How to run a vendor selection process that holds up
Most technology vendor selections go wrong at the brief stage, not the evaluation stage. By the time you are scoring responses, the damage is usually already done. The requirements were too vague, the evaluation criteria were not agreed in advance, or the wrong people were in the room when the shortlist was drawn up. This guide covers the steps that matter most, in the order they matter.
Start with a requirements specification, not a market scan
The instinct when starting a vendor selection is to look at what is available. That instinct is understandable but tends to produce a brief shaped around what vendors offer rather than what you actually need. The requirements specification should be written before you speak to a single vendor, and it should be written by the people who will use the system, not just the people who will procure it.
A useful requirements specification distinguishes between things the system must do, things it should do, and things that would be useful but are not essential. That distinction matters when you are evaluating responses, because vendors will present every feature as essential. Having agreed priorities in writing before the process starts gives you something to hold onto.
Write an RFP that asks the questions vendors find uncomfortable
A well-constructed request for proposal does more than describe your requirements. It asks vendors to demonstrate how they have handled situations that did not go to plan. Ask for references from clients whose projects ran over budget or over time, not just their showcase implementations. Ask what their escalation process looks like when a delivery milestone is missed. Ask who specifically will be working on your account, not just who will be presenting to you.
Ask about their subcontracting arrangements. A significant proportion of technology delivery is subcontracted, and the organisation you sign a contract with is not always the organisation doing the work. If subcontracting is involved, you are entitled to know who the subcontractors are and what oversight the prime vendor exercises over them.
Score responses before you read them
Agree the scoring criteria and their weightings before the responses arrive. This sounds obvious but is frequently skipped, with the result that the evaluation is shaped by whichever response the evaluators read first. A simple weighted scorecard. Covering functional fit, commercial terms, implementation approach, references, and financial stability. Applied consistently across all responses will produce a more defensible outcome than a consensus discussion.
The financial stability check is worth doing separately and early. A vendor who cannot provide audited accounts or who has changed ownership twice in three years is a risk that no amount of functional fit will offset.
What to do when the preferred vendor is not the obvious choice
Occasionally the scoring process produces a result that surprises the people who commissioned it. The vendor with the most polished presentation did not score highest on the criteria that matter most. This is the point at which the process is most likely to be undermined by internal politics.
Document the scoring outcome clearly and present it with the methodology visible. If there are legitimate reasons to override the scoring. A relationship, a strategic consideration, a risk that the scorecard did not capture. Those reasons should be stated explicitly and recorded. An undocumented override is the kind of decision that causes problems two years later when the contract is in dispute.
Vendor selection is one of the areas where independent support pays for itself most clearly. If you are planning a procurement in the next six months and would like to talk through the approach, get in touch directly.