A weighted vendor scorecard with criteria, suggested weights, the evidence to score against and scoring guidance, designed to resist the drift that makes scorecards ratify decisions already made.
Scorecards fail in undramatic ways: weights get tuned until the preferred vendor wins, and scores get filled in from demo impressions rather than evidence. This template builds the countermeasures into the table itself. It is free to copy and adapt, with nothing to fill in first.
Last reviewed 3 July 2026 · Free and ungated
A confidential client brief puts your scorecard and shortlist in front of selected senior operators with direct experience of the vendors on it.
Pressure-test your shortlistFirst, weights are agreed and frozen before any vendor is scored. Weights adjusted afterwards are a decision wearing a methodology. Second, no criterion may be scored without citing evidence: a reference call, a contract clause, a tested scenario. A score with no evidence column entry is an opinion, and opinions were what the scorecard was meant to replace.
| Criterion | Weight | Evidence to score against | Scoring guidance (1–5) |
|---|---|---|---|
| Functional fit against your top ten scenarios | 20 | Scripted demonstrations of your scenarios on your data, not the vendor's standard demo | 5 = all ten work today unmodified; 3 = most work with configuration; 1 = key scenarios need roadmap or customisation |
| Delivery track record at your scale | 15 | Reference calls you sourced yourself, with organisations of similar size and complexity | 5 = three comparable references verified independently; 3 = vendor-supplied references only; 1 = no comparable deployment found |
| Implementation approach and dependency | 10 | The named delivery team, knowledge-transfer plan and what you can run without them at month eighteen | 5 = credible named team with a dated handover plan; 3 = approach clear but resourcing unconfirmed; 1 = delivery model depends on unnamed partners |
| Total cost over the contract life | 15 | Five-year cost model including licences, implementation, support, change requests and internal effort | 5 = full model validated by finance; 3 = major elements priced, internal effort estimated; 1 = year-one pricing only |
| Commercial flexibility | 10 | Contract terms on downscaling, pausing, benchmarking and price caps at renewal | 5 = terms allow volume reduction and cap renewal uplift; 3 = some flexibility, untested; 1 = terms ratchet up only |
| Support and escalation reality | 10 | Resolution-time data, your own test tickets during evaluation, and the named escalation path | 5 = resolution commitments with credits, verified by references; 3 = response-time SLAs only; 1 = support model untested |
| Roadmap credibility | 5 | Delivery record against the vendor's previous two years of public roadmap | 5 = roadmap history checks out; 3 = mixed record; 1 = your requirements depend on undelivered roadmap |
| Exit and data portability | 10 | A rehearsed answer to "what does leaving look like": formats, costs, timelines, in the contract | 5 = contractual exit assistance with defined formats; 3 = export possible, cost unclear; 1 = exit requires vendor professional services at their rates |
| Working relationship signals | 5 | How the vendor behaved during evaluation: responsiveness, candour about weaknesses, redline behaviour | 5 = admitted limitations unprompted; 3 = polished and evasive in equal measure; 1 = every hard question went to follow-up and stayed there |
Convene the scoring group before RFP responses arrive: procurement, the operational owner, IT, finance and one person with no stake in the outcome who acts as evidence challenger. Each scorer completes the card independently before any group discussion, because group-first scoring converges on the loudest voice. Where individual scores diverge by two points or more, the divergence is discussed evidence-first: the question is never "why did you score it low" but "what did you see that I did not".
A finished scorecard answers "which vendor scored best against what we asked". It cannot answer whether you asked the right questions, weighted them sanely, or whether operators who run these platforms daily would recognise your winner. Before the recommendation is written up, that outside reading, from senior operators who have lived with these vendors past the honeymoon, is worth more than another internal scoring round.
Show them the criteria and evidence standards at RFP stage. It materially improves response quality. Never show weights or interim scores; both become negotiation instruments in the vendor's hands.
Treat the feeling as data, not as a verdict. It usually means a real criterion is missing or a weight is dishonest. Fix the instrument openly and rescore. Overriding the result behind closed doors teaches the organisation that the scorecard was never the real decision.
This scorecard ranks vendors comparatively; the risk matrix interrogates what could fail with the one you choose. Sequence them: scorecard to select, risk matrix on the winner before signature. The templates share a philosophy but do different jobs.