Month 2 more than doubled the panel. What follows: the Meta audience the reviews map to, a shortlist of clips worth running as ads, and 8 static concepts written in the same voice as the panel.
At n=9 the panel read as one clean-label PB lover. At n=24 it splits four ways. Each cluster is tight enough for its own creative angle and broad enough to sit inside one Advantage+ audience on Meta.
Health-focused shoppers looking for a clean-label version of a snack they already crave. Ingredient-aware, wellness-led, 25 to 45, majority female.
Reading the pack before it goes in the trolley. "Proper food" is the trigger. 32 to 55, family-first, ingredient-aware.
Vegan or plant-based. Palm-oil-free is table stakes. Cares who owns the brand. Ocado, Waitrose and Holland & Barrett before Tesco.
Ages 23 to 38. Meal-deal spotters. Follows food TikTok. "Try new things" is their number-one shopping priority.
| core audience | UK adults 25 to 45Female-weighted roughly 65/35 to match the panel. Keep men in the pool. |
| structure | Advantage+ base, light interest layerOne soft signal per cluster. Meta learns the audience once, we test the angle inside it. |
| interest pool | Plant-based, palm-oil-free, peanut butter, yoga, sustainable livingRotate the layer to test which cluster ships the cheapest CPA. |
| lookalike seeds | DTC purchase completers, 10%-off email signups, PB&J flavour engagersThree seeds, one lookalike per seed at 1% and 3%. |
| placement | IG Reels + FeedTier A creator clips first. Static concepts follow. |
| read on | CM3 + incrementalityHold the test to ~30 purchases before calling it. |
Below: 6 clips from the panel, a shortlist for paid use, and 8 static concepts written in the same voice the reviewers already use.
6 hero cuts from the 24-strong panel. Every quote is verbatim and short enough to drop into creative. Below the grid: the shortlist of which clips to run, hold or leave alone.
Ranked on emotional payoff, verbal hooks, on-camera reaction and would-buy-again intent. Tier A is ready for full-form ads. Tier B works as 6-second retargeting or pre-roll cutdowns. The skip pile has audio or sentiment that will cost more than it earns.
Each concept is a hook, a supporting line and a CTA. Every line is drawn from what the panel actually said, or what Superfoodio already prints on the pack. Design is left to the designer. What follows is the copy to work with, and the customer voice each concept leans on.
Provisional, and firming. Built on 22 reviewers across 24 reviews, still one recruitment source and a positive lean. The Wellness Seeker spine, the "proper food" hook and the Tesco / Waitrose signal all repeated in Month 2. The four-cluster split is new to this batch. It should be re-tested against Month 3 before it drives meaningful spend. What sits above is behavioural intelligence at Month 2 depth. Finished targeting and finished creative come after the next data cut.