Why Connected Product Platforms Should Charge by Usage
Key Takeaways
- Over 70% of new SaaS entrants now offer usage-based pricing components, up from under 30% five years ago, according to Gartner — reflecting a market-wide shift away from seat counts as a proxy for value.
- Seat-based pricing is structurally decoupled from connected product value: eight platform users managing 400,000 physical products pay the same as eight users managing 4,000.
- Per-product-identity pricing aligns vendor incentives directly with brand outcomes — the vendor earns more only when the customer provisions more connected products and therefore generates more value.
- Usage-based pricing is typically 60–70% cheaper than enterprise seat licensing at volumes of 5,000–50,000 units per year, and scales proportionally with production volume rather than headcount.
The SaaS industry spent two decades telling buyers that seat counts were a proxy for value. They were wrong, and the market is now correcting at scale.
Stripe never charged per employee managing transactions. Snowflake doesn't bill by headcount. Twilio has never cared how many people sit in your engineering org. These companies built their pricing around the actual unit of value — the transaction, the query, the API call — and they became the defining infrastructure platforms of their respective categories.
| Key Metric | Value |
|---|---|
| Usage-based SaaS entrants | 70%+ of new platforms (up from 30% five years ago) |
| Seat pricing accuracy | 0%—completely decoupled from product value delivery |
| Connected products per platform seat | 10,000–500,000 (no correlation) |
| Per-unit ROI | Often higher at SMB scale than enterprise |
| Vendor incentive alignment | Misaligned with seat pricing; aligned with usage pricing |
Pricing philosophy differences: Registria and Narvar use hybrid seat + usage models. NeuroWarranty and Dyrect use seat-only pricing. BrandedMark uses pure usage-based (per-product-identity) pricing—the only model that structurally aligns vendor growth with customer success.
Connected product platforms have been slow to follow. Most still arrive with a deck full of tier tables and seat bundles, as though the number of people who log into a dashboard has anything to do with what a manufacturer actually needs: products connected, customers known, service costs reduced, aftermarket revenue captured. That mismatch isn't just a billing inconvenience. It's a structural misalignment between vendor incentives and brand outcomes.
It needs to change.
Why Seat Pricing Was Never the Right Model
Seat-based pricing made sense for CRM software because CRM value scaled with adoption. Salesforce charged per seat because ten salespeople using the platform captured roughly twice the value of five. Seat count was a reasonable proxy for utilisation, and the model aligned cost with usage. That logic does not transfer to connected product platforms. A connected product platform's job is not to maximise dashboard logins — it is to maximise what happens to every physical unit a manufacturer ships. The value unit is a serialised product leaving the production line with a digital identity attached, not a user opening a browser tab. Eight people managing 400,000 connected products generate fundamentally different business outcomes than eight people managing 4,000 — yet seat-based pricing treats both identically. That is not a minor billing quirk. It is a structural misalignment built into the commercial foundation of the relationship from day one.
The Disconnect in Practice
Consider a mid-sized appliance manufacturer running 400,000 units per year across four product lines. Their connected product platform might be actively used by:
- 2 product managers configuring experiences
- 1 digital ops lead monitoring scan rates
- A handful of brand and compliance stakeholders reviewing content
That's perhaps eight seats on a good day. Yet those eight people are orchestrating digital engagement for four hundred thousand physical products — each capable of generating warranty registrations, support deflections, spare parts orders, and customer relationship data for its entire service life.
Seat-based pricing charges for the eight. It has no relationship whatsoever to the four hundred thousand. If those products generate ten million scan interactions over their lifetimes and drive six figures in direct aftermarket revenue, the platform vendor captures none of that upside and has no incentive to help the brand maximise it.
This isn't just a pricing quirk. It's a misalignment of incentives at the foundation of the relationship.
How the Rest of SaaS Has Already Solved This
The SaaS industry's most durable infrastructure platforms converged on usage-based pricing because the underlying logic is structurally sound: when a vendor's revenue is tied directly to the unit of value it delivers, both parties are automatically on the same side. Stripe prices per transaction. Snowflake prices per compute credit consumed. Twilio prices per API call. None of these companies charge by headcount, because headcount has no relationship to the value their platforms generate. The shift was not accidental — it reflected a deliberate recognition that seat licensing creates a misalignment where vendors are incentivised to expand user counts rather than improve customer outcomes. That pattern is now well understood by enterprise buyers. Three examples from established platforms illustrate how usage-based alignment plays out in practice and why the model is directly applicable to connected product infrastructure.
