125 per user per month, showing that Salesforce can charge more for new value layers. But in the installed core modules, enterprise procurement and budget cycles limit blunt price hikes.

  • New business, meaning AI/data: this is growing fastest but from too small a base. The FY2026 Q4 results disclosed Agentforce ARR of $800 million, up 169% year over year, and combined Agentforce and Data 360 ARR of more than $2.9 billion. Relative to a $41.5 billion revenue base, even rapid growth is unlikely to push the total business to a doubling over the next five years on its own.

  My conclusion: a five-year doubling is unrealistic. A more realistic picture is "mid-single-digit to high-single-digit organic growth plus M&A garnish," with five-year revenue more likely to grow 40%–60% rather than 100%. The growth is real, sustainable, and jointly driven by volume, price, and new business, but the cadence is that of a mature software leader, not the "five-year doubling" high growth Baillie Gifford is looking for.

  Jun 10, 2026
• QFive years from now, what will take over as the next growth engine? Does this "second curve" exist today?5/10

  A

  Bottom line first: the second curve does already exist today, and it sits inside the company. It is the closed loop of "AI agent layer (Agentforce) + unified data layer (Data 360) + Informatica data governance," a "data-workflow-agent" loop. But it is still a curve being validated: the base is too small, and part of the growth comes from M&A consolidation, so it is not yet enough to independently take over as the next growth engine. This is exactly where Salesforce can answer "yes" to Baillie Gifford's question of whether a second curve exists today, but cannot answer that it has already taken over.

  What is the candidate second curve? It is not a completely new business. It is deepening existing CRM customers into "AI execution + data unification." Objective progress, all according to the FY2026 Q4 results:

  • Agentforce ARR reached $800 million, up 169% year over year, with 29,000 cumulative deals, nearly 20 trillion tokens consumed, and more than 2.4 billion "agent work units" generated;
  • Combined Agentforce and Data 360 ARR exceeded $2.9 billion, including about $1.1 billion of Informatica Cloud ARR;
  • Data 360 has processed more than 112 trillion records cumulatively;
  • All top 10 orders in FY2026 Q4 included the combination of Agentforce 360, Data 360, and the core platform, and more than 60% of quarterly Agentforce and Data 360 orders came from expansion among existing customers.

  The point that this comes from existing-customer expansion is crucial. It shows that the second curve's business model is "deepening the wallet inside the same large enterprises." It shares the same customer base, sales motion, and ecosystem as the first curve, CRM seat subscriptions. That is its advantage, with lower customer acquisition costs and faster takeoff, and also its limitation, because the ceiling remains constrained by the same customers' budgets and it is not opening an entirely new customer group. The report's moat characterization as "stable to slightly widening, with the slight widening coming from further bundling of Data 360, Slack, Informatica, and Agentforce" is consistent with this.

  Why has it not yet "taken over" today? Compare scale: second-curve-related ARR totals about $2.9 billion, while FY2026 full-year revenue was about $41.5 billion according to the FY2026 10-K, less than 7%. Even if it maintains high growth, it will be hard to single-handedly reverse the company's overall roughly 10% growth cadence over the next few years. More honestly, about $1.1 billion of the combined ARR comes from Informatica consolidation, which is bought rather than purely incubated organically. The FY27 guidance for about 10%–11% growth also includes about 3 percentage points from Informatica consolidation, and management uses "second half (H2 FY27) reacceleration" to describe when the second curve should truly start contributing. In other words, the curve's "organic reacceleration" still needs observation. The report is right to list it as one of the "biggest uncertainties."

  My conclusion: the second curve is not a story drawn in the air. It already has real ARR dollars running today, and the direction, a data plus AI execution loop, fits the long-term migration in enterprise software. But today it is a "high-value extension of the first curve," not an independent new growth pole, and it depends partly on acquisition contribution. Whether it can move from "exists" to "takes over" depends on whether Agentforce/Data 360's organic expansion can truly pull up companywide growth and margins. That is exactly the most important tracking metric in the report.

  Jun 10, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

  A

  Bottom line first: Salesforce's core competitive advantage is "high switching costs + compound bundling of platform, data, and ecosystem." The moat is real and fairly deep. Over the next three to five years, my judgment is "overall stable and slightly wider at the margin, but not unconditionally wider," because AI raises entry barriers for some new workflows while also lowering the threshold for bypassing some old software layers. Baillie Gifford wants a moat that will widen in the future. Salesforce offers "a stable core, widening incrementally, but with an AI substitution headwind at the same time." It needs to be analyzed layer by layer, not generalized.

