Audience research for a parenting publication that studies the adult reader, not the child

Audience research for a parenting publication should begin with an adult's real decision, not a ready-made image of an ideal follower. Speak to people who have recently faced that decision, use public statistics to check who may be missing, and keep the information collected proportionate. The result should tell an editor what to make, change or stop.

There is no official dataset showing what every parent in England wants from creators. England family surveys answer particular policy questions. UK media studies cover different populations. Platform analytics describe current account activity. Each can inform a research plan, but none can be relabelled as an England-wide creator audience study.

This guide sets out a practical cycle for adult-directed publishing. It does not authorise research with children. Work involving under-18s, vulnerable adults, sensitive family circumstances or special-category information needs appropriate expertise, a defensible lawful basis and safeguards designed for the actual project.

What to take away

Write the editorial decision first

Research becomes vague when the team starts with "Who is our audience?" Replace that with a decision that can be made within weeks. Examples include whether to produce a council-by-council childcare checklist, whether an explanation works better as audio or text, or which part of a school-admissions journey causes readers to abandon a guide.

List what the team currently believes and convert each belief into a question. If an editor believes first-time nursery users need a comparison table, the questions concern how people compare now, where they lose confidence and which details change the choice. Asking whether participants like the proposed table will mostly measure politeness and presentation.

The Government Digital Service recommends defining objectives around problems, assumptions and the information required for the next action in its guidance on planning a round of user research. That page is for government teams, not creator businesses, yet its decision discipline is useful here.

Give every round a small written brief:

If nobody can name the decision, postpone recruitment. Gathering family stories without a defined use exposes participants and leaves the editor with material that is tempting to publish but hard to interpret.

Distinguish the adult reader from the subject child

A page can discuss babies, school pupils or teenagers while serving an adult audience. Specify whether the intended user is a parent, foster carer, grandparent, practitioner or another decision-maker. Also specify whether the content could reasonably attract children despite that intention.

Most early editorial questions can be answered by adults. Ask a carer how they checked an entitlement, assessed a product claim or set a household rule. Do not involve the child merely to create authenticity. A parent's permission is not a complete ethical method for interviewing, filming or profiling an under-18.

GDS guidance on informed consent says research with a child or vulnerable adult requires informed agreement from both the participant and an appropriate responsible adult. Any such plan also needs qualified safeguarding and privacy review. The easiest safe option is often to redesign the question around an adult's decision and use fictional test material rather than family records.

The ICO explains that the Children's code covers online services likely to be accessed by under-18s, including services not explicitly aimed at them.

A creator-run app, membership product or interactive site may need a different assessment from an adult newsletter hosted by a third party. A label in a planning document cannot settle likely access or legal scope.

Use England evidence for context, not imaginary demand

Public records help the team understand the population and form better questions. They do not replace direct investigation of the proposed publication.

The Department for Education's Childcare and early years survey of parents for 2025 covers parents of children aged 0 to 14 in England. Its achieved sample was 4,953 parents.

Its methodology and tables document childcare use, affordability, availability and family characteristics. An editor considering childcare content can use these findings to identify situations worth recruiting across. The survey did not ask whether respondents follow a parenting creator or would pay for a guide.

Ofcom's Children and parents: media use and attitudes report 2025-6 covers the UK. It included 5,008 parent and 3,426 child interviews, with a changed design from 2025.

It can prompt investigation into screen routines or confidence with online safety. It cannot yield an English follower total, and Ofcom treats comparison with earlier studies cautiously because of the methodological change.

Treat account analytics as another bounded source. A platform dashboard may show who saw existing work, not those excluded by the current format or distribution.

Search terms show expressed queries, not whether the answer was useful. Newsletter replies are rich but come from existing subscribers. Write each limitation beside its input, not in a final footnote.

Recruit for the question, not convenience

Define eligibility in behavioural terms. A participant might be an adult in England who applied for funded childcare within the last six months, uses a screen reader to compare local information, or recently tried to understand an advertising disclosure on family content. Collect a broad age band for the child only when family stage genuinely affects the task.

GDS advises recruiting actual or likely users and making deliberate provision for disabled people, limited digital skills and different social circumstances in its guide to finding participants. It also gives a typical range of four to eight participants for an interview or usability round. That range supports close qualitative learning, not statistical estimates about England.

Use several recruitment routes where possible: newsletter invitation, community organisation, relevant service, independent recruiter, and state what each omits.

A social poll disproportionately reaches platform users at the time the post is shown, while a local group offers depth in one setting. Friends of the creator may suppress criticism. None should be described as representative without an appropriate sampling design.

Make participation workable. Offer phone and remote options, ask about communication needs, avoid school-run pressure points and compensate time fairly where feasible. Do not exclude people because recording them is inconvenient. A note-only session may be safer and more accessible than demanding video.

Choose the lightest method that answers the question

Different uncertainties require different evidence.

Use an interview to understand sequence, language and context. Ask for the most recent real example, then follow what happened. GOV.UK's page on using in-depth interviews recommends a discussion guide with open starter questions and prepared follow-ups. It also suggests piloting the flow with a colleague, which checks clarity without turning the colleague's opinions into findings.

Use a short survey when answer options are already reasonably understood and the team needs structured responses from a larger, defined group. A survey distributed only through the creator's channel measures responding followers, not all parents. Report invitations, eligibility, completions, missing data and question wording. Do not create an England percentage from a voluntary online poll.

Use usability observation when the decision concerns a draft guide, calculator, sign-up journey or disclosure. Give the adult participant a realistic task and watch where the design obstructs progress. Use invented names and accounts in prototypes. A real child's school login, health letter or photograph should never be required to demonstrate a content journey.

