Reviews
Part of Why keyword research reviews need a method before you compare tools
Why keyword research best providers need a method before you compare
How to assess keyword research providers: inclusion criteria, a before-and-after table, and due diligence on ad copy, data sources and technology standards.
What to take away
- Set inclusion criteria and scoring weights before any pitch, then score every provider on the same sheet.
- Verify claims with raw query lists, export dates and a fixed-fee pilot, not with demo dashboards.
- Comparative claims sit under the Business Protection from Misleading Marketing Regulations 2008 and the CAP Code, which apply across the UK.
- Test local coverage with named England sources such as ONS business statistics and Google Ads regional data.
- Record each pitch claim beside the verified result so finance or legal can audit the choice.
Criteria and weights to fix before you compare providers
Write the shortlist sheet before you open a single demo. Each provider entry needs four items: a named method, a data provenance statement, a sample deliverable for a topic like yours, and a named contact for procurement and disputes.
Inclusion criteria before comparing
- Written method required
- Named data source required
- Sample output for your topic
- Small paid pilot accepted
- Regional spelling variants shown
- Market advice written for
- Named method covering sampling, de-duplication and branded query filters.
- Data provenance note naming the source, the licence and the export date.
- One sample deliverable on a topic close to your own.
- A named contact and a written dispute route.
Test the sheet against documented products first. Google Keyword Planner, Ahrefs, Semrush, Moz and SE Ranking publish method notes you can read before any meeting.
Check sector wording against the ONS business, industry and trade hub, which uses official classifications rather than marketing shorthand.
Then add two England tests. Ask which regional vocabulary the provider uses, because loose wording such as the Northwest is not the ONS region name. Ask whether volumes split by Google Ads regional data or arrive as UK-wide totals only.
The method for locking weightings before suppliers are scored is set out in keyword research reviews and comparison methods for 2027.
Example: one criterion, one weight
Illustrative weights only, chosen by your own team:
One criterion, one weight
| Criterion | Weight | Score (0 to 5) | Weighted points |
|---|---|---|---|
| Data provenance statement | 25% | 4 | 20 |
| Sampling and de-duplication method | 20% | 3 | 12 |
| Sample deliverable for your topic | 20% | 5 | 20 |
| England regional handling | 15% | 2 | 6 |
| Pilot terms and dispute route | 20% | 4 | 16 |
| Total | 100% | 74 |
Multiply each score by its weight to get points. This provider reaches 74 out of 100, and a rival with a polished demo can still lose on provenance.
Verify claims in a before and after record
Verify claims in a record
| Claim at pitch | Evidence that counts |
|---|---|
| Proprietary search data | Raw query list with source and export date |
| Strong sector rankings | Position data with URLs, dates and an England location setting |
| Nationwide England coverage | Volumes split by ONS region |
| Delivery in three weeks | Pilot milestones agreed in writing before payment |
Claims vs verified evidence
Before the pitch (claimed)
- Coverage
- Full UK coverage
- Data
- Proprietary data
- Compliance
- Compliance assured
- Speed
- Fast turnaround
- Team
- Senior team
After the pilot (verified)
- Coverage
- English regions and variants
- Data
- Named source, refresh, audit
- Compliance
- Checked against CAP guidance
- Speed
- Actual pilot delivery dates
- Team
- Who wrote it and role
Then run the pilot in this order:
Verify claims in a record
- Send one written brief to every provider.
- Request a raw query list for one topic cluster, with export dates.
- Agree a fixed fee, a deadline and one named reviewer.
- Log each pitch claim beside the verified result.
- Keep the evidence pack in one file that finance can read.
The documents to request before signing are listed in our keyword research selection checklist in England, which covers provenance notes and the contact for disputes.
Which rules govern comparative claims in the UK
The Business Protection from Misleading Marketing Regulations 2008 govern misleading comparative marketing in the UK. The CAP Code adds the rules the ASA enforces, and the ASA advertising codes set out the detail.
For non-broadcast copy, the ASA Copy Advice service gives free, confidential checks before publication. Use it before you publish a comparison that names a rival.
US Federal Trade Commission penalty guidance applies in the United States only, so it is not your benchmark in England. Keep a dated record of published copy, because the CAP Code requires evidence for comparative claims.
Pilot scope, effort and cost
Set your own ceiling before you ask for quotes, then request a fixed fee so each quote compares like for like. Ask for one fee covering the cluster, not an hourly estimate.
Planning assumption, set by you and not a market rate: one topic cluster, roughly 20 to 40 seed terms, two to three weeks, and one reviewer at about half a day a week.
Our keyword research budget template in England shows how to line up the brief, the fee and the reviewer hours in one sheet.
Common questions
What if a provider refuses a paid pilot?
Treat refusal as a material risk. A small fixed-fee pilot is the cheapest way to test delivery dates, provenance and written quality.
How do I score a provider that will not share raw data?
Score provenance at zero and note why. Without a query list the claim cannot be tested, and an untestable claim should carry no weight.
Who signs off on the comparison?
Finance or procurement usually owns the final call, so keep the pack auditable. A one-page summary of criteria and verified results beats a slide deck.
Do these checks apply outside England?
Most do. Method and provenance checks travel well. Advertising rules differ by market, so check the local regulator before publishing comparative copy.