  The core source of the moat, the hardest layer, is switching cost. Once an enterprise builds its sales processes, service cases, automation rules, permission system, reports, custom objects, API integrations, partner plugins, and internal training on Salesforce, the switching cost is no longer the license-price difference. It becomes business interruption + data migration + redevelopment + regovernance. Objective support: the FY2026 10-K shows subscription and support revenue at about 95% of total revenue, and the vast majority of subscription contracts are noncancelable. The report cites company-disclosed attrition of about 8%, quite low for a large software company, which corroborates the "strong switching cost" view. This layer is very likely to remain rock solid over the next three to five years.

  The part that can widen, the incremental layer, is deepening the same customers into the "data-workflow-agent" closed loop, making the bundle thicker. According to the FY2026 Q4 results, Data 360 has processed more than 112 trillion records cumulatively, combined Agentforce and Data 360 ARR exceeded $2.9 billion, and more than 60% of related quarterly orders came from existing-customer expansion. The more data customers deposit in the platform, the more agents they train, and the more systems they connect, the harder migration becomes. Add Slack, MuleSoft, Tableau, Informatica, and the AppExchange ecosystem, and this becomes an indirect network effect where customers become harder to leave the more they use it. The direction is wider.

  Brand and ecosystem are medium-strong and stable. According to IDC on a 2024 basis, Salesforce had about 20.7% global CRM applications share, ranked No. 1 for many consecutive years, and exceeded Microsoft + Oracle + Adobe + SAP combined. In enterprise procurement, being a default shortlist candidate is itself a barrier. The ISVs on AppExchange and the system integrators built around Salesforce form channel and ecosystem advantages that new entrants cannot replicate quickly. The report's statement that a competitor would need "years and billions of dollars of capital" to replicate it is a reasonable scale judgment.

  The parts that may narrow or come under pressure, which must be listed honestly:

  • AI substitution risk is a real headwind. In the FY2026 10-K, the company itself acknowledges that the market is "highly competitive, rapidly evolving, fragmented, and subject to low barriers to entry," and directly says generative AI and autonomous agents may change traditional workflows and user interfaces, allowing some old software layers to be bypassed. In other words, AI is both a tool for widening its moat and a potential wall-breaking hammer.
  • The competitive boundary is expanding rather than shrinking. Salesforce spans too many categories. In CRM/sales automation it faces Microsoft, Oracle, SAP, and HubSpot; in workflow automation it runs directly into ServiceNow; in data/AI it cross-competes with Snowflake, Databricks, the Microsoft ecosystem, and the OpenAI toolchain. The report cites recent sell-side views that AI is deepening competitive overlap between Salesforce and ServiceNow. The direction is higher competitive intensity.
  • Scale/cost advantage is only moderate: the multitenant architecture brings delivery-cost sharing and high gross margin, with FY2026 gross margin of 77.7% according to the FY2026 10-K, but the advantage does not show up as the lowest price. Legal barriers, such as patents and licenses, are weak to medium. The real moat is the platform, data, workflows, and ecosystem, not exclusive rights.

  My conclusion: over the next three to five years, the moat's foundation, switching costs + brand + ecosystem, is solid, and Data/AI bundling should make it slightly wider at the margin. But during the same period, AI also lowers barriers in some layers and the competitive boundary is expanding, so this is "net widening, with headwinds," not Baillie Gifford's ideal of "unconditional, accelerating widening." The report's moat strength score of 4/5 and description of "stable to slightly widening" are fair. Salesforce should neither be mythologized nor underestimated.

  Jun 10, 2026
• QIf the core business is disrupted, does it have the genes to reinvent itself? How does it deal with mistakes and bad news?5/10

  A

  Bottom line first: Salesforce has a fairly strong "self-reinvention gene." Historically, it has repeatedly redrawn its boundaries through acquisitions and platform expansion, from a single CRM product to integrated MuleSoft, collaboration through Slack, analytics through Tableau, data through Data 360/Informatica, and then AI agents through Agentforce. It has also shown fast correction when profitability came under pressure. But its attitude toward "mistakes and bad news" is "above average, but not cleanly demanding enough": disclosure is fairly candid, yet capital-allocation correction, especially high-price buybacks, has not been pretty. Baillie Gifford's question here is whether the company can rescue itself if the core business is disrupted. Salesforce's answer leans positive, but deserves discounts for "rewriting itself by buying" and for key-person dependence.