Public competitor review can reveal formats, evidence standards and commercial routes already visible to readers. It cannot show another publisher's conversion rate, private audience or intentions. Do not scrape comments or followers to build household profiles. A joint regulator statement on data scraping and privacy makes clear that publicly accessible personal information remains within data-protection responsibilities.

Design the questions around behaviour

Ask "What did you do next?", "What made that difficult?" and "How did you decide the source was reliable?" These prompts invite an account of events. Avoid "Would you watch a weekly series?" or "Don't you find council pages confusing?" because they invite prediction or agreement.

For sensitive family topics, let participants skip questions without explanation. Do not ask for diagnosis, immigration status, financial distress, domestic circumstances or a child's online identity unless that information is essential, reviewed and lawfully handled. Editorial curiosity is not a sufficient reason.

A discussion guide should contain an opening explanation, topic sequence, neutral prompts, a distress or disclosure procedure, closing reminder and contact route. The facilitator must know which matters they cannot advise on. Research is not a substitute for medical, legal, safeguarding or benefits support.

Pilot the instrument using made-up data. Check whether questions expose a family member, whether answer options force an inaccurate category and whether an accessibility barrier changes who can respond. Revise before sending any public invitation.

Plan privacy before collecting anything

List each field and file the research could create: screeners, contact details, consent records, notes, recordings, transcripts, survey exports and research summaries. For each, record purpose, access, storage location, retention and deletion owner.

The GOV.UK service manual's participant privacy guidance recommends gathering the minimum information, separating access according to need and deleting participant details once they are no longer required. It specifically identifies data about children and health as requiring particular care. Contracts with survey, recruitment and transcription suppliers also need to address permitted use and deletion.

Do not confuse agreement to participate with the lawful basis for processing. The ICO's research-provisions guidance explains that ethical participation consent is distinct from UK GDPR consent as a lawful basis. Its research provisions may not fit ordinary commercial audience work at all. A competent adviser should document the applicable basis and any condition for special-category data before collection.

Use participant codes in working notes, keep contact information separately and restrict raw access. Removing a name does not anonymise a detailed account if a school, town, unusual event and child's age can identify the household. Public reporting should use aggregated or thoroughly anonymised material unless specific, appropriate permission and safeguarding review support another approach.

Compare survey services as processors, not shortcuts

Before selecting software, decide which controls are necessary. Common checks include respondent access, skip logic, answer validation, accessible completion, export, deletion, collaborator permissions, sub-processors, incident handling and data location. A familiar free form may be adequate for a non-sensitive anonymous test, yet unsuitable for identifiable family research.

First-party statements need careful reading. Jisc says its Online Surveys security arrangements store account and survey data in AWS in the Republic of Ireland, encrypt data in transit and at rest, and retain backups for 30 days.

SmartSurvey's customer data guide says its core survey platform keeps data in UK centres but warns that optional integrations, AI, APIs and webhooks can change processing. Those are supplier claims to verify contractually, not independent certifications of a proposed study.

Microsoft and Google publish controls for collecting and exporting form responses. Storage and access depend on account type, tenant configuration and linked files. Check the full journey with dummy records, including whether a spreadsheet survives after responses are removed from the form. Keep a dated assessment because services and terms change.

Analyse without turning anecdotes into a market

Begin analysis soon after each session. Separate what was directly observed from the team's explanation and the action proposed. The GDS analysis guide recommends extracting observations, grouping them, forming findings and then deciding actions, with several observers involved to reduce individual bias.

Keep disagreements. If four people use official guidance easily and two cannot locate a local variation, the finding is not that all parents are confused. It may justify testing clearer signposting with the affected group. State the sample, eligibility and method beside the conclusion.

Survey percentages need the denominator and question wording. Distinguish invited people, eligible respondents, starters and completed questionnaires. Check missing answers and routing before calculating a result. Report voluntary follower polls as exactly that. Do not add decimal places that imply sampling precision the design cannot support.

Persona cards can summarise recurring adult needs after analysis, but they must point back to rounds and observations. Avoid stock photos, fictitious quotes and invented biographies. A label such as "carers verifying local childcare eligibility" keeps attention on the task and is less likely to smuggle stereotypes into commissioning.

Make a bounded editorial choice

End each round with a decision log, not a presentation full of interesting quotes. Record the evidence, limitation, action, owner and test date. Possible actions include narrowing a guide to one life stage, providing an audio version, removing a question that solicits child data, or cancelling an unsupported commercial concept.

Set a threshold before looking at results. A team might require the same task barrier across more than one recruitment route before commissioning a series. It might demand a successful accessibility test before launch, or refuse a sponsor brief that depends on collecting information about children.

The appropriate threshold depends on risk and cost; choosing it early limits motivated interpretation.

Research closes the gap between an editor's assumption and an adult reader's situation. It does not confer ownership of a participant's story, guarantee demand or justify a national claim. Keep the question narrow, the sample described, the family data minimal and the resulting choice open to revision.

Before you act

Common questions

Can I use the Department for Education's childcare survey to estimate how many parents follow parenting creators?

No. The survey covers parents of children aged 0 to 14 in England and documents childcare use, affordability and availability. It did not ask whether respondents follow a parenting creator or would pay for a guide. Use it to identify situations worth recruiting across, not to claim demand.

How many people should I interview for a qualitative round?

The Government Digital Service guide to finding participants gives a typical range of four to eight participants for an interview or usability round. That range supports close qualitative learning, not statistical estimates about England. Recruit actual or likely users and make deliberate provision for disabled people and different social circumstances.

What should I do if my research question involves a child?

Redesign the question around an adult's decision and use fictional test material rather than family records. GDS guidance says research with a child or vulnerable adult requires informed agreement from both the participant and an appropriate responsible adult. Any such plan also needs qualified safeguarding and privacy review.