By Nick French · Founder, StackSwap · 10yrs B2B SaaS GTM (BDR → AE → Head of Revenue) · Methodology →
Affiliate link · StackSwap earns a commission if you sign up for HubSpot via this page (no extra cost to you). We only partner with tools we'd recommend anyway. · Editorial standards →

Operator analysis · HubSpot outcome-based Breeze pricing · pay-per-outcome AI · 2026

HubSpot Just Made You Pay Only When Its AI Works — Here's When That's Actually a Better Deal

At its Spring 2026 Spotlight (announced to press April 2, live April 14, 2026), HubSpot did the thing the whole category has been circling: it moved two Breeze AI agents off usage pricing onto outcome-based pricing. The Breeze Customer Agent went from $1.00 per conversation to $0.50 per resolved conversation. The Breeze Prospecting Agent moved from a recurring monthly charge on enrolled contacts to $1.00 per lead it recommends for outreach. HubSpot's pitch, from Chief Customer Officer Jon Dick: "Outcome-based pricing removes that risk. You pay when it works, full stop."

I run a HubSpot-anchored revenue stack as a daily driver and have shipped the Breeze Prospecting Agent and Breeze Data Agent in production — and I run Intercom 2 with Fin every day at my day job. So I'm pricing this against tools I actually operate, not a press release. Here is the only buyer question this change really moves: when does paying per outcome beat paying per seat — and how do you stop the vendor's definition of "resolved" from quietly inflating your bill?

StackSwap is a HubSpot affiliate, which is why this page exists. The analysis below is the same one I'd give a friend deciding whether to turn these agents on — including the part where the pricing model is a great reason to trial them and a bad reason to buy the platform.

Want to try HubSpot?

Already on HubSpot? Outcome pricing makes the Breeze agents cheap to actually test — so test them on a real queue.

Both agents ship with a free trial window. Wire the Customer Agent to a live support queue and the Prospecting Agent to a real list, run them for a few weeks, then pull the resolved-conversation log and the leads it surfaced and grade them against your own standard. The pricing change lowers the cost of finding out — your job is to confirm the resolutions are real before you scale the spend.

Explore HubSpot →Affiliate link — StackSwap earns a commission if you sign up for HubSpot. We only partner with tools we'd recommend anyway.

What HubSpot actually changed

Two agents, two new meters. The Breeze Customer Agent — the support deflection agent that answers tickets and chats — dropped from $1.00 per conversation to $0.50 per resolved conversation. The Breeze Prospecting Agent — the AI SDR that researches accounts and builds outreach — moved from a recurring monthly fee tied to how many contacts you enrolled to $1.00 per lead it recommends for outreach. The framing from HubSpot is that you now pay for the result, not the attempt.

The numbers HubSpot put behind it: the Customer Agent already resolves 65% of conversations and cuts resolution time 39% across more than 8,000 customers running it, and Prospecting Agent activations were up 57% quarter-over-quarter. Those are HubSpot's figures, measured HubSpot's way — useful as a direction, not as your forecast. Your resolution rate is a function of your ticket mix and your knowledge base, and it will not be 65% on day one.

This did not land in isolation. The same Spotlight shipped what HubSpot markets as 100+ updates (an audited 99: 13 new features, 86 enhancements), headlined by HubSpot AEO — Answer Engine Optimization, built on its October 2025 acquisition of XFunnel — which grades how your content surfaces inside AI answers. If you care about that thread, we cover it directly in our GTM AEO guide and you can run our free AEO Audit on your own site. But the pricing move is the one that changes a budget line, so let's stay on it.

The fine print that decides your bill: what "resolved" means

Outcome pricing only works if the outcome is honestly defined — and the vendor defines it. Per HubSpot's documentation, a conversation counts as resolved when the Customer Agent shares a relevant source or completes an action without a human stepping in (the window is 72 hours), or when it qualifies a lead. Under the hood that is 50 credits per resolved conversation (down from 100), with credits at $10 per 1,000 — which is where the $0.50 comes from.

