Consumption Pricing
Definition
Consumption Pricing
Consumption pricing charges an outsourcing client for measured usage rather than for staff or a fixed period. The meter replaces the headcount as the unit of value, so cost tracks demand and the provider carries the risk that demand never arrives.
Cloud platforms normalised the model and business services followed. Buyers who pay per invoice processed or per ticket resolved are using the same idea.
The appeal is obvious in seasonal work — you pay for December’s volume in December rather than staffing all year for it.
The catch is equally predictable. A model with no floor and no ceiling is a budget nobody can forecast — and finance teams notice.
Key takeaways
- Charging follows measured units of consumption rather than time or headcount.
- Cost moves with demand, which suits volatile or seasonal volumes.
- Unit definition matters more than unit price, because the definition decides what gets counted.
- Metering data often arrives days after the usage it describes.
How it works
Three things have to be settled before the model works: what counts as a unit, how it is measured, and who can see the measurement. Everything else is arithmetic.
Cloud billing shows the pattern at its most developed. Microsoft documents that for enterprise agreements “cost and usage data is typically available in Cost Management within 8-24 hours”, while pay-as-you-go subscriptions can take up to 72 hours.
That lag matters commercially. Estimated charges for the current period are updated six times a day and remain estimates until an invoice is issued, so real-time cost control is never quite real time.
| Design question | Weak answer | Strong answer |
|---|---|---|
| What is a unit? | “A transaction” | A defined, countable event with named exclusions |
| Who measures it? | The provider, reported monthly | A shared system both parties can query |
| What if volume collapses? | Nothing agreed | A floor, or a term that reprices at a threshold |
| What if volume spikes? | Provider absorbs it | Tiered rates and an agreed ramp-up notice |
Public procurement solved a version of this long ago. A requirements contract “provides for filling all actual purchase requirements of designated Government activities” for a stated period, from one contractor.
That is consumption pricing with an exclusivity commitment attached, which is how the public sector buys variable volume without leaving the supplier exposed.
Unit definition is where disputes originate — a “ticket” that counts reopened items separately produces a very different invoice from one that does not, and the difference rarely surfaces until month three.
Examples
Consumption models work best where volume genuinely varies and worst where it merely looks as though it does. The four cases below show both patterns, and the last one shows what happens when nobody watches the meter.
An online retailer pays its support provider per contact handled. Volume triples in the six weeks before Christmas and falls back in January, and the cost profile follows it exactly.
A claims processor is billed per adjudicated claim rather than per full-time equivalent. Productivity improvements now benefit the provider rather than the client, which changes who invests in automation.
A finance function pays per invoice processed and then centralises three subsidiaries onto the same contract. Volume rises, the unit rate drops through a tier, and total cost rises less than proportionally.
A software company moves to usage-based infrastructure billing without setting alerts. A misconfigured job runs for nine days before anyone sees the charge, because the data lagged the usage.
Related terms
Consumption pricing sits among several commercial models that allocate risk differently. The entries below cover the neighbouring structures and the contract artefacts that make any of them workable.
- Managed services: the delivery model most often priced this way.
- Cloud managed services: where metered billing originated and is most mature.
- IT managed services: a common home for per-device and per-user charging.
- Rate card: the input-priced alternative, and the usual comparison.
- Total contract value outsourcing: harder to state when volume is uncertain.
- Service level agreement (SLA): the quality guarantee that has to survive volume swings.
- Unit cost of production: the underlying economics the unit price has to cover.
FAQ
How is consumption pricing different from a rate card?
A rate card prices time. Consumption pricing prices output. One counts hours supplied, the other counts work completed.
Does it always cost less?
No. It costs less when volume falls and more when it rises. Over a stable year it often lands close to a fixed-fee equivalent.
Who carries the risk?
The provider carries volume risk and the client carries unit-definition risk. That split is what makes the definition worth negotiating hard.
What is a tier?
A volume band with its own rate. Tiers give the provider recovery at low volumes and the client a discount at high ones.
Why do buyers dislike the unpredictability?
Because annual budgets are fixed and consumption is not. Minimum commitments and caps are the usual answer.
Can quality be protected under this model?
Yes, through service levels priced into the unit.
Compare providers offering usage-based commercial models in the Outsource Accelerator directory.
Independent