  Start with the evidence of self-reinvention. Facing the generative AI wave that may disrupt traditional software layers, Salesforce did not cling to its old CRM base. It redefined itself as a trusted platform that unifies "people + applications + data + AI agents"; the FY2026 10-K positions the company as a global CRM technology leader and emphasizes the Agentforce 360 Platform. In numbers, the FY2026 Q4 results show Agentforce ARR reaching $800 million within one year, up 169% year over year, proving that it does have the ability to grow the AI layer into its products before its own moat is bypassed by AI. The report's judgment that AI is both "a tool for widening the moat and a potential wall-breaking hammer," and that the company has chosen to embrace it proactively, is defensible. That is exactly what the "self-reinvention gene" looks like.

  Now look at its execution in correcting under pressure. FY2023 was the year of severe margin pressure and intense outside shareholder pressure. The company then launched aggressive cost control, organizational adjustments, and a profit reset. The objective result, according to the standardized financials cited in the report and comparable with the FY2026 Q4 results, was that GAAP operating margin rose all the way from about 2.1% in FY2022 to 20.1% in FY2026, while free cash flow rose from about $5.3 billion to $14.4 billion (+16% Y/Y). Raising the margin structure by an order of magnitude in two to three years itself proves that the organization can hear bad news and then actually change.

  How does it treat mistakes and bad news? My assessment is "candid enough, not demanding enough":

  • The candid side: the report notes that management did not avoid growth slowdown, uncertainty around AI monetization, or capital-return arrangements in earnings releases and the proxy. In FY27 growth guidance, it voluntarily disclosed that "about 3 percentage points come from Informatica consolidation" and described the cadence honestly as "second-half reacceleration" (FY27 guidance). On governance, it also responded to shareholder dissatisfaction. According to the 2026 Proxy, say-on-pay support rose from 46% in 2024 to 77% in 2025, after the company revised executive-compensation design and added metrics tied to margins and growth. That shows it does correct "bad governance signals."
  • The insufficiently demanding side: the clearest weakness in correction is capital allocation. The FY2026 10-K discloses 50 million shares repurchased in FY2026 at an average price of $254.21 per share, and an FY2025 average price of $260.12 per share, while the current share price is about $171. Large-scale buybacks concentrated at prices clearly above today's price show that the company is better at "delivering on a return commitment" than at being contrarian and buying heavily only when shares are cheap. This is not hiding bad news, but it is an inertia of "knowing the timing is not pretty and still executing the plan."

  Two structural risks must be deducted. First, self-reinvention depends heavily on acquisitions, MuleSoft/Slack/Tableau/Informatica, at the cost of about $57.9 billion of goodwill on the FY2026 10-K and negative tangible net worth. Reinvention achieved by "buying" has thin balance-sheet cushion and real integration risk. Second, founder Marc Benioff remains the core of the company's spirit and external image, creating some key-person risk. The report judges that this is "not to the point where the company cannot operate without him," and I agree, but the risk is not zero.

  My conclusion: the self-reinvention gene exists and has been tested in the AI cycle; its handling of bad news is candid in disclosure and willing to correct governance, but weak in self-criticism around capital allocation. Overall, this is a company that can rescue itself and has proven it, satisfying Baillie Gifford's resilience requirement, but discounts are needed for "reinvention through acquisitions," "slow correction on high-price buybacks," and "key-person dependence." It does not reach the very top tier.

  Jun 10, 2026
• QDoes management, especially the founder, have a long-term view and deeply aligned interests with the company? Is it willing to sacrifice current profits for five to ten years from now?5/10

  A

  Bottom line first: founder Marc Benioff unquestionably has a long-term view, having led the company since 1999 and remaining core to its spirit and external image. But "deeply aligned interests with the company" is only medium. His personal ownership is about 2.8%, far from the founder level where almost all net worth is tied to the company and alignment is extremely strong. As for whether management is willing to sacrifice current profits for five to ten years from now, recent facts point the other way: under outside shareholder pressure, the company sharply lifted current margins and implemented systematic buybacks and dividends. It looks more like "making concessions to current returns" than "sacrificing the present for the long term." Baillie Gifford favors founders with deeply aligned interests who are willing to sacrifice the near term for distant outcomes. On this question, Salesforce has enough vision, medium alignment, and weak willingness to sacrifice, so it needs an honest haircut.

