The Moat

The Moat

Gartner's competitive position holds two distinct kinds of protection, and they are moving in opposite directions. The first — clients renewing year after year — is intact: account retention has sat in a narrow 83–89% band for five straight years. The second — the ability to raise spend on those retained accounts — has weakened, with wallet retention sliding for four consecutive years and falling below 100% in 2025. Measured against its only pure peer, the moat is widening; the pressure is on the whole category, and Gartner is losing less of it.

What the moat protects

The asset is a proprietary-data flywheel, not a library. Gartner is in steady contact with over 13,000 distinct client enterprises, runs more than 510,000 direct client interactions a year, and employs more than 2,400 business and technology experts plus 920 consultants across 30-plus countries [1]. Those interactions are the raw material: what 80,000-plus executives are actually deciding feeds the research, and the research is what the next executive pays to read. Scale compounds the loop — more clients generate more signal, which makes the insight more valuable, which attracts more clients.

Sources: FY2025 Annual Report, Item 1 Business [2] and Insights segment metrics [3]; Forrester contract value from its FY2025 10-K [4]. Multiple is Gartner CV $5,155M ÷ Forrester CV $292M.

Gartner is candid that none of this is a legal barrier. Its own 10-K states that "limited barriers to entry exist" in its markets, that it competes with free sources of information available through the internet, and that it anticipates more competition as AI services are adopted [5]. The defense is not a patent; it is switching cost and authority — the Magic Quadrant that vendors organize roadmaps around, and the analyst on the other end of an inquiry call. Those show up in the retention numbers, which is where the moat is measured.

Two retention numbers

Gartner reports two retention rates, and the gap between them is the diagnostic. Client retention counts logos — the share of enterprises that were clients a year ago and still are. Wallet retention counts dollars — the contract value retained from those clients, including whether they spent more or less [6]. When wallet retention runs above client retention, retained clients are spending more; when it falls below 100%, the surviving base is spending less in aggregate before any new business is added.

Source: FY2022 10-K (2021–22), FY2024 10-K (2023–24) and FY2025 10-K (2025) Insights/Research segment metrics [7] [8] [9].

The two lines tell different stories. Client retention (the grey lines) has barely moved — GTS between 83% and 86%, GBS between 86% and 89% — and in 2025 GTS retention actually ticked up, to 85% from 84% [10]. Enterprises are not leaving. Wallet retention (the purple lines) is a steady glide down: GTS from 106% in 2021 to 96% in 2025, GBS from 115% to 99%. The moat is doing its first job — holding the account — while its second job, compounding the spend inside that account, has faded. Management attributes the 2025 drop to "lower levels of spending by existing clients," with a double-digit decline in the US public sector [11].

One quieter figure sits alongside this: the count of distinct client enterprises has fallen from more than 15,000 in 2021 and 2022 to over 13,000 in 2025 [12] [13]. Interactions per year rose over the same span, to a record 510,000, so engagement deepened even as the logo base narrowed — the smaller accounts thinned out while the core relationships intensified.

The peer test

The sharpest test of whether this is Gartner's problem or the category's is Forrester Research — the one indexed peer that runs the same syndicated-research subscription model (research is roughly three-quarters of its revenue). Forrester's retention levels sit materially below Gartner's on every line.

Sources: Gartner FY2025 10-K — total revenue [14], contract value and retention [15], client enterprises [16]; Forrester FY2025 10-K — key metrics [17] and total revenue [18].

Forrester's contract value fell 6% to $292 million, its client count dropped 7% to 1,797, and its wallet retention was 87% against Gartner's 96% [19]. Its client retention of 77% — even after a four-point improvement — sits eight points below Gartner's. Two readings follow. The category-wide one: the whole syndicated-research model is under wallet-retention pressure, so Gartner's sub-100% reading is not idiosyncratic. The relative one: at roughly 17 times Forrester's contract value, with retention nearly ten points higher and contract value still growing where Forrester's is shrinking, Gartner is the clear winner inside its niche, and its lead is widening as the weaker competitor contracts. The moat, defined as relative position, is stronger than it looks in isolation — but it sits inside a category whose ceiling is being tested.

Cyclical or structural

The evidence genuinely cuts both ways, and separating the cyclical from the structural is the harder task.

The cyclical case is specific and recent. The single largest drag was the US federal government: management flagged a double-digit public-sector decline, and by Q1 2026 disclosed that federal contract value was only about $114 million of a $5.3 billion base, with the DOGE-driven cuts having hit from March 2025 and beginning to lap in the second quarter [20]. Strip out federal and Q1 2026 contract value grew 3.5%, not 1%, and GTS wallet retention was 99% rather than 97% [21]. Engagement — the leading indicator Gartner ties to renewal — rose across every month of the quarter, up more than 170 basis points year-on-year [22]. New business slowed in March on macro caution rather than collapsing. On this read, the stall is IT budgets tightening and one government retrenching, both of which pass.

The structural case is that the wallet-retention glide began well before DOGE. The line fell from 106% to 101% between 2021 and 2023 — three years before any federal cut — and the logo count started shrinking over the same period. Forrester's parallel decline points to category maturity rather than a Gartner-specific stumble, and Gartner's own filing concedes that LLMs providing substantive content "could reduce the need to enter our websites," while clients loading proprietary Gartner content into large language models "could reduce the value of our offerings" [23]. If packaged expertise is what a chatbot increasingly approximates, the erosion of pricing power is the leading edge, not a passing dip.

AI on both sides of the ledger

AI is simultaneously the clearest threat to the model and its most-requested subject. Gartner sits between the CIOs buying AI, the vendors selling it, and its own independent coverage of it, and management calls AI the most requested topic across every role it serves — the argument being that an authoritative, vendor-neutral guide is more valuable, not less, when executives face information overload [24]. It launched AskGartner, an AI tool over its research, in August 2025 [25].

The tell is that both syndicated-research players are now embedding AI to defend renewals. Forrester attributes its four-point client-retention improvement partly to the launch of its own AI Access product in the third quarter of 2025 [26]. That both are investing to defend renewals signals each treats AI as a genuine threat to the model. AI is plausibly accretive to Gartner's authority in the near term and corrosive to its pricing power over a longer one, and which force dominates is not yet visible in the numbers.

Reading the stall

The weight of the evidence favors the cyclical read today. The account-level moat — the harder thing to rebuild — is fully intact: client retention is stable to rising, engagement is climbing, and the dollar softness maps cleanly onto federal cuts and a macro pause that management is already lapping. Ex-federal, the engine is still growing mid-single digits. The strongest fact against that read is that wallet retention has fallen for four straight years, starting before the federal shock and alongside a shrinking logo count and an identical decline at Forrester — a pattern more consistent with a maturing, AI-pressured category than a clean cyclical trough.

Two thresholds would settle it. Wallet retention re-crossing 100% as the federal comparison eases would confirm the pricing-power leg was cyclically depressed, not structurally impaired. Client retention breaking below its five-year floor of roughly 83% would mean the durable part of the moat — accounts, not just dollars — had started to give, which is the outcome the bull case cannot absorb. Client retention, more than the wallet number, is the line that would signal structural damage.