Five roles to separate before researching who reads an announcement

Audience research for an announcement service should answer one decision: which route can meet a defined communication job, for whom, under what evidence and permission constraints? It should not invent an ideal customer or treat a journalist as a sales prospect.

This guide designs research for an organisation operating in England that must publish an approved factual announcement and decide among an owned newsroom, direct outreach, technical transmission, paid placement, monitoring and the existing manual route. Earned editorial coverage remains a publisher's independent decision. The design contains no interview, survey or audience finding.

What to take away

Identify five roles before asking questions

The word audience hides different populations. Give each person or organisation one or more explicit roles and record where duties overlap.

Service buyer: the legal entity considering a purchase. The buyer owns the business decision, budget and contract, but may not approve the words.

Internal approver: the person authorised to clear facts, claims, rights, timing and sector issues. An approver can reject publication without being the buyer.

Journalist, editor or media contact: a person or desk making an editorial decision, or receiving material for that purpose. A public email address does not turn its holder into a customer, subscriber, end reader or permitted marketing recipient.

Technical destination: a system, feed, endpoint or buyer-owned page that may accept a file. It is not a human audience, and acceptance does not prove attention.

Public or stakeholder audience: the people intended to find, understand or act on information. Their needs cannot be inferred from a destination list.

This separation prevents a common design error: interviewing communications buyers, then presenting their preferences as evidence about journalists or the public. Each role needs its own population, sampling frame and question set.

Freeze one announcement job

Write the research record before recruitment:

The ONS UK Business: activity, size and location 2025 release distinguishes enterprises and local units and supplies regional tables. Those statistical units may help frame organisations. They do not identify service buyers, approvers or media contacts, and they cannot establish demand.

Build a population and recruitment ledger

Create a separate frame for each role. Define geography by a relevant operating connection to England rather than a convenient registered address. Record the source, extraction date, inclusion rule, exclusion reason, invitation route and non-response.

Recruit buyers from a disclosed organisational frame, not from people who happen to answer a marketing list. Recruit internal approvers through the participating organisation's authorised contact. For media professionals, establish the research purpose and contact provenance before any invitation. Do not reuse research participation as permission for later sales contact or onward sharing.

The ICO's business-to-business marketing guidance explains that publicly available personal data remains subject to UK GDPR and that PECR rules vary with channel and recipient type. The page is marked under review following the Data (Use and Access) Act. A named privacy and PECR reviewer should therefore assess the exact invitation, source and purpose and recheck the guidance before fieldwork.

Give participants usable information

Before participation, explain the research sponsor, researcher, purpose, tasks, duration, recording or transcription, incentive, data collected, recipients, retention, withdrawal route, complaint contact and whether findings may be quoted or published. Separate permission to join the research from the lawful basis for each personal-data operation.

The ICO right-to-be-informed guidance lists purpose, lawful basis, recipients, retention, rights and data source among transparency fields. It too carries a DUAA review notice. The broader data-protection principles guide covers purpose limitation, data minimisation, accuracy, storage limitation, security and accountability.

Professional research standards add a different layer. The MRS Code of Conduct 2023 applies to MRS members and accredited company partners. It emphasises transparent purpose, participant well-being, independent judgement and reproducible methods. It does not replace UK law or certify a project run by somebody else.

Design access into the method

Ask invitees what adjustment or format they need. Offer a route that does not require an inaccessible form, rapid speech, video or a particular device. Record whether the alternative collects equivalent evidence; do not demote it as lower quality merely because its mode differs.

Government accessible communication formats guidance, updated on 17 June 2026, recommends involving disabled people, planning responsibility and providing suitable formats. It was written for government communications, though other researchers can use its practical questions. Qualified accessibility and equality review must address the actual participant population and service.

Safeguarding needs its own route when research could involve children, vulnerable adults, distressing announcements or a power imbalance. Define eligibility, responsible-adult involvement where applicable, distress indicators, withdrawal, escalation and data access before recruitment. Do not collect sensitive detail simply because a conversation permits it.

