NHS Wales
Ready to Talk
Reducing barriers to men aged 40–60 accessing mental-health services.
AI-generated audience profiles for this scan. Each profile predicts how a segment will interpret the campaign, flags what could go wrong, shows the evidence behind the prediction, and recommends specific changes Your Agency can make.
- 28–34 · Cardiff
Rhian
AI ProfilePrimary school teacher, first-time parent.
EducationTrust: mediumInstagramBBC NewsLocal WhatsApp groupsPredicted interpretationRhian reads the hero line as another top-down instruction rather than an invitation, and assumes the ad is aimed at people who are not already doing their bit.
Risks- Disengages if the ask feels vague or preachy.
Why — evidence informing this prediction- Internal audience wave 12: 62% of 28–34 parents describe public-service ads as 'not for me'.
- Previous campaigns to time-poor parents under-indexed when the hero framed the ask as a duty.
- Behavioural science: reactance rises when the audience already self-identifies with the desired behaviour.
Based onYour Agency profilesEvidence records- Successful campaign: 'Check In On Your Mate'
- Parents & effort perception
Recommendations- Rewrite the hero from a duty statement to a concrete first-person moment (e.g. 'The school-run journey that adds up').Why · Concrete, first-person framing raises perceived relevance for time-poor parents.Expected outcome · +8–12pt relevance among 28–34 parents.
- Drop the instructional verb in the CTA; use an invitation ('Try one this week').Why · Reactance falls when the audience feels invited, not corrected.Expected outcome · Higher click-through on social crops.
- 19–24 · Wrexham
Owain
AI ProfileApprentice electrician, lives with parents.
TradesTrust: lowTikTokYouTubeRedditPredicted interpretationOwain reads institutional branding as a signal to disengage before the message lands, and interprets the campaign as being 'for people who already care'.
Risks- Actively negative reshare on social.
Why — evidence informing this prediction- Youth trust study: institutional voice drops trust 14pt among 18–25 with low prior engagement.
- Similar apprentice audience in the 2024 sustainability wave described official campaigns as 'cringe' 71% of the time.
- Behavioural science: identity-threatening frames trigger reactance in low-trust segments.
Based onYour Agency profilesEvidence records- Successful campaign: 'Check In On Your Mate'
- Parents & effort perception
Recommendations- Remove overt government branding from the first 3 seconds of video/social cuts.Why · This audience demonstrates low institutional trust; branding gates attention.Expected outcome · Fewer skips in the first 3s among 18–25 on TikTok.
- Replace the hero on social with a peer-voiced version featuring an apprentice or trades role.Why · Peer messengers outperform institutional voice in low-trust segments.Expected outcome · Reduced negative reshare risk; higher save rate.
- 45–54 · Gwynedd
Meleri
AI ProfileNHS administrator, Welsh first-language speaker.
Public servicesTrust: highS4CBBC Radio CymruFacebookPredicted interpretationMeleri interprets English-first bilingual work as translated-after-the-fact and reads that as a signal the campaign was not made with her community in mind.
Risks- Publicly critiques inauthentic bilingual work.
Why — evidence informing this prediction- Bilingual performance benchmark: Welsh-first hero artwork lifts recall 12–18% in Welsh-speaking regions.
- Gwynedd focus groups named tokenistic bilingualism as the #1 reason to distrust a public campaign.
- Similar profile in the 2023 rural insight programme shared authentic Welsh-first campaigns at 3x the rate.
Based onYour Agency profilesEvidence records- Successful campaign: 'Check In On Your Mate'
- Parents & effort perception
Recommendations- Produce a Welsh-first hero artwork for North and West Wales, not an English translation.Why · Welsh-first framing signals authentic intent, not tokenism.Expected outcome · +12–18pt recall in Welsh-speaking regions.
- Credit a Welsh-speaking community messenger in the caption.Why · Community messengers outperform institutional voice for this segment.Expected outcome · Higher authentic share-rate; reduced public critique risk.
