How this situation actually presents itself
Procurement circulates the evaluation pack on a Friday. Three vendors, weighted criteria, demo scores, a recommendation due at the steering committee in a fortnight. The business sponsor has already had two informal dinners with one of the three. The technical lead prefers a different one because their platform team has used it before. The third vendor is on the list largely to make it a competition. Everyone behaves as though the matrix will decide; nobody quite believes it.
Where the difficulty really sits
This is not an analysis problem. The information that matters most is how each vendor behaves in month nine, after the sales team hands over to delivery, and no demo, RFP response or curated reference call will surface it. Meanwhile the incentives in the room pull in different directions: procurement is measured on savings and process integrity, the sponsor on speed to benefit, the technical team on avoiding an integration they will have to live with. A weighted scoring model averages those interests; it does not reconcile them. And the vendor knows more about its own delivery record than any buyer can discover through the formal process.
What tends to go wrong after award
- The named delivery team from the bid is reassigned once the contract is signed, and the replacement team learns the account on the buyer's time.
- Assumptions buried in the statement of work (data quality, client-side effort, environment readiness) convert into change requests within the first quarter.
- The discount that won the deal is recovered through rate cards on change requests and out-of-scope work.
- The runner-up vendor's stronger point, the one scored down in the matrix, turns out to be the thing that matters in production.
- Nobody re-reads the exit and data-portability clauses until the relationship is already strained.
Questions the evaluation committee should table
- If the favourite did not exist, which of the other two would we choose, and what does that comparison reveal?
- Which criteria were weighted after the preferences formed, rather than before?
- Can each vendor name the individuals who will be on this account in month six, contractually?
- What happened on each vendor's last comparable engagement that did not go well, and how did they behave?
- What would make us walk away from all three?
Pressure-test these before the recommendation goes up
Test the requirements, not just the responses. If the specification was shaped around one vendor's strengths, the evaluation is already answered. Test the commercial construct under stress: what does year three cost if volumes double, if scope shifts, if the relationship sours. Test references the vendor did not supply. And test the scoring itself by asking a colleague outside the project to score blind; a large gap between their ranking and the committee's usually means the matrix is decorating a decision rather than making it.
Who belongs in this conversation before award
Service owners, integration architects and the contract manager who inherits the relationship will operate the outcome, and they deserve a formal voice, not a courtesy briefing. Legal and commercial teams should review the exit provisions while leverage still exists. The gap that internal teams and even procurement advisers struggle to close is lived delivery experience with these specific vendors: executives who have been the client on the other side of each name on the list. That is where independent perspectives from selected senior operators change the decision. They have seen how each vendor behaves when the programme is in trouble, which is the one scenario the evaluation process never simulates.
