MarketSizer's job is to put qualified opportunities in front of your team: accounts in your ICP, showing real purchase intent right now, and likely to be winnable for you based on outcomes in your segment. Nothing that fails any of those three filters reaches your list - ICP fit, real-time intent, and win likelihood are scored together, and only the accounts that clear all three appear at all.
The pricing model is built to complement that. Half of every opportunity's cost is deferred until MarketSizer detects a conversion. If the account never becomes a paying customer of yours, that half is never charged. That is the model in one sentence - and the rest of this post is the reasoning behind it.
The starting constraint was simple: the full cost of an opportunity should only be collected when the customer has evidence it was worth pursuing. Everything else - the upfront fee, the data surfaced, the narratives generated - had to be priced in a way that shared the risk. That constraint ruled out most of the obvious options immediately.
Most B2B software charges customers regardless of outcomes. Seat licences, platform fees, per-API-call meters: the vendor collects revenue whether the product delivers value or not. The buyer carries all the risk, the vendor carries none of it, and the relationship is built around an annual commitment that is settled long before anyone knows whether it worked. MarketSizer's pricing model was designed the other way around. The deferred half is the success fee, and it is the entire credibility lever of the model.
What Outcome-Based Pricing Means for Sales Intelligence
Intercom recently published their thinking on outcome-based pricing for Fin for Sales. Their core principle: you pay when Fin delivers value. For a sales AI, they landed on the qualified lead as the right unit - $10 per prospect that clears your criteria and gets routed to a rep.
It is a sound model, and the reasoning behind it is worth reading. But it illustrates something important about where most outcome-based pricing in sales AI currently draws the line: at readiness for human action. Fin qualifies the lead. The rep takes it from there. The pricing boundary sits at the handoff.
Intercom is not alone in pushing pricing closer to outcomes. Emergence Capital's framework for pricing AI software argues the same direction from a different angle: the more autonomous a product's actions and the more attributable their outcomes, the further pricing should move from seats and toward results.
MarketSizer operates at a different point in the funnel with a different data position, which is what makes a further-deferred fee structurally possible. The outcome MarketSizer charges for is not the qualified lead. It is the converted customer.
Three Models Considered and Rejected
Three alternatives were stress-tested before MarketSizer landed on the current model.
Per-seat or per-user. This is how Apollo, ZoomInfo, and most legacy data tools charge: a flat annual fee regardless of how much of the platform you use or whether any deal ever closes. The incentive structure is misaligned by design - the vendor collects full revenue whether the product delivers outcomes or not, which puts all the risk on the customer.
Per-API-call or pure consumption. You pay for every signal surfaced, every account enriched, every intent flag raised. Consumption pricing is now mainstream in B2B SaaS - OpenView Partners tracks it as the dominant evolution in software pricing - and it works well for infrastructure tools where every call is value delivered. It is the wrong fit for sales intelligence. It penalises teams for exploring the market thoroughly: the more accounts you look at, the more you pay, even when most of those accounts were correctly assessed as unwinnable and left alone. It also charges you for noise alongside signal, which is the opposite of what good data should do.
Revenue share. The purest version of outcome-based pricing: take a percentage of closed revenue from accounts MarketSizer surfaced. Intercom considered and rejected this for Fin, citing two problems: attribution (between qualification and close, many things outside the AI's control affect the outcome) and measurement (you need deep CRM integration to track deals to closure). MarketSizer arrived at the same conclusion, for the same reasons, plus one more: a revenue percentage creates a conflict of interest around which opportunities get surfaced. If the vendor earns more on larger deals, the platform has an incentive to weight itself toward them. That incentive does not belong in the model.
The Four Moments MarketSizer Charges For
MarketSizer's pricing has four moments. Each one is tied to a deliberate action the customer took.
Identify - 1 credit ($0.10). The customer views a qualified opportunity: ICP fit, purchase intent signals, win probability, estimated contract value. Unqualified accounts are free and unlimited. This fee exists because surfacing a pre-scored, pre-ranked account that fits your criteria and is actively in market has real value - but it is priced below the cost of missing it.
Activate - 10 to 50 credits, scaled by Opportunity Score. The customer decides to pursue the account and unlocks the full picture: competitive context, AI-generated narratives, battle cards, workflow assignment. This is half the total cost of an opportunity. The other half is deferred.
Contact - 1 credit per verified email, 5 credits per verified phone number. Unlocking the specific people to reach. Priced à la carte because outreach is selective - you do not need every contact at every account.
Convert - equal to the Activate fee. Charged only when MarketSizer detects that an account you Activated has become a paying customer of yours.
The relationship between Activate and Convert is the core of the model. When you activate a 4.0-scored opportunity, you pay 40 credits. If it converts, you pay 40 more. If it never converts - if you worked the account for months and it went cold - you pay nothing beyond that initial 40. The success fee is deferred indefinitely until the outcome is observable. If it never lands, neither does the fee.
How MarketSizer Detects a Conversion
Revenue-share pricing fails partly because of attribution and partly because of measurement. To know when a deal closes, you need to read from the customer's CRM - and as Intercom noted, that is a significant integration burden.
MarketSizer needed to solve the measurement problem without putting it on the customer.
