Selling Outcomes Not Features: What Intercom Fin Taught Us
Most SaaS companies charge you for the privilege of using their software. Intercom decided to charge you for results instead.
When Intercom launched Fin, its AI customer service agent, priced per resolved conversation, the industry reacted with a mix of admiration and alarm. Admiration because it was intellectually elegant. Alarm because it exposed how much of the market still priced software by headcount rather than by value delivered.
That single pricing decision is one of the more consequential go-to-market shifts in SaaS since the per-seat model became the default. Its implications extend far beyond customer service software. For manufacturers deploying connected products at scale, the same logic applies, with equally disruptive consequences.
Outcome-Based Pricing Models in SaaS
| Pricing Model | Vendor Incentive | Buyer Risk | Transparency |
|---|---|---|---|
| Per-seat (traditional) | Expand headcount; poor quality tolerated | High; paying for promised value | Low; hard to measure ROI |
| Flat-rate platform | Maximize features; real outcomes invisible | Medium; locked into contract | Low; metrics not connected to results |
| Per-resolution/outcome | Fix real problems; direct revenue-to-impact link | Low; pay only for delivered results | High; every sale tied to specific outcome |
| Per-tag/deployment | Maximize successful deployments; scale with customer | Medium; depends on tag usage forecast | High; cost scales with production volume |
The point of the table is not to rank vendors. It is to make the structural difference between these models legible: who carries the risk, and whether the price the buyer pays moves with the value they actually receive.
The Bet That Changed Everything
What Intercom Actually Did
Intercom priced Fin around a simple premise: you pay when the AI resolves a customer issue without human intervention. It is worth being precise about what that did and did not change. Intercom's published plans still charge a monthly per-seat fee and add the per-Fin-outcome charge on top, so the outcome-based element sits alongside seat pricing rather than replacing it.
The arithmetic that makes this attractive is straightforward in principle. Suppose your blended cost per resolved support ticket using human agents is some figure £X. If a per-resolution price comes in below £X, the buyer saves money on every resolved conversation, before even factoring in availability, consistency, and the absence of sick days.
More importantly, the incentive alignment is total. The vendor only gets paid if the product works. Vendor and customer are pulling in exactly the same direction.
Why Per-Seat Was Always a Bad Deal
Per-seat pricing made sense when software was passive. You licensed a word processor, you got a word processor. The value was table stakes, and it was priced by headcount.
But AI changes the equation. When software actively does things, resolving tickets, capturing data, deflecting calls, closing loops, pricing it by the seat is like paying a law firm for the number of lawyers on retainer rather than the cases they win. The metric is wrong. It rewards mediocrity and penalises adoption.
Per-seat SaaS also creates a perverse dynamic in procurement: the more your team grows, the more you pay, even if the software is delivering proportionally more value. Outcome-based pricing flips this. The more value the software delivers, the more you pay, but you are always paying in proportion to what you receive.
Why Outcome Pricing Works: The Alignment Argument
Skin in the Game
The deepest virtue of outcome-based pricing is that it forces the vendor to care about deployment quality, not just contract signing. With per-seat, the deal closes at signature. With per-resolution, the deal is re-evaluated every single day based on whether the product actually works in the customer's environment.
This changes vendor behaviour in ways that matter. Support improves because churn costs real money. Onboarding becomes more rigorous because poor configuration means poor resolution rates. Product roadmaps shift toward reliability and resolution quality rather than feature count.
Nassim Taleb calls it "skin in the game." Vendors with skin in the game behave differently from vendors who have already been paid.
The CFO Conversation Changes
Outcome-based pricing also transforms how software is bought and justified internally. The traditional SaaS pitch to a CFO is: "We cost £X per seat per year, and here are some marketing decks hinting that we might reduce your costs." It is a faith-based argument dressed in spreadsheet clothing.
The outcome-based pitch is different: "We charge you £Y per resolved ticket. You currently pay £Z per resolved ticket using agents. Run us for 30 days and the ROI is self-evident." That is not a belief system. It is arithmetic.
For buyers, this removes procurement risk. For vendors, it removes sales friction. Outcome pricing is simultaneously better marketing and better product discipline.
