Long-Term Thesis
Long-Term Thesis — monday.com Ltd (MNDY)
The underwriting question: what has to be true over the next 5–10 years for monday.com to be a superior investment, and what evidence proves the thesis is working or breaking?
The one-paragraph answer. This is a founder-led, net-cash, ~89%-gross-margin platform that has already done the hard part most software IPOs never reach — it grew revenue to $1.23 billion in FY2025 (+27%) while turning ~26% of it into free cash flow [1] [2], and whose switching-cost moat strengthened through the worst SaaS downturn in two decades. The long thesis is not "buy a hyper-grower" — that era is over, with growth decelerating from 91% to a guided 19–20% [3]. It is "buy a durable, expanding installed base that compounds per-share value through reinvestment and a float-shrinking buyback, while a single binary — whether AI expands the model or reprices it — gets resolved over the next 8–10 prints." Underwrite the moat and the cash; do not underwrite a re-acceleration you cannot yet see. The five conditions below are the spine of that view, each with the multi-year evidence that confirms or kills it.
FY2025 Revenue
FY2025 Free Cash Flow
Net Cash (no debt)
Net Dollar Retention
Source: FY2025 Annual Report (Form 20-F) — revenue [4], adjusted free cash flow [5], liquidity [6], net dollar retention [7].
The thesis spine — five conditions that must hold
A long-term thesis is only useful if it can be falsified. These are the five things that must be true for a multi-year compounding outcome, ranked by how much they decide the result and how much remains genuinely unresolved.
Source: synthesis of cited disclosures throughout this page; status and residual-risk calls are analyst judgment derived from the cited evidence.
The whole thesis pivots on Condition 3. Conditions 1, 2, 4 and 5 are already largely demonstrated in the multi-year record; the AI question is the one variable that can independently turn this into either a re-rating compounder or a value trap — which is why the watch signals at the end weight it most heavily.
Condition 1 — The category keeps expanding, and consolidates onto platforms
At IPO, monday sized its addressable market — summing project/portfolio management, collaborative applications, sales-force productivity, software change management and marketing campaign management — at $56.1 billion in 2020, growing to $87.6 billion by 2024, a 12% CAGR [8]. The composition matters more than the headline: the largest slice was collaborative applications ($21.7B), then marketing campaign management ($14.3B) and sales-force productivity ($11.5B) [9] — i.e. the TAM is defined by the adjacent products monday is now building (CRM, service, dev), not just its original work-management wedge. That is the structural reason the reinvestment runway is real: the company can grow into categories it already addresses.
The more durable point is the demand re-anchoring underneath. monday frames the next decade around three industry trends it expects to govern enterprise software: AI is everywhere but adoption is the bottleneck (the "adoption advantage"); unified platforms beat point tools because AI needs grounded, context-rich company data; and organizations are raising software budgets specifically to fund AI transformation, with Gartner-cited AI spending reaching $4.7 trillion by 2029 [10]. The structural tailwind is intact; the open question (Condition 3) is who captures it.
International mix is an underrated, durable growth lever inside this TAM: the United States was 50% of revenue in FY2025 ($619M of $1,232M), with EMEA, the UK and rest-of-world the other half — and U.S. revenue has compounded from $77.9M in 2020, leaving the non-U.S. base still early in its land motion.
Source: FY2025, FY2022 and FY2021 Form 20-F geographic revenue notes, as reported [11].
Condition 2 — The switching-cost moat holds and widens into a platform
This is the condition with the cleanest multi-year proof, and it is what separates monday from a generic decelerating SaaS name. The test of a real switching-cost moat is what retention does when the cycle turns against expansion. Through the 2022–2025 software bust, the headline expansion signal faded hard — net dollar retention in the over-$50k cohort fell from a 137% peak in 2022 [12] to 116% — but the base itself never left. Blended NDR never fell below 100%, and underneath it enterprise gross retention rose to a company-record 91% [13].
Sources: FY2021 20-F (2019–2021) [14]; FY2024 20-F (2022 over-$50k peak) [15]; FY2025 20-F (2023–2025) [16].
Read the two lines separately: the expansion premium (height above 100%) compressed as the cheap-capital boom ended and 2024–25 list-price increases annualized out — that is the real bear case on growth. But the retention signal that actually measures the moat strengthened through the same downturn. A moat is what you keep when the wind stops at your back; monday kept its customers. The relative control proves it is company-specific, not a category feature: its closest pure-play substitute Asana runs dollar-based net retention of just 96% — a shrinking installed base in the identical model and macro [17]. A 14-to-20-point retention spread is the difference between an industry feature and an edge.
The moat is widening from one product into a platform, account by account — the deepest cohorts are growing fastest, exactly where lock-in compounds.
Source: FY2025 Annual Report (Form 20-F), Consistent Growth of Enterprise Customers [18].
The over-$50k cohort now drives 41% of all ARR (up from 36%), 29% of enterprise customers use two or more products versus only 6% of sub-$50k customers, and monday CRM crossed $100 million ARR in 2025 — the proof that the second and third products become real businesses, not features [19] [20]. The customer base is also unusually diversified — over 250,000 accounts, none above 1% of revenue [21] — so there is no whale risk to the moat.
