Exit provisions are negotiated at the end of a deal, by people who are tired, on a document everyone wants signed. They are also the provisions most likely to be read under time pressure, in a bad mood, several years later. That asymmetry explains a great deal about why supplier exits go badly.
A termination assistance clause, sometimes labeled transition assistance or disengagement services, is the mechanism that keeps a service running while it moves. Without one, termination is a cliff edge, and the supplier whose contract is ending has limited commercial interest in making the transition smooth. With a weak one, the customer discovers at notice that assistance is discretionary, chargeable at rates to be agreed, or limited to a period shorter than the migration takes.
The provision matters most in regulated environments, where supervisors expect institutions to demonstrate that critical services can be moved or brought back in-house without disruption. But the underlying mechanics are the same in any outsourcing arrangement, and the questions a vendor manager should ask do not change with the sector.
What a workable assistance clause contains
The clause is a set of promises about behavior during a period when goodwill is scarce. Each element should be specific enough that performance can be assessed rather than argued.
- A defined assistance period, a fixed duration running from notice or expiry, with an extension right held by the customer at pre-agreed rates rather than by mutual agreement.
- Pre-agreed commercial terms, rate cards or a formula for assistance services, set at signature, because rates negotiated after notice is served reflect the leverage of the moment.
- Data return in a named format, not simply a return obligation but a specified format, medium and deadline, plus a certification that no copies are retained beyond agreed retention.
- Successor supplier cooperation, an obligation to work constructively with an incoming provider, including knowledge transfer sessions and documentation, which many clauses omit entirely.
- Assistance on any termination, the obligation should survive termination for the supplier's breach, since that is precisely the circumstance in which assistance is most needed and least likely to be volunteered.
A useful test is to read the clause and ask what the customer would do on the first morning after serving notice. If the answer involves negotiating anything at all with the supplier, the clause is incomplete and should be treated as a renewal-time amendment target.
Rights to data extracts, documentation and knowledge transfer are frequently exercised for the first time during an actual exit, which is when any gap becomes expensive. Requesting a full data extract and a current runbook during normal operations converts an assumption into a known quantity, and costs very little while the relationship is healthy.
The exit artifacts to hold before you need them
Contract language sets the entitlement. These artifacts are what make the entitlement usable, and each should be a tracked deliverable with a refresh cycle rather than a one-time request.
- A current exit plan. A short document describing how the service would move, who would receive it, how long it would take and what it would cost, refreshed annually rather than written once during onboarding.
- A tested data extract. Evidence that a complete export has been produced in the contracted format within the contracted timeframe, with a note of anything that proved unavailable.
- Current service documentation. Runbooks, configuration records and process maps held by the customer, not only by the supplier, and updated on an agreed cadence.
- A dependency map. The subcontractors, integrations and third party licenses the service relies on, since these determine whether a migration is weeks or quarters of work.
- An assistance cost estimate. What the contracted assistance period would actually cost at the agreed rates, so the exit decision can be made with a number attached to it.
A five-step roadmap to exit readiness across the portfolio
Step 1, tier the portfolio by exit difficulty
Criticality alone is the wrong lens. Rank suppliers by how hard the service would be to move, which combines data volume, integration depth, regulatory constraint and the availability of alternatives. The hardest exits deserve attention first regardless of contract value.
Step 2, extract the exit terms into fields
Pull the assistance period, rate basis, data return format and deadline, and successor cooperation obligations from each agreement into structured data so the portfolio can be assessed in one view rather than one document at a time.
Step 3, score each clause against a standard
Define what good looks like for your institution and score existing clauses against it. The output is a prioritized amendment list, which is far more actionable than a general observation that exit terms are weak.
Step 4, fix the worst clauses at renewal
Exit terms are cheapest to improve when the supplier wants the renewal. Carrying the amendment list into renewal negotiations converts a compliance gap into an ordinary commercial ask at the one moment leverage exists.
Step 5, rehearse annually for critical services
Run a desktop exercise on the hardest exit each year. It surfaces missing documentation, unknown dependencies and unusable data formats while there is still time to address them contractually.
See exit terms across the whole supplier book
Aavenir extracts assistance periods, data return obligations and exit rates from executed agreements into one reviewable portfolio view.
Why exit terms drift out of date
Exit provisions age badly for a reason that has nothing to do with drafting quality. The service described in the original agreement is rarely the service being delivered five years later. Volumes grow, new integrations are added, the supplier acquires a subcontractor, additional data types come into scope, and none of that flows back into the exit provision or the exit plan. The clause continues to describe a migration of a service that no longer exists in that form. Institutions that stay ready treat the exit plan as a living artifact tied to the contract record, refreshed when the service changes materially rather than only at renewal, and they keep the dependency map current because that is the part most likely to have moved. It is the same discipline that keeps any other post-signature commitment honest, described in obligation management.
Exit provision benchmarks by service type
Assistance periods and data complexity vary by what is being outsourced. The patterns below give a starting point for assessing whether an existing clause is proportionate to the migration it would have to support.
| Service type | Typical assistance period | Main migration constraint | Priority artifact |
|---|---|---|---|
| Core processing | Twelve months or more | Data model and integrations | Tested extract and schema |
| Cloud infrastructure | Three to six months | Architecture dependencies | Current dependency map |
| Business process outsourcing | Six to twelve months | Trained staff and knowledge | Current runbooks |
| SaaS application | Thirty to ninety days | Data export format | Verified export sample |
| Managed security | Three to six months | Continuity of monitoring | Parallel running plan |
Exit readiness is a contract management discipline rather than a legal one. The clause creates the entitlement, but the entitlement is only worth what the supporting artifacts make it worth, and those artifacts decay quietly between renewals. The practical program is small: know which exits would be hardest, know what each contract actually entitles you to, hold a current plan and a tested extract for the services that matter, and carry the amendment list into every renewal. Institutions that do this are not the ones with the best drafted clauses. They are the ones for whom serving notice is a decision rather than a leap. The work is cheap while the relationship is healthy and impossible once notice has been served, which is the entire argument for doing it early. For the wider vendor oversight picture, see third party risk contract automation.
For related guidance, see The DORA guide to ICT contract registers and exits, Vendor contract controls for OCC and FFIEC exam readiness and Complementary user entity controls in vendor contracts.
