A ninety-day parenting pilot that tests an adult need and sets the safeguards first

Use 90 days to make one evidence-based decision, not to promise a follower or revenue result. The pilot should show whether a defined group of adults can use a small body of work, whether the creator can maintain it safely and what the real production and review costs are.

What to take away

Days 1 to 15: choose the problem and limits

Write the adult reader, recent family task, England-specific element and decision the pilot will inform. Turn assumptions into research questions. GDS guidance on planning user research recommends prioritising questions and organising rounds around specific user groups.

Create the child boundary before choosing formats. State that no under-18 will be recruited, filmed or profiled unless a later, specialist-approved method makes participation necessary. List family details the project will not collect. Assign someone with authority to stop production if a safeguard fails.

Open a source register for every factual proposition. Give entries a geography, publication date, owner, review date and limitation. Do not convert UK-wide media findings into an England audience estimate.

Days 16 to 30: listen to adults and set a baseline

Recruit adults who recently faced the chosen task through more than one route. Record eligibility and recruitment bias. The GDS page on finding participants gives four to eight people as a typical qualitative round, not a statistically representative sample.

Ask about the last real attempt, sources checked, obstacles and outcome. Use informed participation materials, collect only what the question needs and agree deletion dates. Summarise patterns and disagreements without publishing identifiable family stories.

Measure the current starting point only where data exists: relevant page visits, qualified replies, corrections, production hours and direct costs. Do not backfill a target from another creator's public counters.

Days 31 to 60: publish the smallest useful series

Choose one primary format and produce a short connected set that solves the same adult task from different stages. Every item should state its evidence boundary, use an accessible alternative where needed and lead to the complete source-backed answer.

If an item has a sponsor, affiliate link, gifted product or own-brand promotion, design the disclosure into the opening. The ASA's current influencer advertising guidance normally expects an obvious upfront "Ad" label. Record the benefit and approval trail even when a planned item is cancelled.

Collect task evidence with low-risk methods. A voluntary adult follow-up asking whether the checklist was completed is more relevant than raw impressions. Do not request a child's image, diagnosis, account or school as proof.

Days 61 to 75: inspect failures

Check factual corrections, misunderstood steps, accessibility barriers, moderation time, unsafe submissions and source changes. Interview a small number of eligible adults who did not complete the task. Compare actual hours and spending with the plan.

Review any sign-up route against the ICO's April 2026 electronic-mail marketing guidance. Confirm that choices, notices, suppression and deletion work in practice. Remove test exports and contact data whose purpose has ended.

Days 76 to 90: make the next decision

Bring together reader outcomes, qualitative evidence, content quality, safety incidents, workload, costs and attributable income. GDS performance guidance recommends using metrics with user research and changing measures as understanding develops.

Choose one outcome: continue unchanged, revise the audience or format, pause for qualified review, or stop. Set future targets only from the observed baseline and available capacity.

Preserve the decision record, not unnecessary participant material. A stopped idea is a valid result when evidence shows weak need, an unsustainable route, or family risk that cannot be controlled.

Before you act

Common questions

What should the pilot measure at the start?

Measure the current starting point only where data exists, such as relevant page visits, qualified replies, corrections, production hours and direct costs. Do not backfill a target from another creator's public counters, because that figure does not describe your own audience or capacity.

How should a sponsor or affiliate link be handled?

If an item has a sponsor, affiliate link, gifted product or own-brand promotion, design the disclosure into the opening. The ASA guidance normally expects an obvious upfront Ad label. Record the benefit and approval trail even when a planned item is cancelled.

What counts as a valid result after 90 days?

Choose one outcome: continue unchanged, revise the audience or format, pause for qualified review, or stop. A stopped idea is a valid result when evidence shows the need is weak, the route is unsustainable or the family risk cannot be controlled.