The cap on the moat — why it is narrow, not wide. monday concedes it still derives the majority of its revenue from the original monday work management product [22], and the substitution ceiling is permanent: it "directly compete[s] with several large technology companies … including Google and Microsoft" and expects that competition to increase [23]. Microsoft can bundle good-enough work management (Planner, Project, Teams, Copilot) into licenses a CIO already owns, behind research-and-development spend of $32.5 billion — roughly 26x monday's entire revenue [24]. The switching cost protects the installed relationship; it does nothing to win the next deal against a free bundle. That is the structural cap on pricing power and the reason this is a narrow — if widening — moat.
Condition 3 — AI: expansion vector or repricing event? (the unresolved swing variable)
This is the variable that decides the thesis, and it is honestly unresolved in the numbers. Every lever that drove the model is bending at once, and the new vector barely contributes yet. Management is migrating from predictable per-seat billing to seats-plus-consumption "credits," so that "revenue expands naturally without requiring additional seats purchases" as AI agents do more of the work [25]. At the same time, gross margin is guided down from ~90% toward the mid-80s on AI compute cost [26], and the self-serve top of funnel is being disrupted — the company discloses it has "experienced declines in web traffic from google searches, primarily due to AI-generated search updates" [27]. Yet AI drove only ~3% of net-new ARR in Q1 FY2026 [28].
Sources: Q1 FY2026 Earnings Call — pricing [29], gross margin / AI compute [30], AI net-new ARR [31]; FY2025 20-F web-traffic risk factor [32].
The bull reading: a platform that already holds a company's work data is the natural home for agents that act on it — consumption pricing layers a new value-capture vector on a sticky base, and the data lock-in is "a data advantage that no point solution can replicate" [33]. The bear reading: AI that makes building bespoke internal tools trivial erodes the value of a no-code platform; the consumption switch swaps visible subscription billings for opaque usage; and the same compute permanently clips the margin while the giants distribute AI free across far larger captive bases. Both readings are coherent today. The 5–10 year outcome hinges on which is right, and the only honest position is to make AI's share of net-new ARR the central thing you watch.
Condition 4 — ~89% gross margin must convert into real owner FCF
The franchise economics are genuinely strong, but the long-term thesis must underwrite the quality of the cash, not just the headline. monday converts revenue to cash at an excellent rate — adjusted FCF of $322.7 million on $1,232 million of revenue (~26% margin), on capex below 2% of revenue [34]. The multi-year inflection is real: FCF margin went from 1.6% (FY2022) to the high-20s, and cumulative free cash flow across FY2023–FY2025 is ~$811M — running ahead of the ~$1B-over-2023-2026 pace management promised at its 2023 Investor Day.
Source: derived from reported financials, FY2022–FY2025 Form 20-F cash-flow and selected-data disclosures [35].
But the caveat is structural and durable: the GAAP operating line was a $(1.7)M loss in FY2025, and stock-based compensation of $177.0M — roughly 14% of revenue — almost exactly equals all of non-GAAP operating income ($175.3M) [36]. Net the SBC out and owner cash is closer to ~$130–145M, a pedestrian ~4% yield rather than the headline ~26% margin — and FY2025's reported GAAP net income of $118.7M was 100% non-operating (interest on the IPO cash plus a one-time tax benefit), so it "will not recur at that level." The durable question for Condition 4 is whether the GAAP operating line turns durably positive — Q1 FY2026 flipped to a small GAAP operating profit — and whether SBC trends down as a share of revenue as the company scales. Both are observable, and both are on the watch list.
A Rule-of-40 lens still flatters the business — ~27% growth plus ~26% FCF margin is ~53 — and the balance sheet gives years of runway to get the conversion right: deferred revenue of $411.6M and remaining performance obligations of $838.9M (81% recognizable within 12 months) [37] [38] provide unusually high forward visibility.
Condition 5 — Founder-led capital allocation compounds per-share value
The fifth pillar is the one most within management's control, and the record so far is shareholder-friendly. The balance sheet holds $1,503.1M cash plus $162.3M of marketable securities with zero debt — net cash of ~$1.67B, roughly half the market capitalization [39]. In September 2025 the board authorized the first capital return in company history — an $870M repurchase program — and management is deploying it aggressively into the de-rating: $135.0M in Q4 2025 [40], then an accelerated Q1 2026 program that left only ~$182M remaining [41] — roughly $688M repurchased in two quarters against a ~51M-share float. On a tiny float at a trough multiple, that is a powerful per-share lever; the honest qualifier is that with SBC near 14% of revenue, a meaningful part of the buyback neutralizes dilution rather than shrinking the share count outright.