Stripe: Charge Per Transaction
Stripe's model is elegant precisely because it requires no justification. Every payment processed is a unit of business value for the merchant. Stripe takes a small slice. If the merchant processes more payments, Stripe earns more. The vendor's growth is structurally coupled to the customer's growth. Neither party needs to renegotiate to stay aligned.
Stripe doesn't charge per developer who integrates the API, per finance team member who checks the dashboard, or per geographic market the merchant operates in. The transaction is the thing. Everything else is overhead.
Snowflake: Charge for Consumption
Snowflake's consumption credit model was controversial when it launched. Analysts questioned whether variable pricing would make enterprise buyers nervous. Instead, it became a competitive moat.
The model works because it eliminates the negotiation overhead of seat licensing. Buyers don't have to forecast headcount twelve months in advance. They consume what they need. Snowflake's revenue scales with data volume — which scales with business activity — which is exactly what should drive infrastructure pricing.
Intercom: Per Resolution, Not Per Agent
Intercom's shift to per-resolution pricing for its Fin AI agent is the most instructive recent example for the connected products space. The legacy model charged per support seat — a proxy for team size, not for outcomes. The new model charges per resolved conversation. If Fin resolves a customer issue, Intercom earns. If it doesn't, they don't. Intercom has publicly stated that this pricing shift, which aligns vendor revenue directly with customer outcome, accelerated Fin adoption faster than any feature release in the product's history.
That model changes the entire product conversation. Intercom has a direct financial incentive to make Fin better at resolving issues. The customer has a direct financial incentive to deploy Fin on more channels and more use cases. Both parties grow together.
The parallel to connected products is exact. A platform that charges per product identity created has a direct incentive to make that identity as valuable as possible — better scan experiences, richer warranty capture, more effective support deflection, higher aftermarket conversion.
The Right Unit for Connected Products: The Product Identity
The right pricing unit for a connected product platform is the serialised product identity — a unique digital record created at manufacturing or packaging time that ties a specific physical item to a digital experience, a lifecycle record, and a customer relationship. Stripe's unit is the transaction. Intercom's unit is the resolved conversation. For connected products, the equivalent is the moment a physical unit gains a scannable digital identity and enters the world capable of generating value. Every identity created represents a warranty registration opportunity, a self-service support channel, a spare parts discovery surface, a Digital Product Passport compliance record, and an authenticity signal against counterfeiting. These are not one-time events — they compound across a product's entire service life. Pricing on this unit means a vendor's revenue is directly tied to the volume of value-generating touchpoints a brand deploys, not to how many employees access the configuration dashboard.
What Per-Identity Pricing Looks Like
A production-volume model prices platform access based on the number of product identities provisioned. A manufacturer running 100,000 units a year pays for 100,000 identities. If they scale to 500,000, they pay proportionally more — and they get proportionally more value from the platform in the form of warranty registrations, support deflections, and aftermarket sales.
The model has several structural advantages over seat licensing:
Predictable budgeting tied to production. Manufacturers know their production volumes. Budgeting for a platform tied to that volume is no different from budgeting for labels, packaging, or components. It's a per-unit cost of goods consideration, not a murky headcount forecast.
Incentive alignment at the vendor level. A vendor earning revenue per product identity provisioned has every reason to maximise the value of each identity. Scan rate improvements, experience quality, conversion optimisation — all of these directly serve the vendor's commercial interest because they make brands want to provision more identities.
Scales with the business, not the org chart. As a brand grows, production volumes grow. The platform cost scales with the business in a way that seat pricing never could. There's no renegotiation required, no awkward conversation about whether a new product line counts as a new seat tier.
Aligns with how manufacturers think. Brand and product teams at manufacturing companies don't think in terms of software seats. They think in terms of SKUs, production runs, and units shipped. A pricing model that maps to production volume speaks their language.
The TCO Argument: Why Cheap Seats Aren't Actually Cheap
Seat-based pricing often appears cheaper at contract signature. That comparison rarely survives contact with the actual cost picture. The true cost of disconnected products includes customer acquisition spend that warranty registration data would have offset, support call volumes that self-service deflection would have reduced, and aftermarket revenue that a competitor captured because there was no scan-to-buy touchpoint on the physical product. None of these costs appear in a SaaS contract line item, but all of them are real and measurable. Per-product-identity pricing changes the evaluation framing entirely. Instead of asking what the software costs per year, the question becomes what return each connected unit generates — in warranty registrations, support deflections, and aftermarket transactions — against a known per-unit provisioning cost. That is a healthier question for every stakeholder in the decision, and it surfaces the full economic case for connecting products rather than obscuring it behind a seat count comparison.