Read that definition like an operator, not a buyer. A conversation where the agent confidently posts a help-doc link and the customer silently gives up still looks resolved to the meter. A "qualified lead" is qualified by the agent's own scoring. None of that is sinister — every outcome-priced vendor has the same structural issue, and HubSpot's $0.50 is real — but it means the number you must trust is not the price, it is the true resolution rate on your tickets. Before you scale, pull a sample of billed-resolved conversations and read them. If the agent's "resolved" matches yours, the economics are excellent. If a meaningful slice is deflection theater, you are paying $0.50 to annoy a customer, which is worse than free.

Does outcome pricing actually beat per-seat? The honest read by shape

Strip away the "pay when it works" language and the practical question is narrow: does this change your next move? Here is the read by situation.

Your situationWhat actually changedYour move
Already on HubSpot Pro/Enterprise, not yet running Breeze agentsReal win — the pilot just got cheap to run and easy to justifyTurn on the free trial, wire to a real queue/list, measure genuine resolution and audit the billed log before scaling
Running a rival support agent (Fin, Zendesk, Agentforce)HubSpot's $0.50 sticker undercuts the field on unit priceCompare on true cost-per-real-resolution and switching cost, not the headline rate — the cheapest meter can lose on definition and accuracy
Evaluating CRMs from scratchPricing model is a feature competitors will copy, not a moatPick the platform on data model, ecosystem, and motion fit; treat outcome-based agents as upside, not the deciding factor
Sub-team, cost-bound, or on a Starter/free tierLittle — the agents require Professional or EnterpriseDon't let the $0.50 unit pull you up a tier the seat math doesn't justify; a cheaper point tool may still win

The lesson lives in the top and bottom rows: outcome pricing is a reason to trial agents you were already positioned for — never a reason to climb a tier or switch platforms for a meter that competitors will match within a quarter.

The contrarian take worth holding

Jason Lemkin (SaaStr) makes the argument I find hardest to dismiss: outcome pricing is the right model today, but it is a transitional one. As resolution rates climb toward 80–90%, nearly every "use" becomes a "resolution," so the gap between per-use and per-resolution collapses — and the model becomes a footnote. Bret Taylor has called outcome pricing "obviously the correct way to build and sell software," and for support deflection in 2026 he's right. Just don't mistake a clever transitional meter for a permanent discount. The structural truth: AI support gets cheaper as it gets better, and you want a vendor whose incentives ride that curve with you. HubSpot's does. So does Intercom's with Fin — which is exactly why the AI customer-agent category is where the pricing innovation is concentrated.

If you want to pressure-test where these agents fit against the rest of your stack — and whether turning one on lets you retire a separate tool:

The honest caveats

Three to hold loosely. One: the headline stats are HubSpot's. 65% resolution and 39% faster are real averages across 8,000 customers, but your number depends on your ticket mix and knowledge base — budget for a lower rate until you measure your own. Two: "resolved" and "qualified" are HubSpot-defined billable events; sample the log and confirm they match your standard before you trust the meter. Three: this is two agents, not the whole platform. The Breeze Data and Content agents, and the broader suite, price differently, and the agents require a Professional or Enterprise tier — the $0.50 unit does not change the seat math underneath it.

None of that makes the move anything but smart. Aligning what you pay with what the agent actually delivers is the right direction for AI software, and HubSpot planting the flag as the category leader matters. It just means the buyer-side answer is "trial it and measure," not "switch to HubSpot for the pricing."

Running HubSpot and weighing an AI support or SDR agent? Pay-per-outcome makes the test nearly free — so run the test.

Affiliate link — StackSwap earns a commission if you sign up for HubSpot. We only partner with tools we'd recommend anyway.
Explore HubSpot →

FAQ

Yes. At its Spring 2026 Spotlight (announced to press April 2, live April 14, 2026), HubSpot moved two Breeze AI agents off conventional usage pricing onto outcome-based pricing. The Breeze Customer Agent went from $1.00 per conversation to $0.50 per resolved conversation. The Breeze Prospecting Agent moved from a recurring monthly charge based on enrolled contacts to $1.00 per lead it recommends for outreach. HubSpot's Chief Customer Officer Jon Dick framed it as: "Outcome-based pricing removes that risk. You pay when it works, full stop." HubSpot says the Customer Agent already resolves 65% of conversations and cuts resolution time 39% across more than 8,000 customers who have it live.