Ask about observed work, not imagined preference

Interview questions should move from a real recent task to evidence and only then to options. Useful prompts include:

Avoid asking whether someone likes a hypothetical feature or would pay an unspecified price. A journalist can describe receipt and editorial assessment; they cannot prove public readership without separate evidence. A buyer's account of audience needs remains an inference until researched with that audience.

Keep notes in four columns: what the participant said, what artefact was observed, the researcher's interpretation and what remains unknown. Never manufacture a quotation from a paraphrase.

Compare research routes on common fields

The tool decision follows the method. Compare a buyer-owned manual log, an approved existing service, a form platform, a moderated interview and an accessible alternative on the same fields:

Leave price, service levels, thresholds and scores blank until current evidence exists. No platform was tested for this guide. NCSC guidance for using SaaS securely addresses access, data location, provider access, sharing, logging, recovery and service change. It does not prove that any form product is secure for this job.

Analyse without turning stories into prevalence

Publish the recruitment and analysis flow: records approached, eligible, participating, withdrawing, incomplete and not responding. Explain coding, second-review procedure and how contradictions were retained. One detailed account may expose a process defect; it cannot show how common that defect is.

Use a decision table rather than a composite persona score. For each route, record evidence that supports it, evidence against it, unresolved gates and the owner of the next check. Stop if the sample excludes a role essential to the decision, participants misunderstood the research purpose, accessibility failures distorted participation, or missingness makes a comparison unfair.

Keep research and promotion apart

Research findings do not authorise outreach, syndication or paid placement. CAP Code section 2 requires identifiable marketing communications within its remit. If recruitment material promotes the buyer or a supplier, qualified advertising and research reviewers should examine whether the activity remains research and how commercial intent is presented.

Likewise, observing a competitor's public claim does not verify capability. Corporate filings, supplier pages, paid content and independent editorial items belong in separate evidence fields. Copying contact databases or protected material creates privacy and rights questions rather than better research.

Turn findings into a bounded decision

The decision owner should predefine what evidence can change the route. A proceed decision requires a verified buyer need, accessible research coverage across required roles, no unresolved privacy or rights gate, and a successful synthetic correction test.

A hold follows when a necessary population is absent or a tool's data and exit terms remain unknown. Stop when the research purpose drifts into undeclared marketing or participants cannot withdraw safely.

Before publication, assign the author and fact-checker. Named UK research-methods, media, privacy, PECR, accessibility, equality, safeguarding, security, IP and commercial reviewers must check their parts. Re-open every ICO page because the cited guidance is under review, and do not publish findings that have not actually been collected.

Questions for the research owner

Is a media contact part of the target audience?

Only if the study defines that role and purpose. The person is not automatically the buyer, public audience or permissible marketing recipient.

Can persona cards be written before interviews?

They can be hypothesis cards with sources, unknowns and expiry dates. They must not be presented as discovered people or findings.

Does survey consent permit later sales contact?

Research participation and later direct marketing are separate purposes. The latter needs its own assessment and cannot be hidden in an invitation.

What is the minimum useful output?

A transparent evidence map that changes or stops one route decision is more useful than a confident-looking profile built from untested assumptions.

Before you act

Common questions

Why should a journalist not be treated as a sales prospect?

A journalist, editor or media contact makes an editorial decision or receives material for that purpose. A public email address does not make its holder a customer, subscriber, end reader or permitted marketing recipient. Mixing that role with buyer research produces evidence about the wrong population.

What must be settled before recruitment starts?

The research record should state the buyer legal entity and England operating location. It should also state the announcement type, approval owner, latest useful date and intended public population. Routes under consideration, decisions the research may change and excluded questions belong there too.

How should accessibility be handled in the method?

Ask invitees what adjustment or format they need, then offer a route that avoids inaccessible forms, rapid speech, video or a single device. Record whether the alternative collects equivalent evidence. Do not treat it as lower quality just because the mode differs.