  Start with alignment. According to the beneficial ownership table in the 2026 Proxy, Benioff held 22,791,882 shares, representing 2.8%; current directors and executives as a group held about 3.5%. This means management is not without skin in the game, but compared with founder-controlled companies with very high ownership, the degree of alignment is only medium. The report's wording that "one cannot say interests are aligned to an 'extremely strong' level" is accurate.

  Then consider the governance discount for long-term investors. Benioff serves as both Chair and CEO. The 2026 Proxy shows that aside from Benioff, co-founder Parker Harris, and Robin Washington, all other director nominees meet NYSE independence standards, and 77% of the board is independent. Overall independence is not weak, but the combined Chair and CEO role is a natural discount for conservative investors. More telling is the passive correction feature of governance: say-on-pay support fell as low as 46% in 2024, and only after the company revised compensation design and added "north star" metrics tied to margin and growth did it recover to 77% in 2025. This shows that the company responds to long-term shareholder demands, but often under pressure rather than through proactive self-discipline.

  The key question: is it willing to sacrifice current profits for five to ten years out? Based on actions over the past three years, the answer leans "no." Baillie Gifford's ideal founder would depress current profits and go all-in on reinvestment when facing a high-growth opportunity. Salesforce has taken the opposite route. After FY2023 margin pressure, it chose to sharply lift current margins, with GAAP operating margin rising from about 2.1% to 20.1% in FY2026 according to the FY2026 Q4 results, and returned substantial cash to shareholders. The FY2026 Q4 results disclosed $14.3 billion returned to shareholders in FY2026, including $12.7 billion of repurchases and $1.6 billion of dividends. In February 2026, it raised repurchase authorization to $50.0 billion and the quarterly dividend to $0.44 per share, and in March 2026 launched its largest-ever accelerated share repurchase of about $25.0 billion, or about 103 million shares (Salesforce announcement). This package is standard for an excellent mature software company, but the orientation is "prioritizing current returns," not "sacrificing the present for the future."

  The two sides of capital-allocation rationality also support the view of "medium alignment and discipline below the top tier." The good side is a clear return path. The bad side is specific: the FY2026 10-K shows an FY2026 buyback average price of $254.21 per share and an FY2025 average of $260.12 per share, both clearly above the current share price of about $171, showing that management is not good at contrarian repurchases. The acquisition gene is heavy, with historical deals leaving about $57.9 billion of goodwill and negative tangible net worth. Equity compensation remains meaningful: the report cites FY2026 SBC expense of $3.509 billion and $6.042 billion of unrecognized SBC remaining as of the end of January 2026. Long-term shareholders cannot treat it as a zero-cost item.

  My conclusion: long-term vision is sufficient; interest alignment is medium, with 2.8% ownership and no control-level stake; willingness to sacrifice the present for the distant future is weak, as the company has clearly shifted in recent years toward a mature model of "reinvestment + systematic shareholder returns." The report's management credibility score of 3.5/5 and capital-allocation rationality score of 3/5 are fair. This is a credible management team that will correct course, but on Baillie Gifford's most valued standard of "deep founder alignment + willingness to depress current profits for ten years out," Salesforce cannot score highly.

  Jun 10, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or regulation?6/10

  A

  Bottom line first: if Salesforce disappeared tomorrow, its large-enterprise customers would miss it considerably, but not in a fatally irreplaceable way. It is deeply embedded in customers' sales, service, marketing, and data processes, and switching costs are extremely high. That is real indispensability. But it is still an enterprise productivity tool, not critical infrastructure whose disappearance would cause society to stop functioning, and Microsoft, Oracle, SAP, ServiceNow, and others can provide substitute categories. Its growth model is highly healthy and sustainable, and does not depend on harming society or stepping over regulatory lines. Baillie Gifford asks for both indispensability and social/regulatory sustainability. Salesforce is close to full marks on the second dimension, clean sustainability, and above average on the first, "customers would miss it a lot, but substitutes exist."

  Start with "how much would they miss it," the indispensability layer. The evidence is switching cost and stickiness: the FY2026 10-K shows subscription and support revenue at about 95% of total revenue, with the vast majority of subscription contracts noncancelable, and the report cites company-disclosed attrition of about 8%, very low among large software companies. Customers build sales processes, service cases, automation rules, permissions, reports, custom objects, API integrations, partner plugins, and training systems on Salesforce. If it disappeared, they would not face "changing software." They would face business interruption + data migration + redevelopment + regovernance. That is the substance behind "they would miss it." The report is right to judge switching cost as a "very strong" moat layer.

