"The phone can wait. People can't." Pressure-testing a digital-wellbeing campaign
A company preparing a public campaign on responsible internet use wanted to know whether the public would believe it — before spending the budget. So we tested it on 300 True Twins™: who they'll trust as the messenger, what they'll dismiss as corporate spin, and which exact line travels. One message won on every metric. Another was branded "marketing language, seen a thousand times."
A responsibility campaign only works if the public believes the messenger
A corporate online-safety campaign is a credibility bet. Done well, it signals genuine responsibility; done wrong, it reads as the brand polishing its image on a problem it profits from. The company knew the topic mattered — but a campaign on digital wellbeing is exactly the kind of message a skeptical public is primed to dismiss as spin. Before committing budget and brand reputation, they wanted to know what the audience would actually accept.
Derisk the campaign before launch
The goal was a pre-launch reality check: not whether the issue resonates, but whether this brand can speak to it credibly, through whom, and in what words. Getting any of those wrong in public is costly to walk back.
A rare consensus — and a sharp warning.
Three candidate lines, one clear winner
The same 300 respondents rated three campaign messages on attention, relevance, authenticity and shareability. The gap between them was decisive — and instructive.
The lesson is unambiguous: direct, human language that names a relationship wins; the abstract "balance / wellness" register loses. Respondents described the losing line, verbatim, as "marketing language, seen a thousand times" — the same instinct that flags corporate campaigns as hollow.
The messenger matters as much as the message
Credibility was almost unanimous, and it has a clear shape: put expertise and lived experience on camera; keep influencers off it entirely.
Psychologist = #1 choice for 81.6%; influencer = the worst choice for 98.1%. Parent and peer voices rank as the strongest supporting messengers — the audience wants expertise plus authentic lived experience.
What the study told the brand to do
You don't need to convince anyone the problem is real — skip straight to "what we're doing about it"
With 83.5% calling excessive use a serious problem and 96.3% rating it personally relevant, the persuasion battle is already won. Spending the campaign establishing that screens are a problem would waste the budget on a point the audience concedes instantly. The opportunity is to move directly to credible action.
"Tools, not talk" — the anti-greenwashing trap is the central risk
69% want concrete tools (filters, time limits, controls) from the company, not campaign slogans; 96.7% flag "empty talk without action" as the #1 red flag. A digital-wellbeing message from a brand the audience associates with the problem is credible only if it's paired with something real. The audience even articulated the honest framing themselves.
Lead with the psychologist; never the influencer
The credibility hierarchy is nearly unanimous and directly castable: a psychologist as the central authority, parents and peers as supporting voices for authenticity, and influencers avoided entirely. On this topic an influencer doesn't just underperform — they actively undermine the message for 98.1% of the audience.
Optimism is conditional — and the conditions are the brief
69.7% believe education and campaigns can help, but only under explicit conditions: honesty, leading by example, and concreteness. That conditional optimism is the whole strategy in one sentence — the audience is willing to be moved, provided the brand earns it. The winning message and the demand for real tools are two halves of the same expectation.
A campaign derisked before a cent of media spend
The brand went into launch knowing what to say ("the phone can wait, people can't"), who should say it (a psychologist, with parents and peers alongside, never an influencer), what to avoid (abstract "balance" language and any message unbacked by real tools), and why the audience would accept it. The single most expensive mistake on a campaign like this — sounding like corporate spin on a problem you profit from — was identified and designed out in advance.
The audience already agrees there's a problem. What they won't forgive is a campaign that talks about it without doing anything — from a brand that profits from their attention.
Test the campaign on the audience before the audience ever sees it
Atani True Twins™ are individual AI models, each built from a real, recruited respondent profiled across 100+ dimensions. For pre-testing a public campaign, that meant the brand could fail safely in private — and iterate at no marginal cost:
A/B/C message testing, instantly
Three candidate campaign lines tested head-to-head on the same 300 respondents — attention, relevance, authenticity, and shareability — in hours, before a cent of media spend.
Grounded in real Serbians
Every twin traces back to a real recruited respondent, so the credibility judgments reflect the actual market's skepticism — not a generic model's guess.
Fail safely, in private
A weak message or a wrong messenger could be caught and discarded internally — instead of on a billboard, where it becomes the story.
The why, not just the score
Open-ended answers coded across 14 dimensions explained why a line worked — so the winner came with a usable creative rationale, not just a higher number.
We fielded a tightly structured 13-question instrument to 300 validated True Twins™ spanning Serbia's regions, genders, age groups, education levels and settlement types. The instrument combined closed scales (1–10, 1–5), ranking, forced choices, and open-ended questions across four themes: personal stance on the problem, which age groups are most vulnerable, credibility and sources of voice, and operational recommendations for the campaign — including a head-to-head test of three concrete messages.
For every question we produced a full-sample distribution plus breakdowns by age, gender and region. Open-ended answers were coded by LLM across 14 dimensions — sentiment, problem framing, belief in change, credibility criteria, attitude to influencers, what people demand from the company, and attitudinal typology — using language comprehension rather than keyword matching. Every verbatim quote was validated against its source response before inclusion.
This is a pre-launch diagnostic on a validated panel, not a probability sample of the national population — read the percentages as strong directional signal on attitudes and message reception, not as official population estimates. The message test measures stated attention and shareability, not in-market behavior or recall. As with all pre-testing, it derisks a decision; it does not replace measuring the live campaign once it runs.
We hold no identity data. Our AI twins are built from anonymised interviews conducted by an independent research agency, processed under strict data separation, and governed by a Code of Ethics that covers every client, every use case, and every prompt. GDPR-compliant by design.