This is the part to read closely, because the vendor defines the billable event. Per HubSpot's documentation, a conversation counts as resolved when the Customer Agent shares a relevant source or completes an action without a human stepping in (the window is 72 hours), or when it qualifies a lead. Mechanically that is 50 credits per resolved conversation (down from 100), with credits priced at $10 per 1,000 — so $0.50. The operator point: "resolved" is HubSpot's definition, not yours. A deflection that technically closes the ticket but leaves the customer unhelped still bills. Before you trust the model, sample your own resolved conversations and confirm they were actually resolved to your standard, not just to the meter's.

On the sticker, yes. Per SaaStr's comparison set, Intercom's Fin is roughly $0.99 per resolution, Zendesk is about $1.50 committed (closer to $2.00 pay-as-you-go) plus a per-agent AI add-on, and Salesforce Agentforce runs around $2.00 per conversation. HubSpot's $0.50 undercuts all of them. But price-per-unit is not total cost. The number that decides your bill is your resolution rate times your volume times what each vendor counts as a billable resolution — and those definitions differ. I run Intercom 2 with Fin every day at my day job, so I price these against a tool I actually operate: the headline rate is the easy part; the definition of the billable event and the real deflection rate on your tickets are where the money is.

Both can be true. The honest counter-take comes from Jason Lemkin (SaaStr): as agent resolution rates climb toward 80–90%, the gap between "per use" and "per resolution" shrinks toward zero, because almost every use is a resolution anyway — so the model becomes a footnote for mature agents. Where outcome pricing genuinely helps you is at the start: low or uncertain volume, a team that has not proven the agent on its own data, and a CFO who does not want to pay for attempts that fail. It de-risks the pilot. It does not lower the price once the agent is good and busy. Buy it for the trial economics, not as a permanent discount.

HubSpot shipped what it markets as 100+ updates (an audited count of 99: 13 new features plus 86 enhancements, per Vantage Point's inventory). The headliner besides the pricing change was HubSpot AEO — Answer Engine Optimization, built on its October 2025 acquisition of XFunnel — which grades and improves how your content shows up in AI answers (ChatGPT, Perplexity, Google AI Overviews), priced around $50/month standalone and free in Marketing Hub Pro/Enterprise. Other notable adds: Smart Deal Progression (reads meeting transcripts plus deal history, drafts follow-ups, suggests CRM updates) and a free AEO Grader. The unifying pitch was "context" — that HubSpot's agents act on your full CRM history, not just a prompt.

By shape. If you already run HubSpot, this is a low-risk reason to actually trial the Customer and Prospecting agents — both ship with a free trial window, so wire them to a real queue, measure the genuine resolution rate on your tickets, and audit what "resolved" billed for before you scale. If you are evaluating CRMs from scratch, do not pick HubSpot because of the pricing model — pricing models are easy to copy and competitors will. Pick the platform on data model, ecosystem, and motion fit, then treat outcome-based agents as upside. If you are sub-team or cost-bound, note the agents require a Professional or Enterprise tier; the $0.50 unit does not rescue a plan whose seat math never worked.

Related reading

Canonical URL: https://stackswap.ai/hubspot-outcome-based-pricing. Sources: HubSpot company news — the outcome-based pricing announcement and the Spring 2026 Spotlight post (HubSpot, April 2026), including the Jon Dick quote and the 65% / 39% / 8,000-customer figures; pricing-model comparison and the "does it matter" counter-take per SaaStr / Jason Lemkin; effective-date corroboration via MarTech; the audited 99-update breakdown via Vantage Point. The 72-hour "resolved" window and credit mechanics are HubSpot's documented definition. Disclosure: StackSwap is a HubSpot affiliate. The analysis above is the same operator read we'd give a friend deciding whether to switch these agents on — including the team shapes where the answer is "don't." We affiliate with most vendors in this category and earn the same disclosed commission across them, so the read isn't shaped by who pays us.