Long-Term Thesis

Long-Term Thesis — What Has To Be True By 2035

The 5-to-10-year question for Toast is not whether it is a good company — the share leadership, the profitability inflection, and the rising take rate are settled facts covered elsewhere. The question is narrower and harder: can a U.S.-restaurant-POS franchise compound recurring gross profit at ~20% for a decade, from ~$1.8B today toward management's stated $5B then $10B-plus, while holding a 40%-plus EBITDA margin and finally retiring shares — and is the switching-cost moat durable enough to underwrite that path against better-capitalized payment rivals? Management has put the destination in writing: scale the business to "$5 billion and $10 billion and beyond" [1], at a "long-term 40-plus percent adjusted EBITDA margin" [2], "sustaining high growth for the next five to ten years" [3]. The thesis is a bet on whether that target survives contact with maturation and competition.

The destination management is underwriting

Read the long-term story as one equation taken to scale. FY2025 recurring (subscription plus fintech) gross profit was ~$1.82B [7]; the path to $5B implies ~22% compounding for ~5 years, and to $10B-plus another ~5 years in the high teens — exactly the "five-to-ten-year" window management frames [8][9]. Because FY2025 adjusted EBITDA of $633M already runs ~35% of recurring gross profit [10], the 40%-plus margin target is a modest expansion, not a heroic one — which is why the growth rate, not the margin, is the variable that sets the outcome.

Source: derived from reported FY2025 recurring gross profit [11] compounded at illustrative rates anchored to management's stated $5B/$10B path [12] — bull ~24%, base ~21% (tapering), bear ~16% decelerating to ~10%.

The three lines are not forecasts; they are the spread of outcomes the durable thesis has to navigate. The bull path reaches ~$12.6B of recurring gross profit by 2034 (and, at 42%, ~$5.3B of EBITDA); the base path lands near management's $10B with ~$4B of EBITDA; the bear path — deceleration into the low teens as the U.S. core matures faster than new verticals offset — stalls near $5B, never reaching the second milestone. The gap between the blue and red lines, roughly 2.5x of terminal earnings power, is the entire investment debate, and it is decided by the four pillars below.

Source: base-case recurring gross profit derived from reported FY2025 figures [13]; 40%-plus margin target per management [14].

The four pillars — what has to be true, and the evidence on each

Everything in the durable frame reduces to four load-bearing assumptions. Each has a mechanism, an evidence base, and a status: which are proven, which are in progress, and which are unproven and therefore the swing factors.

Sources: retention and share [15][16][17]; take rate [18]; new-TAM economics and Drive-Thru/grocery [19]; cash and capital return [20][21][22].

Pillar 1 — the moat is proven where the evidence was generated, and unbuilt where the capital is going

The hard evidence is a net revenue retention rate that held above 110% every single year since 2015 — 119%, 122%, 114%, 110%, 114% for base years 2015–2019, the last carrying a 2020 comparison through the COVID restaurant shock [23]. That franchise still reads 109% SaaS net retention in 2025 [24], and Toast now powers ~20% of U.S. SMB and mid-market restaurants, roughly doubled in three years [25]. Retaining above 110% while taking share is the combination a commodity reseller cannot produce.

But the moat is mechanism-specific and uneven. Toast's own 10-K concedes SMB customers "are more readily able to change their payment processors than larger organizations" [26], that SaaS terms run only 12–36 months and customers "are not obligated to, and may not, renew" [27], and that rivals enjoy "substantially greater financial, technical, sales, and marketing, and other resources" [28]. The decade of retention proof was generated in the U.S. independent core; in enterprise, international, and retail the switching-cost clock resets to zero. The moat is durable enough to underwrite for the core and an open question for the adjacencies.

Pillar 2 — the rising take rate is the clearest pricing-power tell, and the most contestable variable

Monetization crossing 1% of payment volume for the first time — 103 bps, up 5 bps year-over-year [29] — is leverage stacked on ~20% volume growth, and it is the single fact a truly competed-away spread should not produce. But because ~82% of revenue rides on that spread, the same line is the thesis's soft underbelly: if a better-resourced rival forces basis-point give-back, it hits the bulk of the P and L, not a fringe. The long-term bull layer here is Toast IQ and outcome-priced AI agents — a way to raise the take rate on software value rather than the contestable processing spread — but that is optionality, not yet a proven moat extension.

