Reviews
Part of How UK buyers can assess prospecting software reviews
Prospecting software selection checklist: nine checks from data provenance to references
A nine-check prospecting software selection checklist for England: data provenance, third-party consent, TPS and MPS screening, scoring and references.
What to take away
- Most teams start with a demo and work backwards, so the shortlist reflects who presented best, not what the data lets you do.
- Start with data provenance, then compliance, then pricing, and only then the product tour.
- Every claim about coverage, match rates or accuracy needs a named source or a test you can repeat in-house.
- Write the scoring model before the first call, so scores are not adjusted once a preferred supplier emerges.
What are the nine checks?
Work through the nine checks in order. Each one has a pass or fail test that you can write down before any demo.
- Data provenancesource, age and lawful basis for each contact field.
- Coverage fitdoes it hold the sectors and job titles you sell to?
- Accuracy evidencea repeatable test on a sample of 200 records.
- Compliance featuressuppression lists, consent flags and audit trails.
- Integration depthwhat the CRM sync writes, not just that it exists.
- Pricing mechanicsseat, credit and overage costs at your real volume.
- Contract termsexit, data export and notice periods in plain English.
- Support modelwho answers when enrichment breaks mid-campaign.
- Reference checkstwo customers of similar size, contacted directly.
How do you check data provenance and consent?
Ask each vendor where contact records originate, how old they are and what consent or lawful basis sits behind them. A large database with unclear origins is a liability. See data sources for the privacy basics.
Consent can arrive through a third party. People can consent to your marketing via a third party, but you must make sure the consent was valid, as the direct marketing guidance sets out. Relying on third-party consent given more than six months earlier is generally not recommended.
What screening should the tool support?
Whatever the tool promises, your team sends the messages. Live marketing calls need numbers checked against the TPS and CTPS registers and against your own do not call list. Postal marketing should be checked against the Mailing Preference Service before it goes out.
Ask each vendor how those checks work inside the tool, and how suppression and unsubscribe requests reach it. Request the answer in writing. The security claims guide shows how to test what a vendor says about its controls.
How do you test the workflow, not the tour?
Give each vendor the same brief: import 500 records, enrich 100, push them to the CRM and list what failed. Ask for the failure log. A vendor that cannot produce one has told you something useful. Check what the CRM integration writes to each field.
For accuracy, check a sample of 200 records by hand. As illustrative arithmetic, if 170 of 200 sampled records are correct, accuracy is 85 per cent. Run the same sample against every vendor so that the numbers are comparable.
How do you score before you demo?
Write the scoring weights before any demonstration. If data quality is the main risk, weight it above interface polish. A simple model, for example 40 points for data and compliance, 30 for workflow fit and 30 for commercial terms, keeps the conversation on evidence.
The reviews and comparison methods guide explains how to hold weights steady across vendors whose demos differ in style. The review methodology article covers how to grade evidence when vendors supply their own numbers.
What belongs in the contract and reference checks?
Price a quiet month and a busy month. As illustrative arithmetic, a team paying £400 a month for 5,000 credits pays £4,800 a year before seat fees and top-ups. Ask for both scenarios in writing, and read the commercial contracts checklist for exit and notice terms.
An organisation is not by its nature either a controller or a processor, as the controller or processor test sets out, so the contract should record the role for each activity. Then call two reference customers directly.
Common questions
How long should the checklist take?
Two to three weeks for a shortlist of four or five vendors, because the trial and reference steps take longest. Block the calendar for the trial first, then fit the demos around it, so the evidence arrives before the opinions do.
Should legal review it?
Yes, if you process personal data at scale. A quick review of the provenance, consent and suppression questions prevents the most expensive mistakes, and it gives you a record of who approved which answer.
What if a vendor refuses the trial?
Treat it as a scoring failure rather than a negotiation point. A supplier confident in its data will run the test, and a refusal tells you how it is likely to behave once the contract is signed.