  But "missing it" does not mean "irreplaceable." Two points must be stated honestly. First, it spans too many categories, and every category has strong substitutes. The FY2026 10-K itself lists competitors including Microsoft, Oracle, SAP, HubSpot, ServiceNow, and AI-native startups, and acknowledges that the market has "low barriers to entry" and that generative AI and autonomous agents may bypass traditional workflows. Second, it is a front-office enterprise productivity platform, not a critical node like the power grid or payment clearing, where being cut off would halt society. Customers would suffer and spend one or two difficult years migrating, but society would not stop functioning because Salesforce disappeared. That makes its indispensability "strong commercial stickiness," not "civilization-level critical infrastructure."

  Now consider whether the growth model is sustainable and not dependent on harming society or regulation. Salesforce performs very well here. Its money comes from voluntary enterprise purchases, renewals, expansion, and cross-selling of multiple products under a subscription model. The FY2026 10-K says professional services are mostly used to support adoption and drive larger subscription contracts. It is not regulatory arbitrage, not consumer harm, and not monetization through privacy erosion. Quite the opposite: one of its selling points in the AI/data era is helping customers unify and use data within a trusted, governable, permissioned framework, through the combination of Data 360 + Informatica + Hyperforce. Data compliance and security are part of the moat, not a liability. The report also does not list any major regulatory litigation or social externality risk as a current risk. Its risks are competition, AI substitution, acquisition integration, and valuation rerating, not "growth built on harming society or crossing regulatory red lines."

  Customer structure also supports the health of the business. The FY2026 10-K discloses customers across almost every industry, paid subscriptions per customer ranging from 1 to hundreds of thousands, and no single customer accounting for more than 10% of revenue in FY2026/FY2025/FY2024. This is a broadly penetrated, diversified business that does not depend on a few large customers or gray areas. Its growth sources are both diversified and clean.

  My conclusion: on the second dimension, social/regulatory sustainability, it is excellent: clean growth, voluntary payment, and compliance as a selling point. On the first dimension, indispensability, it is very strong but not irreplaceable: customers would miss it deeply and migration would be painful, but substitute categories exist and it is not societal critical infrastructure. Overall, it is "a sticky platform customers would genuinely miss, with a monetization model that passes ethical and regulatory scrutiny." That satisfies Baillie Gifford's health requirement for long-term holding, though it leaves one notch of room below the highest standard of "irreplaceable."

  Jun 10, 2026
• QWhat are the unit economics of this business, including gross margin and incremental returns? Do they improve or deteriorate as scale grows? Where does the money it earns go?7/10

  A

  Bottom line first: Salesforce's unit economics are excellent on gross margin, cash generation, and capital intensity: high gross margin, asset-light operations, and strong cash flow, and these metrics improve as scale grows. But "incremental returns," using accounting ROIC, are only medium and have not reached the top tier of software companies, mainly dragged down by large M&A goodwill and ongoing stock-based compensation. The money it earns mainly goes to two places: ongoing operating investment in sales, R&D, customer success, and similar functions, and large-scale buybacks plus dividends returned to shareholders. Baillie Gifford asks whether unit economics improve or deteriorate with scale and where the money goes. Salesforce's answer is "core unit economics improve, but incremental capital returns are diluted by acquisitions and SBC, and a growing share of cash flow is directed to shareholder returns rather than high-return reinvestment."

  Start with gross margin and cash: excellent and improving with scale. According to the standardized financials cited in the report, comparable with the FY2026 Q4 results, gross margin rose from 73.5% in FY2022 to 77.7% in FY2026; GAAP operating margin rose from 2.1% to 20.1%; and FCF margin rose from about 20% to about 34.7%. The multitenant architecture, with the FY2026 10-K describing one instance serving many organizations and delivery costs spread over a larger user base, is the structural source of "the larger the scale, the better the margin." Capital expenditure is extremely light. The FY2026 Q4 results show FY2026 capex of only $594 million, about 1.4% of revenue, with OCF as high as $15.0 billion and FCF of $14.4 billion. In one sentence: it is a classic "high gross margin, sales-heavy, capital-light" software business. Growth does not require large fixed-asset investment, which is the root of its friendly unit economics.

  But "incremental returns" need an honest discount: ROIC is medium, not top tier. The report cites comparable data showing Salesforce ROIC at about 7.6%, below SAP at about 13.2% and Microsoft at about 22.0%, and close to Oracle at about 9.4% and ServiceNow at about 7.9%. Why does such a high gross margin not translate into top-tier ROIC? The root is M&A: the FY2026 10-K shows about $57.9 billion of goodwill and about $6.8 billion of other intangible assets, with negative tangible net worth. Large amounts of capital are embedded in "bought" assets, enlarging the denominator and diluting returns on incremental capital. The report lists capital allocation as the area where it is "most reserved" on Salesforce, and that matches the lowish ROIC fact.