Pillar 3 — the new TAMs are the swing factor, and they are a promise being kept quarter to quarter, not a settled fact

The reinvestment runway is genuinely large, and it is where most growth capital now flows. Below.

Pillar 4 — the capital-allocation turn is real but unfinished

FY2025 generated $608M of free cash flow, above net income of $342M [30][31], on a net-cash balance sheet, and stock-based compensation has fallen to 11% of recurring gross profit, roughly half its level two years earlier [32]. The board added $500M to the buyback in February 2026 on top of the original $250M [33]. The catch the durable frame must monitor: diluted shares still rose from 591M to 607M in FY2025 [34]. The buyback is slowing dilution; it has not yet reversed it. Genuine per-share-count shrinkage is the confirming signal that the compounding accrues to owners, not just to the enterprise.

The reinvestment runway — long, but each new vertical resets the clock

Toast's own IPO math sized the U.S. restaurant-technology opportunity at roughly $55B of annual spend by 2024 [35], against which it served only ~6% of ~860,000 U.S. locations and ~3% of a then-$15B serviceable market at IPO [36]. The penetration curve since is the runway made concrete: from ~48,000 locations and $38B of payment volume at IPO [37] to ~164,000 locations and $195B of GPV in FY2025 [38], with ARR compounding to ~$2.05B [39].

Sources: IPO-era locations and GPV [40]; FY2025 locations, GPV [41] and ARR [42]; intermediate years company filings, as reported.

The new TAMs extend the runway well beyond the core. Management says that in each new TAM, ARR is growing faster and at higher SaaS ARPU than the core did at the same age; Toast for Drive-Thru opens ~140,000 additional locations, and grocery alone is over 20,000 independent U.S. operators representing ~$250B of sales [43]. The right skeptical question is not whether the TAM exists — it plainly does — but whether the unit economics in drive-thru, London, and grocery ever match the beloved U.S.-independent core where the moat is deepest. This is the assumption with the widest error bars and the largest payoff, and it is the one a long-term holder must re-underwrite every quarter.

Does management earn the benefit of the doubt? Yes — with a 2028 governance catalyst

A 5-to-10-year hold is a bet on the people as much as the model, and the credibility record is unusually clean. Management promised a quarterly adjusted-EBITDA profit "by the end of 2023" and delivered it two quarters early; the strategy never pivoted — only the execution proved out. The one self-inflicted misstep (a mandatory $0.99 online-order fee in July 2023) was reversed in writing within a week — "We made the wrong decision" [44] — and the 2022 material weakness in internal controls was formally remediated by year-end 2023 [45]. The company is founder-led continuity, not a turnaround by outsiders: co-founder Aman Narang has been CEO since January 2024 [46].

The one structural governance knock is also the one with a published expiry. Insiders control 40.6% of the vote on just 2.0% of the Class A economics [47], but the super-voting Class B sunsets automatically on September 24, 2028 [48]. For a decade-long holder, that is a positive embedded in the horizon: the central control discount mechanically falls away mid-hold, and one-share-one-vote arrives.

What proves the thesis working — and what proves it breaking

The discipline that separates long-term thesis evidence from quarterly noise is a fixed set of leading indicators with thresholds set in advance. The order below is the order in which they would actually warn you — retention is the moat's vital sign and fires first; the financials confirm later.

Sources: thresholds are this analyst's framing; underlying facts — retention [49], take rate [50], share count [51], and the Toast Capital 15% loan-buyback obligation [52], the auditor's sole Critical Audit Matter [53].

Bottom line

Toast clears the bar for a long-term compounding thesis on the evidence that is hardest to fake: a decade of above-110% retention through the worst restaurant shock on record [55], a take rate rising to 103 bps while it gains share [56], and a model that has crossed from cash furnace to self-funding compounder throwing off $608M of free cash flow [57]. The destination management has committed to — $5B then $10B-plus of recurring gross profit at a 40%-plus margin [58][59] — is arithmetically reachable from a ~$1.8B base at the growth rates the franchise has already demonstrated [60].

What keeps this a high-conviction-with-confirmation thesis rather than a certainty is that two of the four pillars are not yet proven: the new TAMs that carry most of the incremental capital have core-like economics only on management's word, and the de-dilution that would make the compounding accrue to owners has slowed but not reversed [61]. The durable thesis is intact as long as net retention stays above ~108% and the take rate keeps climbing; it is breaking the moment those two roll together. Underwrite the core with confidence, the adjacencies with skepticism, and watch retention before you watch anything else.