Audience research
Part of Audience research for social listening starts with the decision, not the data
Researching social listening competitors without overstepping the public record
Research social listening competitors in England with dated public evidence, like-for-like criteria and firm boundaries around testing and inference.
Social listening competitor research should record what a named organisation publicly offered on a stated date. It should not imply product testing, customer experience or inside knowledge when the work consists of reading websites, filings and advertisements.
Use this checklist to build an evidence table that another researcher can verify. Legal, privacy, intellectual-property and advertising conclusions require named qualified review.
What to take away
- Record what a named organisation publicly offered on a stated date, not implied testing or inside knowledge.
- Link every capability claim to the provider's own current page and label it first-party.
- Reserve tested for a declared protocol, date, configuration, dataset and researcher.
- The completed record can support a shortlist but not a winner or performance claim.
Define the comparison
- Write the buyer question and common unit before selecting competitors. One unit might be a twelve-month software specification, a defined research project or a monthly managed service. Do not compare these as if they contain the same work.
- Set the England connection. This might be an English registered office, delivery availability to an English buyer or evidence of an English contract. State which one is used.
- Publish inclusion and exclusion rules. Avoid a hand-picked set that makes the commissioning organisation look favourable.
Verify identity without overstating it
- Record the legal name, trading name, source URL and access date.
- Check filed identity and status where relevant. Companies House explains the fields in its public register and its limitations. A filing can support identity; it does not prove skill, solvency, service quality or market share.
- Keep UK-wide registration evidence separate from actual delivery in England.
Three England and UK sources are worth checking, and each has a limited reach.
- Companies House holds the filed recordlegal name, company number, registered office, filing status and filed accounts. It does not prove trading performance, service quality or that the company delivers in England.
- The ICO register of data protection fee payers shows whether an organisation has paid the fee and named a data protection officer where required. It does not prove that processing is lawful or that marketing consent exists.
- ASA rulings set out published adjudications under the CAP Code, upheld or not upheld. They do not prove that a product works or that a provider suits a particular buyer.
Capture public claims faithfully
- Link each capability, coverage or commercial statement to the provider's own current page.
- Save the wording and date, then label it "first-party claim". Do not rewrite "available" as "effective" or infer completeness from a long source list.
- Mark missing information as undisclosed. An absent public price is not evidence that the service is expensive or negotiable.
A single evidence-table row might look like this.
| Provider | Claim | URL | Access date | Label |
|---|---|---|---|---|
| [legal name exactly as filed] | [exact wording copied from the provider's own page] | [full address of that page] | [date you opened it] | first-party claim |
The row is complete only when another researcher can open the same address on the same date and read the same wording.
If the final material becomes promotional, the CAP Executive's comparison guidance says objective claims need evidence. Its verifiability advice also calls for checkable support where a competitor is identifiable. These pages are non-broadcast advertising guidance, not individual legal advice.
Respect source and platform boundaries
- Use the access method permitted by the website, platform and contract. Record login, geography and edition limits.
- Do not copy screenshots, reports, database extracts or substantial text merely because they are visible. GOV.UK's copyright guidance says protected material generally needs permission, ownership or an applicable exception.
- Separate organisation facts from personal profiles. Minimise names and role-holder details that are unnecessary for the comparison.
Public availability does not authorise later prospecting. The ICO warns in its lead-generation guidance that public social information is not automatically free to use for direct marketing. Keep research records out of sales systems unless a separately reviewed purpose permits the transfer.
State the testing boundary
- Label screenshots, demonstrations and provider-led tours as observed presentations, not independent tests.
- Reserve "tested" for a declared protocol, date, configuration, dataset and researcher.
- Do not report sentiment accuracy, source coverage, alert speed or service quality without results produced under that protocol.
- Record conflicts, free access, samples and supplier contact beside the affected evidence.
A minimal acceptable protocol names the question tested, the date, the product edition or configuration, the dataset or query set, the steps taken, the researcher and where the raw output is kept. It is short enough to publish beside the test.
Analyse and update
- Compare every included provider on the same material criteria and unit.
- Show unknowns and contradictory records rather than awarding assumed scores.
- Distinguish observation, interpretation and the buyer's eventual decision. GOV.UK's research-analysis guidance offers that separation for research evidence.
- Recheck when a page, product edition, contract term or comparison purpose changes.
The completed record can support a shortlist or new research question. It cannot support a winner, performance claim or statement about an organisation's internal practice without further evidence.
Before you act
- Confirm every claim in the record carries a source URL, an access date and a label.
- Delete any winner, performance or internal-practice statement the evidence cannot support.
Common questions
What should a social listening competitor research record?
It should record what a named organisation publicly offered on a stated date. It should not imply product testing, customer experience or inside knowledge when the work consists of reading websites, filings and advertisements. The record must be verifiable by another researcher.
How should public claims be captured?
Link each capability, coverage or commercial statement to the provider's own current page. Save the wording and date, then label it first-party claim. Do not rewrite available as effective or infer completeness from a long source list. Mark missing information as undisclosed.
What are the limits of the completed record?
The completed record can support a shortlist or new research question. It cannot support a winner, performance claim or statement about an organisation's internal practice without further evidence. Legal, privacy, intellectual-property and advertising conclusions require named qualified review.
