Earn-Back Clauses
Definition
Earn-Back Clauses
Earn-back clauses let a provider recover the service credits it has already accrued by delivering sustained performance over a later period. The credit is suspended rather than forgiven until the recovery test has either been passed or failed outright.
The logic is recovery, not forgiveness — a provider that misses a target in March and then exceeds it through April, May and June has arguably restored the value the buyer lost.
Buyers agree to earn-back because it changes provider behaviour after a failure. Without it, a missed month is a sunk cost, and the incentive to recover quickly is weaker than the incentive to manage the next month’s numbers.
The risk is obvious enough. An earn-back window that is too generous converts every service level into a rolling average, and rolling averages hide exactly the failures a service level exists to catch.
Key takeaways
- Earn-back recovers credits already accrued, rather than preventing them arising.
- Recovery normally requires sustained over-target performance across consecutive periods.
- Generous windows dilute service levels into rolling averages.
- Critical service levels are usually excluded from earn-back entirely.
How it works
A credit accrues on a miss and is held rather than deducted. If the provider then meets or exceeds the target for an agreed run of consecutive periods, the held credit is released. If not, it is deducted.
Public contracting uses an almost identical device for payments. Under the federal construction payments clause, the contracting officer may retain up to ten percent where “satisfactory progress has not been made” — and releases it when progress recovers.
The release condition is the important half. That clause requires the officer to “release to the Contractor all the remaining withheld funds” once the work is substantially complete, which is what makes the withholding a lever rather than a fine.
Design has to resist the obvious gaming route. The UK Sourcing Playbook’s instruction to “incentivise delivery of the things that matter” cuts against earn-back on measures where a single miss does real damage.
| Design parameter | Tight version | Loose version |
|---|---|---|
| Recovery window | Three consecutive periods | Any three months in a year |
| Recovery threshold | Above target, not merely at it | Meeting target is enough |
| Eligible measures | Non-critical service levels only | All measures including critical ones |
| Proportion recoverable | Half the accrued credit | All of it |
| Repeat misses | Earn-back right lapses | Right resets each time |
The eligible-measures row is where most damage is done — allowing earn-back on a safety, security or regulatory measure tells the provider that a breach is temporary in cost as well as in fact.
Examples
Earn-back works where recovery genuinely restores value and fails where the damage was done at the moment of the miss. These four cases show both.
A software provider misses availability in one month, then exceeds target for three consecutive months, and recovers half the accrued credit. The buyer gets a restored service and the provider gets an incentive to fix it fast.
A contact centre provider recovers credits by beating its answer-time target across a quarter. Service is measurably better than it was before the miss.
A payments processor negotiates earn-back across all measures including a regulatory reporting deadline. The deadline is missed, reported to the regulator, and the credit is later recovered anyway.
A logistics provider faces a clause where earn-back lapses after two misses in a rolling year. It uses the right once, then manages carefully, which is the behaviour the clause was designed to produce.
Related terms
Earn-back is one of three mechanisms that move money after a performance failure, and the differences are about timing. The entries below place each one on that timeline.
- SLA linked pricing: fee levels that track attainment continuously rather than by credit and recovery.
- Service level agreement (SLA): the agreement setting the targets earn-back is measured against.
- Penalty hold: money withheld pending an outcome, which earn-back closely resembles in practice.
- Performance based pricing: a model placing fee at risk against results rather than against misses.
- Gain sharing outsourcing: sharing a saving achieved, not recovering a credit lost.
- Key performance indicator (KPI): the measures eligibility for earn-back is decided across.
- Service level compliance: the attainment record a recovery claim is tested against.
FAQ
How is earn-back different from a service credit?
A service credit is the money the provider loses on a miss. Earn-back is the route by which that money can be recovered through later sustained performance.
Which measures should be excluded?
Anything where the harm is done at the moment of failure: safety, security, regulatory deadlines and data breaches. Recovery cannot undo those.
How long should a recovery window be?
Three consecutive periods is the common setting. Windows allowing any three months in a year turn a monthly service level into an annual average.
Should all of a credit be recoverable?
Often not. Recovering half preserves some consequence for the original miss while still rewarding a fast fix.
Does earn-back weaken an SLA?
It can, if drafted loosely. Tight thresholds and excluded critical measures keep the service level intact while still rewarding recovery.
Do buyers gain anything from agreeing to it?
Yes, when it works. A provider with a route back concentrates on recovery rather than writing off the month and defending the next one.
Source partners prepared to work under earn-back terms are listed in the Outsource Accelerator hub directory.
Independent