  Where does the money specifically go? Three buckets:

  • Ongoing operating investment, the largest item: the cost structure is channel + R&D. The report cites FY2026 sales and marketing expense of $17.345 billion and R&D expense of $5.993 billion. The main costs are not factory raw materials, but sales systems, R&D, customer success, and ecosystem/infrastructure investment. This is necessary reinvestment to sustain growth and the moat.
  • M&A: the cash-flow statement in the FY2026 10-K shows about $9.27 billion of net cash outflow in FY2026 for "business combinations, net of cash acquired," mainly Informatica. This directs cash toward "buying growth/capabilities." Strategically, there is data-layer synergy, but the cost is goodwill accumulation and negative tangible net worth.
  • Shareholder returns, increasingly large: the FY2026 Q4 results disclose $14.3 billion returned to shareholders in FY2026, including $12.7 billion of repurchases and $1.6 billion of dividends. In February 2026, the company raised repurchase authorization to $50.0 billion and the quarterly dividend to $0.44 per share. But buyback quality is not pretty: the FY2026 10-K shows 50 million shares repurchased in FY2026 at an average price of $254.21 per share, and an FY2025 average of $260.12 per share, both far above the current share price of about $171.

  There is another "real economic cost" that cannot be ignored: stock-based compensation. The report cites FY2026 SBC expense of $3.509 billion and an unrecognized balance of $6.042 billion. This means that under an owner-earnings lens, SBC should be deducted from $14.4 billion of FCF, giving conservative owner earnings of about $10.9 billion. The unit economics are good, but reported FCF cannot all be treated as distributable cash. This deduction is necessary when evaluating incremental returns.

  My conclusion: the unit economics are top-class on gross margin, cash, and asset-light structure, and improve with scale. But incremental capital returns, ROIC, are diluted by large M&A goodwill and SBC, and are only medium. The money earned is partly reinvested efficiently in operations, partly put into M&A, where there is synergy but thin balance-sheet cushion, and increasingly returned to shareholders through buybacks and dividends, directionally right but often at expensive buyback prices. The report's conclusions, "can generate stable free cash flow: pass" and "whether capital returns are excellent: uncertain," fully match the facts above.

  Jun 10, 2026
• QWhat conditions must hold simultaneously for it to rise fivefold in ten years? Are those conditions realistic? What expectations are implied by today's share price?3/10

  A

  Bottom line first: for Salesforce to rise fivefold in ten years, about 17.5% annualized, three engines, revenue, margin, and valuation multiple, all need to move up simultaneously and approach an optimistic scenario. Given its maturity today, that combination is not realistic. It looks more like requiring several unlikely things to happen together. As for what today's share price implies: at about $171 and about $140.0 billion of market cap, the market is assigning a low multiple to a "mature cash cow + uncertain AI realization + capital-allocation discount" and is not implying a high-growth expectation of "five-year doubling/ten-year five-bagger." That is part of the safety, but it also means the fivefold story requires the market to first reclassify it as a growth stock. Baillie Gifford's standard answer to this question is to list the conditions needed for a ten-year five-bagger, judge whether they are realistic, and unpack the current price's implied expectations. Here is the breakdown.

  First translate "fivefold in ten years" into the conditions that must hold simultaneously, starting from about $171 per share and about $140.0 billion of market cap today, where fivefold means roughly $700.0 billion of market cap and about 17.5% annualized total return:

  1. Revenue engine: organic growth needs to reaccelerate from the current about 7%–8%, since FY27 guidance of about 10%–11% includes about 3pts of Informatica consolidation, and then remain in double digits for a long time. Reality: FY2026 revenue grew only about 10% according to the FY2026 Q4 results, while the CRM industry itself is growing at about 14.6% according to Grand View Research, and Salesforce has consistently grown slower than the industry. Revenue alone cannot support a ten-year fivefold outcome.

  2. Margin engine: FCF margin needs to move another step up from about 34.7% and actually accrue to shareholders. But it is already on a high base, with FY2026 GAAP operating margin of 20.1% according to the FY2026 Q4 results, leaving limited room for further expansion, while SBC of $3.509 billion in FY2026 continues to erode truly distributable cash.

