Start with the deliverables, not a multiplier
The UK government's Digital, Data and Technology Playbook says exit plans should cover activities, milestones, resources, roles, risks, dependencies, and transfer of data and knowledge assets. Its published exit schedule also calls for asset, contract, configuration, and operating-procedure records. These public-procurement terms are not a contract template for every buyer, but they expose work that a one-line “handoff included” promise can hide.
Estimate at least these rows:
| Work package | Quantity to request | Evidence |
|---|---|---|
| Asset and dependency inventory | hours | reviewed register |
| Data export and validation | hours plus storage or transfer | sample export and reconciliation |
| Environment reconstruction | hours plus services | non-production build |
| Knowledge transfer | sessions and preparation hours | agenda, recording, open questions |
| Parallel operation | people, hours, and service overlap | dated cutover plan |
| Acceptance and recovery drill | hours | signed test results |
| Contract and license transfer | adviser time and fees | written disposition |
| Decommissioning | hours | deletion and access-revocation evidence |
Owning source code helps, but it does not fill these rows automatically.
Build three visible scenarios
For each row, enter low, base, and high quantities and rates. Then calculate:
row cost = quantity × rate
scenario subtotal = sum of row costs
scenario total = subtotal × (1 + chosen contingency percentage)
Do not use one blended rate if buyer staff, the outgoing provider, a replacement provider, and a specialist adviser charge differently. Record whose rate it is. For fixed fees, use a quantity of one and keep the quoted amount visible.
The accompanying DealDigg worksheet is blank by design. Its dependency-free calculator passed three tests on Node.js 22.18.0: scenario arithmetic, buyer-selected contingency, and rejection of negative inputs. The numbers in the tests are synthetic arithmetic fixtures, not market evidence.
Ask the current provider for evidence
Turn unknowns into priced requests:
- What data, files, configuration, histories, and identity mappings can be exported, in which formats, and how long does a representative export take?
- Which cloud, domain, repository, package, monitoring, and support accounts are controlled by the buyer?
- Which third-party licenses can transfer, must be repurchased, or require a substitute?
- Can a replacement operator build, deploy, restore, observe, and change the application from the proposed handoff?
- How many included transition hours are available, at what notice, and what rate applies afterward?
Unanswered questions belong in the high scenario or the risk register—not in a footnote that silently assumes zero cost.
Compare the estimate with the quote
Keep implementation, ongoing operation, and replacement as separate columns. A proposal may reasonably have a higher initial cost and lower replacement cost because accounts, documentation, tests, and recovery are delivered as the work proceeds. Another may be inexpensive to start and rationally expensive to leave. The worksheet makes that trade visible; it does not decide it for you.
Government guidance on technical lock-in recommends identifying switching costs early and recognizes that a less portable technology may still create enough value to justify the trade. The goal is an informed decision, not portability at any price.
Acceptance rule
Accept the replacement estimate when every material row has an owner, source, quantity, rate, evidence target, and low/base/high assumption. Revisit it when architecture, data volume, critical integrations, or account ownership changes.
This is a commercial worksheet, not legal advice or a price benchmark. Have qualified advisers review material ownership, termination, data, licensing, confidentiality, and transition terms.