Criteria written before the shortlist exists.

The evaluation confirmed
a decision already made.

Requirements and weightings agreed with every stakeholder before vendors are approached, so the process cannot be steered afterwards.

The same questions, the same scenarios and the same scoring for every candidate, run against your real use cases rather than a scripted demo.

Licence, implementation, integration, internal effort, support and contractual increases, in a model finance can re-run with its own assumptions.

Calls we arrange and run, including former customers, asking what went wrong rather than what went well.

What leaving costs in data, integration and engineering weeks. Usually the strongest lever you have in the negotiation.

A ranked matrix with the reasoning written down, so nobody relitigates the decision when the champion moves on.

CRM, ERP, commerce, data platform — anything with a three-year contract and a migration cost at the end of it.

Where the honest comparison is not two vendors but a vendor against your own engineering capacity.

A renewal approaching with no credible alternative assessed, which removes all negotiating leverage.

Two teams want different platforms and the argument has stopped being about requirements.

Several vendors with three-year terms and no consistent framework for comparing what they actually do.

Two overlapping systems and a decision required about which survives before integration cost compounds.

Every contract read,not just the pricing page.

Clauses that cost money later

Master agreements and schedules read in full for escalators, auto-renewal, data rights and exit terms. These are the clauses that decide the real price and they are rarely on the pricing page.

Claims against documentation

Vendor claims compared against their own technical documentation and API reference, which routinely surfaces gaps between what is sold and what is supported.

What customers say unprompted

Public reviews and community discussion read at volume to find recurring complaints — the ones no reference call the vendor arranges will mention.

Reference calls and negotiation are done by people.

What a customer actually thinks emerges from a conversation without the vendor present, and weighting trade-offs is a judgement about your organisation. Faster reading is what buys time for both.

When a formal selection
is not worth it

A monthly tool with an easy export does not justify a six-week process. Trial two and pick one.

A selection commissioned to legitimise a choice wastes weeks and damages trust in the process afterwards.

In some categories the market has a single viable answer. We will confirm that quickly rather than manufacture a comparison.

Selecting a platform for a process nobody has defined means scoring vendors against a guess.

Actual three-year spend versus what we projected, including the implementation and internal effort nobody usually counts.

Requirements the platform turned out not to meet. Never zero, but they should be minor rather than structural.

The difference between opening quote and final terms. Exit cost and a credible alternative are what create that leverage.

Whether the choice gets relitigated when the champion leaves. Recorded reasoning is what prevents it.

Whether the platform is actually used as intended. A technically correct choice nobody adopts is still a failure.

How long from signature to the platform doing the job it was bought for, against what the vendor promised.

What selection
sits between