The Criticism: Bill Shock and Assumed Resolutions
It would be dishonest to present outcome-based pricing as a clean solution without trade-offs. Per-resolution models attract real criticism, and those criticisms deserve engagement.
Bill Shock Is Real
When your cost structure is variable, forecasting becomes harder. A spike in product issues, a bad firmware update, a viral complaint thread, a seasonal surge, translates directly into a larger bill. For companies with tight monthly budgets, unpredictable software spend is a legitimate operational problem.
Caps and hybrid tiers can soften this, but the underlying tension remains. Outcome pricing benefits businesses with stable, foreseeable support volumes more than those with erratic demand.
What Counts as a Resolution?
The harder problem is definitional. What exactly constitutes a "resolved" conversation? A model might credit the AI when a customer confirms satisfaction, or when the conversation closes without escalation. But closed is not always resolved. A customer who gives up in frustration and calls the phone line instead is not a success story. Measuring outcomes requires measuring the right outcomes, and that measurement is harder than it looks.
Any vendor moving to outcome-based pricing needs honest, auditable definitions. Otherwise, you are not selling outcomes. You are selling metrics that look like outcomes.
The Minimum Viable Spend Problem
For smaller deployments, per-resolution pricing can be more expensive than flat-rate alternatives. If you are handling a low volume of support tickets, a fixed platform fee often beats paying per resolution. Outcome pricing tends to scale well at volume and can be inefficient at low volume. This is why many serious outcome-based models include a platform component to cover base infrastructure, with variable resolution fees layered on top.
The Connected Product Parallel
Here is where the Intercom story becomes directly relevant to manufacturers.
Connected product platforms, software that powers warranty registration, self-service support, parts ordering, and lifecycle engagement for physical goods, have traditionally been sold on seat or flat-rate models. Pay a monthly platform fee, get unlimited scans, and the vendor collects the same cheque whether you are capturing a handful of warranty registrations a month or a great many.
That pricing model can leave money on the table for successful vendors and create resentment among buyers who feel they are paying for a promise rather than a result.
What Outcomes Look Like for Physical Products
The Intercom model translates naturally into the connected product context. Consider three specific outcome metrics:
Registrations captured. Every warranty registration is a customer relationship that did not exist before. A manufacturer with no connected product platform has no first-party customer data from the point of sale. A platform that captures a registration has delivered a measurable, attributable piece of value. Pricing per registration, or per block of registrations, directly ties vendor revenue to customer acquisition value.
Tickets deflected. When a customer scans a product QR code, finds a guided troubleshooting flow, and resolves their issue without calling the support line, that is a deflection. The cost of a handled support call varies by industry and company size, but each deflected contact removes a unit of cost. A platform that deflects a meaningful volume of calls makes that saving explicit, provided the deflection is verified and auditable.
Parts ordered. When a product scan triggers a spare parts purchase, a filter replacement, a motor brush, a cable, the platform has generated direct revenue. Commission-per-transaction is the oldest form of outcome pricing, and it applies cleanly here. The platform earns when the manufacturer earns.
These are not theoretical constructs. They are measurable events that connected product platforms already track. The only question is whether the pricing model reflects them.
What a Hybrid Model Has to Get Right
The lesson generalises past any one vendor's price list. A model that ties some of the cost to the value actually delivered puts the vendor's revenue on the same side as the customer's outcome. A model that charges purely for access does not. Most workable structures end up as a hybrid: enough fixed revenue to cover genuine infrastructure cost, enough variable revenue to keep the alignment honest. Where a given platform draws that line is a question to put to the vendor directly.
The Future of Pricing: Everything Is a Conversion Funnel
Intercom's bet on outcome pricing was not just a tactical pricing decision. It was a statement about what software is for.
Software, at its best, is not a tool that humans use. It is an agent that produces results. The shift from "tool you use" to "agent that delivers" changes the appropriate pricing metaphor from a hardware rental to a performance contract.
This shift is visible across several SaaS categories, where vendors have begun experimenting with pricing tied to results rather than seats: recruitment platforms exploring per-hire pricing, revenue tools tied to pipeline rather than seats, legal AI priced per matter or per document, and manufacturing software experimenting with per-unit-produced fees.
