A subscription commerce product comparison should be built on evidence, not sales decks

What to take away

Name the decision the comparison must settle, such as renew, replace or add a second supplier, before you open any vendor material.
Weight your criteria first, then score every product against the same list, so the weighting cannot be reverse-engineered to fit a favourite.
Separate demonstrated capability from marketing claims by asking for evidence in writing, such as a sandbox or a reference call.
Treat pricing, data handling and exit terms as scored criteria, not footnotes, because they decide total cost and risk after signature.
Record who scored what and when, so a later challenge can be answered from the file rather than memory.

Why does the decision come before the shortlist?

A comparison with no decision attached drifts into feature collecting. Write one sentence naming what changes when the work finishes, for example replacing a billing provider before a renewal peak or adding a second tool for a new market.

Decision First, Then Shortlist

Write one sentence: what changes?
Set shortlist size
Define evidence needed
Identify sign-off people

That sentence sets the shortlist size, the evidence you need and the people who must sign. For a wider view of how reviews and comparisons fit together, read subscription commerce: reviews and comparison methods before you book demos.

Which criteria should carry weight?

Build the list from your own operations, not from vendor categories. Typical groups cover billing and dunning, plan changes, tax and invoice handling, integrations and contract terms.

Weight each group out of 100 before scoring. If data handling matters, say so now: the ICO sets out controllers and processors guidance that clarifies which party carries which duty in a subscription data supply chain.

How do you compare claims fairly?

Give every product the same tasks and the same time. Ask each vendor to run a defined scenario, such as pausing a plan mid-cycle or issuing a partial refund, and watch the result rather than the slide.

Score only what you saw. Where a claim cannot be shown live, mark it unverified and request written confirmation. This habit is the core of subscription commerce review methodology, which sets out how to keep scoring consistent across products.

Where does marketing evidence fit?

Retention and cancellation messaging sits inside the comparison because wording affects churn and complaints. The CIPR's professional standards and ethics guidance is a reasonable benchmark for judging whether a supplier's messaging templates meet UK expectations.

If paid acquisition is part of the plan, size the channel before you commit. IAB UK publishes UK digital adspend and channel research that helps you test whether a vendor's growth assumptions match the market you actually sell into.

What does a worked comparison look like?

Suppose a team pays £400 a month for a billing tool and £150 a month for a separate dunning add-on. A replacement quotes £520 a month but includes dunning.

On sticker price the replacement looks dearer. Once the £150 add-on and, say, eight hours of monthly admin at £35 an hour are counted, the incumbent costs £830 a month against £520. Over a year that gap is £3,720. These are illustrative figures.

Put that arithmetic in the scoring sheet, not a separate note.

How should the file be stored?

Keep one master sheet with criteria, weights, raw scores, evidence links and the date of each entry. Add a short note for every score below the midpoint explaining why.

Before the final meeting, re-check the three criteria most likely to change after signature: price escalation, data portability and notice periods.

Common questions

How many products should a comparison cover?

Three to five is usually enough to expose real differences without exhausting the team.

Should we score price or total cost?

Score total cost of ownership across at least two years, including add-ONS, integration work and admin time.

Who should sit on the scoring panel?

Include one operator who uses the tool daily and one person who owns the contract. A finance or procurement voice keeps the evidence standard honest.

When should we stop and choose?

Stop when every criterion has evidence and the top two are within a few points. Further demos rarely move the ranking; a reference call usually settles it.