  3. AI/data second-curve engine: Agentforce + Data 360 must scale from current combined ARR of about $2.9 billion, including about $1.1 billion of Informatica, to a level that materially rewrites the overall growth curve. Agentforce ARR rising 169% in a year to $800 million according to the FY2026 Q4 results is genuinely impressive, but it remains small relative to the $41.5 billion revenue base. For it to pull the whole company into double-digit growth, it needs to sustain very high growth for many years, possible but not easy.

  4. Valuation-multiple engine: the market must reprice Salesforce from a "mature cash cow" back into a "growth platform," with meaningful multiple expansion. But the report points out that the market is currently doing exactly the opposite, applying a lower multiple for Salesforce's maturation, uncertainty around AI realization, and capital-allocation discount.

  5. Capital-allocation engine: net repurchases need to keep offsetting SBC and truly grow per-share value, ideally with buybacks executed at cheap prices. But the FY2026 10-K shows that FY2026/FY2025 repurchase average prices of $254.21 / $260.12 were both far above the current roughly $171. Historical buybacks were not cheap.

  Are these conditions realistic? Each one is possible in isolation, but a ten-year five-bagger requires nearly all of them to land simultaneously in the optimistic range. For a mature leader whose growth has fallen to about 10% and is being supplemented by M&A, that is not realistic. The report's three-scenario valuation also supports this: even the optimistic scenario, involving reacceleration from the second half of FY27, Agentforce and Data 360 amplifying platform effects, and an 8.5% discount rate, implies intrinsic value of only about $260–$310 per share. Relative to the current roughly $171, that is about 1.5–1.8 times, far from fivefold. In other words, even the report's own optimistic case does not treat fivefold as a reasonable expectation.

  So what does today's share price imply? It implies low, conservative expectations:

  • Relative valuation is low. The report cites comparable data showing current CRM P/FCF of about 10.2x, EV/EBITDA of about 12.4x, and PB of about 2.8x, significantly below software leaders such as ServiceNow at 22.7x P/FCF, SAP, and Microsoft that still carry AI narratives. The market is not giving a growth premium; it is applying a maturation discount.
  • Yield perspective: based on the report's conservative owner-earnings estimate, the static yield is about 7%. Relative to the U.S. 10-year Treasury yield of about 4.56%–4.57%, the risk premium is not extreme. Relative to the S&P 500 forward earnings yield of about 4.24%, based on FactSet's early-May 2026 data cited by the report and an S&P 500 forward P/E of about 20.9x, it is only "slightly better, not obviously superior."
  • Implied market narrative: the current price broadly implies "mature, moderate growth, with returns mainly from cash flow and buybacks." It does not imply "renewed high-speed growth." The report places the current roughly $180 price, and even cheaper after updating to about $171 today, near the lower end of the fair-value range and clearly above the conservative range. That is exactly the implied expectation of "not expensive, but not cheap enough to ignore execution risk."

  My conclusion: a ten-year fivefold outcome requires five things to happen at once: revenue reacceleration, another step up in margins, scaling of the AI second curve, multiple rerating, and buybacks enhancing per-share value. That is not realistic for a mature leader, and even the report's optimistic case points only to about 1.5–1.8 times, not fivefold. Today's share price instead implies low expectations. That provides downside protection, making it less likely to collapse on a single quarterly miss, but it also means the prerequisite for "fivefold" is a narrative rerating in which the market first reclassifies it from a cash cow into a growth stock.

  Jun 10, 2026
• QWhy has the market not realized all this yet? Is it failing to understand, looking down on it, or unable to look far enough? What would become the "narrative inflection point"?3/10

  A

  Bottom line first: the market is not "failing to understand" Salesforce. It is "looking down on it + not looking far enough." It has reclassified Salesforce as a cash cow with maturing growth, AI realization that needs time, and imperfect capital allocation, and has lowered the multiple accordingly. That judgment is quite reasonable, not simply a market mistake. So strictly speaking, this may not be a "market misjudgment waiting to be corrected." It is more like "the market's pricing is broadly right, but it assigns conservative weight to long-term AI realization." The real "narrative inflection point" is whether Agentforce/Data 360's organic growth can prove that it truly pulls up companywide growth and margins. Once the data speaks, the market may reclassify Salesforce from "cash cow" back to "growth platform." Baillie Gifford's question asks whether the market does not understand, looks down on it, or cannot look far enough, and what would create the narrative inflection point. Here is the layer-by-layer breakdown.