Governance reinforces, rather than undermines, the alignment. monday is run by co-founders Roy Mann and Eran Zinman, who together own ~13.0% (Mann 9.6%, Zinman 3.4%), with all officers and directors at 13.9% [42]. Crucially, the founders chose not to adopt a super-voting dual-class structure; control instead runs through a single non-tradable "founder share" that carries no economic or ordinary voting rights and only vetoes change-of-control and a few charter items, self-destructing if the founder leaves or dilutes below a threshold [43]. Co-CEO pay is modest and ~90% equity, and in the worst loss year (FY2022) the founders voluntarily waived their cash bonuses — a small but telling signal of owner-operator mindset over a multi-year horizon.
Track record — what management has actually delivered
Five-to-ten-year underwriting depends on whether management does what it says. The multi-year record is mostly promises kept, with one clear recent ding that matters for how you weight forward guidance.
Sources: Investor Day 2023 — cumulative FCF target [44] and Rule of 60+ [45]; Investor Day 2025 — FY2027 $1.8B revenue target [46] and 20-25% long-term margin target [47]; Q4 FY2025 call withdrawal [48].
The withdrawn FY2027 target is the single blemish, and it cuts both ways. It set a $1.8B FY2027 revenue base case and a 20–25% long-term operating-margin ambition in September 2025 [49] [50], then pulled both barely five months later [51] — a genuine hit to forward credibility and a reason not to underwrite any specific 2027–2030 number. The charitable read is that management chose honesty over defending a stale figure once the AI transition made multi-year visibility unreliable. Either way, the lesson for the long-term holder is to underwrite the moat and cash, which are demonstrated, rather than management's growth targets, which they have just told you they cannot see.
The competitive frame over 5–10 years
Among pure-play work-management peers, monday pairs the fastest growth with the best profitability — a combination none of Asana, Atlassian or Freshworks matches today. That relative position is the durable structural fact; the durable structural risk is the platform giants (Microsoft, Salesforce, ServiceNow), who are larger, distribute AI across captive bases, and can bundle.
Sources: company FY2025/FY2026 annual reports — Asana 10-K [52], Atlassian 10-K [53], ServiceNow 10-K [54], Microsoft 10-K [55]; monday FY2025 20-F [56].
The enduring asymmetry: monday wins on retention and profitable growth among pure-plays, but lags badly on enterprise depth — it has only ~87 customers above $500k ARR versus ServiceNow's hundreds above $5M ACV, and ~250,000 customers versus Atlassian's 300,000-plus. That gap is simultaneously the bear's ceiling (it may never crack the largest enterprises) and the bull's runway (the up-market motion is young and compounding 45–74% at the top cohorts).
What you are paying — the valuation frame
The reason the setup is interesting is that the multiple already prices a stall. After a ~78% de-rating from the 2025 peak, monday trades around 1.1–1.3x EV/sales and a mid-single-digit EV/FCF — multiples normally reserved for no-growth or eroding businesses, on a franchise still growing ~20% with ~26% FCF margins and net cash worth half the market value. The asymmetry: the bull re-rate (toward ~2.7x forward EV/sales, still a discount to a 20%-grower-with-26%-FCF peer) implies ~$115; the structural bear (capitalize stressed, SBC-adjusted operating owner-FCF at a no-growth multiple plus net cash) implies ~$48. The long-term holder is paid to wait via the buyback and the net cash floor while Condition 3 resolves.
The multi-year scorecard — signals that prove or break the thesis
These are the durable signals to track over the next 8–12 quarters, ranked by how much they decide the 5–10 year outcome. They are deliberately separated from quarterly noise: each maps to one of the five conditions.
Source: framework derived from monday.com's disclosed key business metrics, pricing transition and risk factors cited throughout this page [57] [58].
Verdict — the durable frame
monday.com clears four of its five long-term conditions on the evidence available today: the category is structurally growing and consolidating onto platforms; the switching-cost moat is real, company-specific and widening account by account; the cash conversion is strong even after the SBC haircut; and a founder-led, net-cash balance sheet is being weaponized into a per-share-accretive buyback. The single unresolved condition — whether AI expands the seat model through consumption monetization or reprices and substitutes it — is also the one that most decides the 5–10 year return, and it will not be settled by argument, only by two-to-three years of prints showing AI's share of net-new ARR and the gross-margin floor.
The top long-term driver is the compounding, switching-cost-protected installed base monetized through multi-product attach and (if it works) AI consumption. The top failure mode is AI dissolving the no-code value proposition and the seat economics together while Microsoft bundles the category away — turning a cheap compounder into a value trap. The durable, thesis-breaking line in the sand is enterprise gross retention rolling below ~90% or blended NDR breaking under ~105%; the near-term confirmation marker is AI's share of net-new ARR climbing past ~3% toward double digits. Underwrite the moat and the cash, demand the AI evidence before paying up for growth, and let the buyback and net-cash floor carry you while the binary resolves.
Bottom line. A durable, founder-led, net-cash compounder whose moat strengthened through a downturn, priced for a stall its own renewal data contradicts — with one genuinely binary AI question that decides whether the next decade is a re-rating or a slow erosion. The evidence that proves it is observable and close: watch enterprise gross retention and AI's share of net-new ARR above all else.