Conversion detection works differently here. The primary path is objective subscription data: MarketSizer maintains a subscription intelligence dataset of 72 million historical subscription outcomes, tracking when companies start and stop paying for software products. When an account you Activated begins paying for your product, MarketSizer observes that transition in its own data - no self-reporting required, no CRM read access needed, no dependency on your pipeline updating on time.
Where the subscription dataset does not cover the market - non-SaaS products, off-platform contracts, vendors not yet ingested - you can mark a conversion manually in-app with the destination deal ID and contract date. That is the fallback, not the primary mechanism, and that distinction is what makes the model credible: the 72-million-row dataset is what removes the "you are just billing for whatever you say closed" objection.
When a conversion is detected, customers receive an in-app and email notification with the signal type, the account, the timestamp, and the credit cost. They have 30 days to dispute it. If the evidence is inconclusive, MarketSizer defaults to refunding the customer.
The Risk on the Other Side
The honest critique of any model that bills on detected outcomes is symmetric to the critique of revenue-share. If MarketSizer earns more by detecting more conversions, MarketSizer has an incentive to over-detect: to call conversions that did not really happen, or to attribute conversions to accounts that closed for unrelated reasons.
That risk is real, and the model has to take it seriously to be credible.
The mitigations are deliberate. Detection is grounded in objective subscription data, not self-reported pipeline. Every detected conversion comes with an audit-trail entry showing the signal type, the timestamp, and the Activate record it traces back to. The 30-day dispute window is generous, and the tie-break defaults to the customer: where the evidence is inconclusive, the charge is refunded. The reasoning behind that policy is straightforward - a refunded conversion costs less than a lost customer.
None of this eliminates the asymmetry. It just keeps the cost of getting it wrong squarely with the vendor.
Why the Fee Scales with Quality
The Opportunity Score is a composite of three signals: ICP fit (40%), purchase intent (35%), and win likelihood (25%). It runs from 0.0 to 5.0 in half-point steps.
The Activate fee scales directly with it. A 5.0-scored opportunity costs 50 credits to activate. A 2.5 costs 25. A 0.0 costs nothing.
This matters for two reasons. First, pricing reflects expected value at the moment of activation, not in hindsight. A 5-star account is more expensive because it converts at a higher rate - the fee is proportional to what the data suggests it is worth. Customers who activate lower-scored accounts are taking a more speculative bet, and they pay less for it.
Second, scaling the fee with quality keeps the scoring honest in both directions. A flat activation fee would create an incentive to inflate scores to justify the cost. Scaling means the scoring has to be calibrated correctly: over-scoring would make the product expensive for no reason, and customers would notice.
The total cost of a full-funnel 5-star converted opportunity is around $10. That is deliberate. The per-credit price is set at the customer's subscription tier, and the Activate-Convert symmetry is the credibility mechanism - not the extraction ceiling. Pricing below the unit cost of a lost deal is the whole point.
What You Never Pay For
It is as important to say what the model does not charge for as what it does.
You do not pay to browse the market. Unqualified accounts are free and unlimited to view - you can explore your entire ICP universe before spending a credit.
You do not pay a success fee after 12 months. The conversion window closes a year after activation. A deal that closes three years later, through channels disconnected from the account MarketSizer surfaced, does not trigger a fee.
You do not pay a second fee on expansions. If an Activated account converts and later expands, the conversion fee is not charged again. Re-Activating the same account for a different product line or business unit is a new Activation and a new conversion fee may apply - but the Activation record, not the company, is the unit of accounting.
You do not get a refund if the customer churns. The conversion fee fires once, at the moment of conversion, and is not clawed back if the account subsequently churns. The fee was for delivering the conversion, not for retention. Retention is your domain, not MarketSizer's.
You do not pay for seat licences or platform presence. Every credit charge is tied to an action you chose to take.
The Model in Practice
The critique of outcome-based pricing from the vendor side is usually that the vendor is carrying too much risk. Revenue is unpredictable. Long sales cycles mean fees land months after the work was done. Both are true. The model accepts them as part of the deal.
The case for accepting them is that the incentives line up throughout. MarketSizer has a direct financial interest in the quality of the opportunities surfaced, not just in the volume. A direct interest in accurate scoring, because inflated scores make the product expensive relative to outcomes. A direct interest in detection quality, because a conversion missed is revenue never collected. Every one of those interests is aligned with the customer's.
For a 5-star opportunity that converts with a verified email contact unlocked, the total is 102 credits end-to-end: 1 to identify, 50 to activate, 1 for the contact, 50 when MarketSizer detects the conversion. At entry-band pricing that is $10.20. At volume that is less.
For a 5-star opportunity that never converts: 51 credits. That is your downside on a top-rated account. Your downside is capped at the Activate fee. Your upside is not.
That is the model. The subscription intelligence data MarketSizer has built is what makes it possible to defer further than any other sales-intelligence vendor we have observed.
If the model works for you, the pricing page has the full breakdown. If it does not, that is worth knowing before you commit to anything - which is exactly why the 500-credit, 14-day trial requires no card. Book a demo when you are ready to see the accounts in your market that are actively evaluating right now.