A Simplified ROI Model
Take a manufacturer provisioning 200,000 product identities per year at a modest per-identity cost. If those identities drive:
- 8% warranty registration rate — 16,000 direct customer relationships captured per year, each with meaningful lifetime value
- 12% support deflection rate — roughly 24,000 support interactions resolved via self-service rather than call centre
- 2% scan-to-purchase conversion on spare parts and accessories — 4,000 direct transactions per year
The connected product ROI from those three vectors alone typically dwarfs the platform cost by an order of magnitude. The pricing structure should reflect that leverage — and per-identity pricing does, because it scales with the volume that generates the return.
What This Means for Platform Selection
When evaluating connected product platforms, the pricing model is not a secondary consideration — it is a leading indicator of whether vendor and brand incentives are structurally aligned. A vendor charging flat seat fees has no financial stake in your scan rates, warranty capture, or aftermarket conversion. Their revenue is fixed regardless of platform performance. A vendor charging per product identity provisioned earns more only when you provision more units — which happens when the platform is performing and you are confident enough in the return to expand. That is the self-reinforcing relationship a long-term infrastructure partnership should have. The post-purchase infrastructure a brand builds today will shape its direct customer relationships, service economics, and aftermarket revenue for years. Choosing a commercial model that keeps the platform partner invested in those outcomes from day one is not a procurement detail — it is a strategic decision about who your vendor is actually working for.
The Broader Shift Is Already Underway
Gartner estimates more than 70% of new SaaS entrants now offer usage-based pricing components, up from under 30% five years ago. That shift reflects a structural change in how enterprise buyers evaluate commercial models — procurement teams are asking earlier in the process whether vendor revenue tracks directly to the value the customer receives. Connected product platforms have been slower to follow than other SaaS categories, but the pressure is building as brands become more sophisticated about what they are actually purchasing. The answer should be clear: a brand is paying for product identities — for the infrastructure that gives each manufactured unit a digital life, a scannable touchpoint, and a direct customer relationship. Seat counts measure how many people opened the dashboard. Product identities measure how many products were connected, how many customers are known, and how many aftermarket opportunities were created. One is a proxy. The other is the commercial purpose of the platform.
FAQ: Usage-Based Pricing for Connected Products
How is "per-product-identity" defined? Does it count every SKU variant separately?
A product identity is one scannable, serializable unit. For non-serialized products (e.g., model-level SKUs), you're provisioning one identity per SKU per year. For serialized products, it's one identity per physical unit manufactured. A manufacturer shipping 50,000 units of 10 SKUs has 50,000 identities. If they ship 25,000 units next year, they pay for 25,000. It's proportional to volume—exactly how manufacturing costs scale.
What if we have seasonal volume spikes? Do we pay for peak or average?
You pay for actual units provisioned. If you provision 100,000 units in Q4 and 20,000 in Q1, you pay for 120,000 total annually. Platforms that charge monthly can have a mismatch between your production calendar and billing calendar. Ask whether the pricing structure has monthly reset, annual smoothing, or some hybrid. BrandedMark allows quarterly true-ups—you forecast annual volume, adjust quarterly as needed, true-up at year-end.
How does this compare cost-to-cost with seat-based platforms?
At 5,000–50,000 units/year, usage-based is typically 60–70% cheaper than enterprise seat pricing. At 200,000+ units, seat licensing may appear cheaper on first glance—but per-identity pricing compounds with value: every additional unit drives warranty registrations, support deflections, parts revenue. Seat pricing is flat regardless of whether your connected product ROI is 2x or 10x. That's the fundamental difference.
Is there a maximum volume where usage-based becomes disadvantageous?
No. Companies shipping 5 million units/year still benefit from usage-based pricing because their cost scales with their value generation. A manufacturer shipping 2x the units should expect to provision 2x the identities and pay proportionally more. That's natural. Seat-based pricing breaks at scale because eight product managers managing 5 million units is the same cost as eight managing 500,000 units—which is clearly wrong.
Can we lock in a price if we commit to annual volume?
Yes. Most usage-based platforms offer committed volume discounts—if you forecast 500,000 units and commit, you lock a per-identity price. This gives you budget certainty while maintaining the incentive alignment of usage-based pricing. If your actual volume exceeds the commitment, you pay overage at a predetermined rate. It's more flexible than a fixed contract and more predictable than pure monthly metering.
BrandedMark prices by production volume — product identities provisioned, not dashboard logins. If you're evaluating how to build a connected product strategy that aligns cost with value from day one, explore how BrandedMark works.