Connected product platforms are not immune to this trend. The manufacturers buying these platforms today are increasingly sophisticated buyers. They have been through SaaS procurement cycles before. They know how to demand accountability. They will increasingly expect vendors to price for outcomes, and vendors who cannot demonstrate measurable outcomes will find it harder to justify flat-rate fees.
The Accountability Imperative
This is the deeper lesson from Intercom Fin. Outcome-based pricing is not just a revenue model. It is a public commitment to accountability. When you charge per resolution, you are saying: we are confident enough in our results to tie our income to them.
For connected product platforms, that same confidence needs to be demonstrated, in warranty registration rates, in support deflection data, in parts attachment rates, in customer lifetime value uplift. The pricing model and the metrics need to be coherent.
Vendors who can show that confidence, who can point to per-registration data, per-deflection figures, per-tag deployment ROI, will win the accounts that matter. Vendors who can only point to feature checklists will lose them.
What This Means for Manufacturers Evaluating Platforms
If you are a product manager or after-sales director evaluating connected product platforms, the Intercom lesson has a direct implication for your RFP process. Ask your vendors:
- What outcomes do you actually measure? Not features. Not capabilities. Outcomes: registrations, deflections, revenue.
- How does your pricing relate to those outcomes? Does the vendor have skin in the game?
- Can you show me third-party auditable outcome data? Not case studies written by the vendor's marketing team. Actual numbers.
- What happens to your pricing if your platform doesn't deliver? If there is no answer to this question, the pricing model is not aligned.
A platform that can answer these questions confidently is one that has thought seriously about value. One that deflects them with feature demos has not.
The Bottom Line
Intercom priced Fin per resolution and changed how the industry thinks about AI pricing. The insight was simple: if your product delivers value, price for the value. If it does not, price for the features and hope nobody notices.
The connected product industry is approaching the same inflection point. Manufacturers are beginning to demand evidence, not promises. Platforms that can tie their pricing to measurable outcomes, registrations, deflections, parts revenue, will define the category. Those that cannot risk being renegotiated out of enterprise accounts.
Selling outcomes instead of features is not a pricing strategy. It is a business model built on confidence that your product actually works.
The vendors who win the next decade will be the ones who were confident enough to charge for results, and disciplined enough to deliver them.
FAQ: Outcome-Based Pricing and Vendor Alignment
How do I know if a vendor's outcome pricing is real or if they're just hiding per-seat cost in different language?
Ask three specific questions: (1) Show me your per-customer auditable outcome data, registrations, deflections, revenue, across multiple reference customers. (2) What happens to your pricing if you miss your committed outcome targets? (3) Can you show me a case where a customer paid less because outcomes underperformed? Vendors confident in outcome alignment will have transparent, third-party auditable data and a clear answer to "what do you owe us if you don't deliver?" If they deflect to features or case studies written by their marketing team, they are selling outputs, not outcomes.
If I sign an outcome-based contract, how do I protect myself from bill shock if outcomes spike?
Legitimate outcome models include cost caps, tiered pricing, or hybrid structures (base platform fee plus outcomes). A vendor that offers unlimited per-outcome pricing without caps may be protecting itself with aggressive pricing assumptions. Ask for a time-boxed pilot with a clear, capped cost. Measure outcomes yourself in parallel with the vendor's measurement. Demand data export and third-party audit rights. A vendor confident in their outcomes will accept these terms.
What's the minimum scale where outcome pricing makes sense for connected products?
It depends on volume. At low production volumes, fixed platform fees often result in lower total cost, because the variable component is too small to offset the overhead of metering it. As annual tag or unit volume rises, per-tag pricing becomes more attractive and more transparent, and per-unit economics tend to improve through volume discounts. For smaller volumes, negotiate a fixed base fee plus lower per-outcome rates (registrations, deflections) rather than per-tag; this aligns incentives while keeping cost predictable.
BrandedMark is the Product Operating System for manufacturers of physical goods: serialised product identity, connected experiences, warranty registration, and Digital Product Passport readiness in one platform. See how it works at brandedmark.com.