The score should explain the next action, not merely name a winner. These fictional proposals illustrate the calculation; they are unrelated to any named supplier.
Assume both scored proposals pass the gates, have complete evidence and pass A2, B1 and B2.
| Criterion | Weight | Direct plan: raw score | Direct: points | Hosted plan: raw score | Hosted: points |
|---|---|---|---|---|---|
| A. Catalogue | 25 | 4 | 20.0 | 4 | 20.0 |
| B. Reliability | 20 | 4 | 16.0 | 5 | 20.0 |
| C. Integration | 15 | 3 | 9.0 | 4 | 12.0 |
| D. Metadata | 10 | 4 | 8.0 | 4 | 8.0 |
| E. Controls | 10 | 4 | 8.0 | 4 | 8.0 |
| F. Support | 8 | 2 | 3.2 | 4 | 6.4 |
| G. Commercial | 7 | 5 | 7.0 | 4 | 5.6 |
| H. Exit | 5 | 3 | 3.0 | 4 | 4.0 |
| Total | 100 | 74.2 | 84.0 |
- Direct plan: remediate support and integration before approval. Its attractive commercial score does not cure its 2/5 support result.
- Hosted plan: eligible for approval within the tested scope, subject to budget and accountable sign-off.
- A third proposal with unresolved upstream authority: not eligible, regardless of how impressive its provisional technical score appears.
For a hybrid plan, score the combined implementation. Deduplicate overlapping games and include the buyer’s work coordinating contracts, feeds, monitoring and incidents. Do not average two supplier scores and call that a portfolio assessment.