Reviews
Part of How to weigh evidence in proposal software reviews before buying
Seven checks behind a proposal software review methodology
A seven-step proposal software review methodology for England that separates vendor claims from verifiable evidence and records what was not assessed.
What to take away
- The most common mistake is treating vendor material as evidence. A methodology separates what a supplier says from what an independent source confirms.
- Every score needs a written basis, a named source and a date, so a colleague can reproduce it later.
- Certifications are checkpoints, not endorsements. They confirm a control exists, not that the product suits your sales cycle.
- State plainly what the review did not cover, including anything needing a live pilot.
- A published method protects the buying team when a preferred vendor loses on score.
Start by writing down what the review is not
Most teams begin with a feature grid and a shortlist. That order produces a comparison shaped by whichever supplier sent the best demo.
Write the exclusions first. A desk review cannot test performance under load or confirm how a quoting engine handles your pricing rules. Those need a pilot.
Once the limits are agreed, scoring stops drifting. Reviewers assess evidence quality, not personal preference.
Separate three kinds of input
Vendor claims are marketing, documentation and sales answers. They frame questions and never carry a score on their own.
Independent evidence comes from regulators, standards bodies and published analyst material. Editorial judgement is the reviewer's reasoning applied to both, recorded in a sentence or two.
Keep the three in separate columns. Where a claim and an independent source disagree, the source wins and the conflict is noted.
Seven steps to a repeatable review
- Define the buyer scenario in one paragraph, including team size, deal value band and the systems a proposal tool must connect to.
- List the criteria before contacting any supplier, so the shortlist cannot reshape the questions.
- Gather vendor claims and label each with its source and date.
- Check security and staff-vetting positions against published schemes, not assurances given on a call.
- Score each criterion from 1 to 5, with a written reason for anything at or below 3.
- Record the counter-argument for the highest-scoring option, the cheapest way to expose a weak winner.
- Date the review and set a review-by date, because pricing and ownership change.
Step four is where desk research earns its place. The Cyber Essentials scheme overview sets out the certification a supplier can hold, a reasonable baseline for a tool handling commercial pricing. For providers handling sensitive bids, the BS 7858 staff vetting standard covers personnel screening rather than technical controls alone.
Seven steps to a repeatable review
- Define buyer scenario in one paragraph
- List criteria before contacting suppliers
- Gather vendor claims with source and date
- Check security against published schemes
- Score each criterion 1 to 5
- Record counter-argument for top option
- Date review and set review-by date
Weight the criteria before you score
Equal weighting flatters broad, shallow products. Weighting exposes the trade-offs a buying team faces.
For example, a team sending 40 proposals a month might weight integration above analytics. A team sending 400 might reverse that. Publish the weights with the scores so the reasoning survives staff changes.
If criteria keep shifting between reviews, the framework is often the problem, which is why the Proposal software: reviews and comparison methods for 2027 pillar sets out the wider structure this method sits inside.
Handle market position claims carefully
Supplier pages often cite analyst recognition as proof of suitability. That is a claim about market position, not about fit for your pricing model.
Analyst reports validate the market position of leading sales software vendors. Pages such as Salesforce's published analyst recognitions confirm what a vendor claims about itself, not product quality.
Treat every recognition as one line of evidence. Weight it low unless the buyer scenario depends on ecosystem scale.
Common questions
What was assessed in this methodology?
Published vendor material, independent sources and reviewer reasoning applied to a defined buyer scenario. Each score carries a written basis so another reviewer can reach a comparable result.
What was not assessed?
No hands-on testing, no live pilot and no load testing. Those need a working environment and a real pricing dataset, so they belong in a trial stage after shortlisting.
How are vendor claims separated from evidence?
Claims are logged with source and date, then scored only where an independent source or documented reasoning supports them. Conflicts are recorded rather than averaged away.
How often should a review be repeated?
At least annually, and whenever pricing, ownership or certification changes. A dated review tells the next reader how stale the evidence is. When you move from method to shortlist, the Proposal software product comparison in England article covers the comparison stage.