  First exclude "does not understand." Salesforce is a highly understandable business. The FY2026 10-K shows subscription and support revenue at about 95% of total revenue, diversified customers with no single customer above 10% of revenue, and a clear business model. The report also scores business understandability at 4.5/5. Institutional coverage is extensive and financial disclosure is detailed. There is no information asymmetry where "the market does not understand and therefore has mispriced it." This is not a "does not understand" opportunity.

  The main reason is "looking down on it," a maturation discount. The market sees FY2026 revenue growth of only about 10% according to the FY2026 Q4 results, FY27 guidance of about 10%–11% growth that still includes about 3 percentage points of Informatica consolidation, implying organic growth of about 7%–8%, and management describing the cadence as "second-half reacceleration." All of this leads the market to conclude that Salesforce is "no longer a structurally high-growth platform." So it receives a low multiple: the report cites current CRM P/FCF of about 10.2x and EV/EBITDA of about 12.4x, significantly below ServiceNow at 22.7x P/FCF, SAP, and Microsoft. The objective backdrop confirms this sentiment: Salesforce's share price has fallen about 30% since the start of 2026, and the company even responded to the weak valuation and market skepticism about the growth story with its largest-ever accelerated repurchase of about $25.0 billion, announced in March 2026. The report describes this discount as "reasonable, but also partly reflected in the price." I agree. This is exactly "looking down on it," and doing so for reasons that make sense.

  The secondary reason is "not looking far enough," meaning the market assigns conservative weight to the AI second curve. The market is skeptical about the realization cadence of Agentforce: the report cites an external survey reported by Barron's showing Agentforce paid adoption and deployment moving slower than the most optimistic expectations. Objectively, Agentforce ARR rose 169% in one year to $800 million, and combined Agentforce and Data 360 ARR exceeded $2.9 billion according to the FY2026 Q4 results, but that is still small relative to $41.5 billion of revenue and includes about $1.1 billion of Informatica consolidation. The market chooses to "see it first, then give credit," discounting the ten-year value of the AI closed loop. That is "not looking far enough," but without realization evidence, conservatism is rational.

  Why say this may not be a "wrongly punished" stock? We must be honest: all three reasons for the lower multiple, maturation, uncertain AI realization, and capital-allocation discount, are supported by facts. Growth has indeed fallen to about 10% and is being supplemented by M&A; AI remains a minority contribution; and buyback average prices of $254.21/$260.12 according to the FY2026 10-K were indeed far above the current roughly $171. So this is closer to "pricing is broadly right." The report's Watch rather than Buy rating, and its emphasis that the margin of safety is insufficient, are based on this. The opportunity is not "the market is wrong and will correct." It is "if AI/data realizes in the future, the market's classification may shift from cash cow back to growth stock."

  What could become the narrative inflection point? I see four types of signals, any of which could trigger a rerating if verified by data, consistent with the report's core tracking and trigger indicators:

  1. The organic-growth inflection point: after excluding Informatica consolidation, organic growth in core subscription and support revenue returns sustainably to double digits, rather than being hidden by acquisitions. This is the hardest inflection point.
  2. AI pull-through is proven: rapid expansion of Agentforce/Data 360 truly translates into simultaneous improvement in companywide growth and margins, with cRPO/total RPO, currently $35.1 billion / $72.4 billion according to the FY2026 Q4 results, accelerating alongside it.
  3. Management's promise of "H2 FY27 reacceleration" is delivered: if the second-half reacceleration embedded in FY27 guidance appears on schedule, the narrative can shift from "mature" to "regrowing."
  4. Capital-allocation quality improves: net repurchases truly offset SBC and raise per-share value, and the company stops consistently buying back at clearly expensive prices. This would repair the capital-allocation discount.

  Conversely, the report also identifies reverse triggers that would turn the narrative decisively bearish and create permanent capital-loss risk: ex-consolidation core growth slipping to about 5% or lower, attrition worsening above 10%, Agentforce/Data 360 failing to drive total customer expansion, Informatica integration damaging margins, and SBC not being offset by net repurchases so that per-share value fails to rise or declines.

  My conclusion: the market's view of Salesforce is "looking down on it + not looking far enough," and both are based on real facts, not simple mispricing. Current pricing is broadly reasonable and the discount is evidence-based. The narrative inflection point is not "market awakening," but "the AI/data second curve proving through organic data that it can drive companywide growth and margins." Until then, Salesforce looks like what the report says: worth studying, worth lining up for, but not yet something that must be bought immediately.

  Jun 10, 2026