Most software evaluations are a decision already made, dressed in a scoring matrix afterwards. The two things that prevent that are locking weights before you see proposals, and requiring evidence rather than impressions in each cell.
Lock the weights before you see proposals. Agree and record the weights while nobody knows which vendor would benefit. Weights set afterwards get tuned, consciously or not, toward the preferred option.
Define what each score means. Without a rubric, one evaluator's 3 is another's 5. Write what 1 through 5 mean for each criterion, usually in terms of whether the need is met natively, via configuration, via customisation, or not at all.
Score independently, then meet. Have three to five evaluators score alone before discussing. Scoring as a group means the most senior or most vocal person's view propagates, and the matrix just records it.
Here's a preview. Unlock the free download to get all 3 sections (2 more below).
Complete and sign off before any proposal is opened. Weights must total 100.
| Criterion | Weight % | What a 5 looks like | What a 1 looks like | Who evaluates |
|---|---|---|---|---|
| Functional fit: must-haves | 30 | All must-haves native, no workarounds | Multiple must-haves unmet | Process owner |
| Total cost over 3 years | 20 | Lowest TCO including implementation and support | Highest TCO or unclear pricing | Finance |
More free templates IT teams use alongside this one.
Licence price is the number vendors compete on and the smallest part of what you will spend. This worksheet forces in the costs that decide which option is actually cheaper: implementation, internal time, integration, and getting out again.
A POC without written success criteria becomes an extended demo that everyone enjoys and nobody can draw a conclusion from. Deciding in advance what would make this a failure is what turns a trial into evidence.
RFPs produce incomparable responses because each vendor answers in their own structure and prices in their own format. Fixing the requirement format and the pricing table is what makes a side-by-side comparison possible at all.
Functional fit usually carries the most weight at around 25-35%, with cost at 15-25% and security, integration, implementation effort, and vendor viability sharing the rest. What matters more than the exact split is agreeing it before any proposal is opened.
Because group scoring propagates the most senior or most vocal evaluator's view. Independent scoring first surfaces genuine disagreement, and the disagreements are usually where the real decision information is.
A specific observation with a date: which demo, which document, which reference call. 'Looked good' is not evidence. The column exists so the decision can be defended months later when someone asks why this vendor was chosen.
