Normalise before you compare
The first task is not scoring. It is making the proposals comparable. Suppliers will have responded to the same brief with different assumptions, so begin by listing the assumptions each supplier made, particularly about volume, review rounds, language coverage, and what the price includes.
Where a proposal is materially narrower than your requirement, either accept the difference as a genuine choice, or ask the supplier to requote on the fuller scope. Do not silently reconcile the difference in a spreadsheet, because that hides the very trade-off you are trying to see.
A useful discipline is to give every proposer the same written set of assumptions, including the review rounds, language coverage, and access to subject matter experts, so that differences in the responses reflect differences in approach rather than differences in what each supplier quietly assumed. Where a proposal rests on an assumption you did not state, write it into the clarification request and get an answer before scoring.
Build a comparison table with one row per requirement or deliverable and one column per supplier, marking included, excluded, assumed, or quoted separately. Most decision errors in content procurement are discovered at this stage, and they are almost always scope errors rather than quality errors.
Score against criteria fixed in advance
Use the criteria and weights you published in your RFP, and score each supplier against them independently before discussing as a group. A pre-scoring step protects the process from the anchoring effect of the first proposal read, which is the most common way a good decision gets distorted.
Score to a simple scale with written justification for every score. The justification is what makes the decision reviewable by people who were not in the room, and it prevents a comfortable-sounding narrative from outranking a well-evidenced one.
Have each scorer work independently and then compare. Large divergence between two reviewers on a criterion usually means the criterion was ambiguous, which is worth fixing before the scores are used, not after.
Compare cost drivers, not totals
When a proposal is lower, ask what it excludes. A lower total can reflect fewer review rounds, a lighter media method, no accessibility conformance testing, translation outside the quoted languages, or a rate structure that charges extra for changes. None of these is necessarily wrong, but each is a decision you are making.
Build a comparison of cost drivers across suppliers, and then build your own version of the scope using a consistent rate structure where possible. If your rate structure is different, model the impact: for example, a per-hour model will behave very differently from a per-module model as scope changes, and the difference becomes material at exactly the point you most need to know it.
Our guide on what affects the cost of custom eLearning development explains the variables behind these numbers, and our guide on building an enterprise content production budget shows how to hold a position against a target.
Treat the demonstration as part of the evaluation
Written proposals describe capability; demonstrations reveal it. Give every finalist the same short scenario and the same amount of preparation time, then ask them to present their approach. A common failure mode is a spectacular demo of a piece they have already built, which tells you nothing about how they would approach your problem.
Score the demonstration on the same criteria as the written response, and include a question that requires a genuine trade-off discussion. Ask the team what they would cut if the budget were reduced by a defined percentage. How they handle scarcity tells you more about the partnership than how they handle abundance.
Also use the demo to check team chemistry with your stakeholders, and to test whether the people presenting are the people who would deliver.
Do the reference and risk checks before deciding
Call references with a specific question set rather than a general one. Ask what the scope was, what the budget and timeline turned out to be compared with the original plan, what the supplier did when something went wrong, and whether the client would hire them again for the same scale of work.
Run the risk checks in parallel with the scoring, not after it. These include business continuity arrangements, security and data handling, and any dependency the proposal creates on a single tool, platform, or individual. A strong proposal with an unresolved dependency risk is a different decision from a strong proposal with no such exposure.
Finally, check contract alignment. If the proposal promises things the contract does not cover, the contract governs. Read the two side by side, as our guide on questions to ask a courseware development vendor outlines, and convert every material proposal commitment into a contractual term.
Finally, record the decision rationale in a short written note, including what you chose to accept, what you conceded, and what you will monitor after launch. The note is valuable when the relationship continues, because it turns the selection into an explicit set of expectations rather than an unspoken judgement, and it gives a future buyer a defensible record of why the choice was made.
Commonly overlooked pitfalls
- Comparing proposals at the level of totals without first checking the assumptions each one made.
- Scoring collaboratively in a single meeting, where the loudest early opinion anchors the rest.
- Treating a low price as a strength without identifying which scope elements were reduced to reach it.
- Letting the strongest written proposal win when the demonstration showed a weak fit for the team.
- Deferring reference and risk checks until after a preferred supplier has been informally chosen.
Next steps
- Build a normalised comparison table with a row per requirement and a column per supplier.
- Have two or three reviewers score independently against pre-agreed criteria, with written justification.
- Model each proposal on a common cost-driver basis so trade-offs are visible rather than implied.
- Run identical, time-boxed demonstrations and score them with the same criteria as the written responses.
- Convert every material commitment in the winning proposal into a contractual term before signature.
Related guides
Put this into practice
The vendor evaluation checklist and the RFP template turn a guide into a process you can run this week.