Case study
Pepsi owned the day. Night was the open occasion.
Pepsi Night Edition launched with a Snapchat Lens that reads how awake you are. A filter like that only rings true if the person running it is genuinely in that moment, so the roster was the campaign.
The occasion Pepsi never claimed
Everyone in the market knows Pepsi. Almost nobody associates it with 2am.
The brand is anchored to daytime, and has been for decades: meals, breaks, afternoon heat, in every asset, every year. Meanwhile the consumption was already happening after dark. Gaming sessions, late food runs, football watch parties, gatherings that start when the day is over.
None of it was claimed by the category, and none of it appeared in anything the brand had ever put out. Pepsi Night Edition was the product answer. The campaign question was narrower: make night read as a Pepsi occasion, using the people already living it.
Somebody has to run the Lens
The launch mechanic was a Snapchat Lens. Point it at your face and it reads how awake, or how far gone, you actually are. Three steps: open the Lens, get your read, post the result.
No database indexes who is awake at midnight
Every creator discovery database indexes the same four fields: follower count, category, geography, engagement rate. Not one of them tells you whether a creator's life happens after 10pm.
“Lifestyle” describes what somebody posts about. It says nothing about when they post it, or when their audience is awake to see it. Cast a daytime creator, ask them to run an awake-or-asleep test at 3pm, and the mechanic exposes the mismatch before the audience does. So we selected on behaviour.
| Signal | What it answers |
|---|---|
| Posting-time distribution | When a creator actually publishes, rather than what they publish about |
| Audience active hours | When their followers are awake, online and watching |
| Existing back catalogue | Whether nighttime settings were already there before Pepsi arrived |
Snapchat Spotlight carried it, because that is where the Lens lives, where the audience already is after dark, and where there is no detour between using a filter and posting the result.
Two waves, the second briefed by the first
The roster was not fixed at launch. Wave one put 16 creators live, all 16 of them publishing, and the point of it was the read: which night contexts actually carried.
Wave two was then cast against that answer rather than against the original plan, and took the campaign to 32 Spotlights live in total. A single 43-creator launch buys the same roster without ever learning which half of it was working.
| Wave | Creators | Spotlights live | What it was for |
|---|---|---|---|
| Wave 1 | 16 | 16 | Establish which night contexts carried |
| Wave 2 | 27 | 16 | Cast against wave one's read, remainder scheduled |
The ecosystem behind the cast
Spotlight carried the mechanic. The 43 creators carried an audience across three platforms.
| Platform | Combined following |
|---|---|
| TikTok | 6.04M |
| 5.67M | |
| Snapchat | 3.21M |
| Total | 14.9M |
Selected content
Published Spotlights from the cast. Each links to the live post.
1spdrSnapchat Spotlight · UAEView the post Rami alostaSnapchat Spotlight · UAEView the post
Run this for your brand
Halo AI sources, vets, briefs and chases the creators, and you keep approval on the roster and the content.
Book a demo
