SaaS Gross Retention Slips to 88% as AI Substitution Bites

What happened

Benchmarkit's latest SaaS performance data shows median gross revenue retention (GRR) sliding to roughly 88%, down from about 90% in 2022. Two percentage points may sound trivial, but Benchmarkit itself flags the three-year drift as a potential "canary in the coal mine", a small, early signal of a larger structural shift rather than ordinary noise.

The drivers named in the research are two-fold. The first is a continued post-boom rationalization: buyers who over-accumulated software during the 2020–2022 expansion have spent the years since consolidating their stacks and cutting tools they do not use. The second, and newer, pressure is AI substitution, low-switching-cost software being displaced by AI-native alternatives, or absorbed into broader platforms that now bundle a capability a standalone tool used to sell. Net revenue retention has drifted the same direction, with median NRR for private B2B SaaS reported to have fallen from roughly 105% in 2021 to about 101% in 2024.

The distribution underneath those medians is where the warning sharpens. NRR splits hard by segment: roughly 118% for enterprise accounts (ACV above $100K), about 108% for mid-market, and only around 97% for SMB (under $25K ACV). AI-native SaaS companies, meanwhile, post strikingly weak retention as a group, median NRR near 48% and GRR around 40%, consistent with fast adoption followed by rapid churn once novelty fades. ChartMogul data cited alongside the benchmarks shows the switching-cost gradient explicitly: plans under $50/month retained at about 32% NRR, versus roughly 85% for plans above $250/month.

Why it matters for practitioners

For a bootstrapped SaaS founder, GRR is the metric that determines whether your business compounds or leaks. Net revenue retention can be flattered by a handful of big expansions; gross retention cannot be gamed, it is the share of revenue you keep before any upsell, and it is the truest read on whether customers actually need your product. A slipping industry median is a signal to pressure-test your own moat before the trend reaches you.

1. Switching cost is the whole game now. The ChartMogul gradient is the single most actionable finding here: cheap, shallow plans retain at ~32% while deeper, higher-priced ones retain near 85%. AI substitution does not threaten all software equally, it threatens software that is easy to leave. If a customer can replicate your core value with a prompt and a spreadsheet, low price will not save you; depth of integration, data gravity, and workflow entanglement will. Founders should be auditing, honestly, how many days of pain it would cost a customer to rip them out.

2. Retention is a product-led metric before it is a customer-success one. The product-led companies that defend retention best do it inside the product, not through save calls at renewal. When users reach genuine value early and weave a tool into their daily workflow, gross churn falls as a byproduct. That is the whole logic of product-led growth: the product itself does the onboarding, the expansion, and much of the retention work. In an AI-substitution environment, a strong activation experience that gets users to a habit quickly is a defensive moat, not just a growth tactic.

3. Expansion cannot paper over a leaky base. Expansion revenue now represents roughly 40% of new ARR across companies, rising to 58–67% above $50M ARR, which is exactly why a soft GRR is so dangerous. If you are refilling a leaking bucket with upsells, you are running to stand still, and every point of gross churn raises the bar on how much expansion you need just to hold flat. Fixing the leak is almost always cheaper than out-expanding it.

Key details

• Median GRR: ~88% in the latest data, down from ~90% in 2022 (Benchmarkit)
• Framing: Benchmarkit calls the three-year decline a potential "canary in the coal mine"
• Median NRR: fell from ~105% (2021) to ~101% (2024) for private B2B SaaS
• NRR by segment: ~118% enterprise (ACV >$100K), ~108% mid-market, ~97% SMB (<$25K)
• AI-native SaaS: median NRR ~48%, GRR ~40%, fast adoption, fast churn
• Switching-cost gradient (ChartMogul): ~32% NRR on sub-$50/month plans vs. ~85% on $250+/month plans
• Expansion mix: ~40% of new ARR overall, 58–67% above $50M ARR
• Named drivers: post-boom stack consolidation and AI-native substitution of low-switching-cost tools

Market implications

The most important reframing in this data is that retention has become a proxy for defensibility against AI. For a decade, gross churn was mostly a story about onboarding, support, and pricing fit. In 2026 it is also a story about whether your category can be swallowed by an AI-native alternative or a platform bundle. The tools posting 40% GRR are not badly run, many are simply positioned in a place where the switching cost is near zero, and that is now a category-selection problem as much as an execution problem.

For bootstrapped founders, the strategic response is to compete where depth beats novelty. AI-native entrants win the demo and lose the year; their strength is a fast first impression and their weakness is durability. A product-led business that invests in data accumulation, integrations, and genuine workflow entrenchment builds exactly the retention that AI-native tools struggle to match. The gap between a 32% and an 85% retention curve is not mostly about price, it is about how hard you are to leave.

That defense starts at the top of the funnel. Retention is largely determined in the first session, which is why activation and free-tier design deserve as much rigor as any renewal motion. A well-constructed free tier that gets users to real value fast, and to a habit faster, does more for gross retention than any downstream save play, because it front-loads the moment where a user decides the product is worth keeping. In a market where the industry median is drifting down and AI substitution is the named culprit, the founders who treat activation and switching cost as first-class priorities will be the ones whose retention curves hold while everyone else's slip.

Related resources

• Product-Led Growth (Glossary), Why expansion and retention sit at the center of PLG economics
• PLG Companies Analysis, How product-led businesses defend gross retention against AI-native entrants
• How to Launch a Free Tier, Activation and free-tier design as a retention lever, not just an acquisition one