2022 – present · Solo developer — the delivery mechanism, the microsite, and the per-campaign builds
Cavai Anywhere
A way to run our interactive creatives in places that can't host them, from social feeds to a QR code on a poster.
- Engineering
- Product
The brief
Cavai's whole product is the interaction. The problem is that a lot of the places clients actually want to buy won't run one. Facebook and Instagram give you an image or a video and a click out, and that's the entire surface you get. A poster on a wall gives you less than that. So whenever someone asked if we could run inside a walled garden the answer was no, and that's an awkward thing to keep saying about a large share of where the money goes.
The approach
The creative doesn't run in the placement. It runs on a microsite, and the placement's only job is to get you there. That part is straightforward.
The harder part was making the jump not feel like a jump. If a Facebook ad throws you out into Safari you lose most people, and the ones you keep now feel like they've left the app and ended up somewhere else. So on Facebook it opens in the app's own built-in browser instead. You tap the ad, the creative comes up, and you're still inside Facebook as far as it feels. The whole goal is for the handoff to be boring enough that you don't notice it happened.
Because the creative is really just sitting on a URL, it doesn't care much what sent you there. Social was the reason I built it, but the same mechanism now runs behind QR codes on printed posters, which works exactly the same way.
My role
This one is mine end to end. I built it and I've maintained it since the first commit in October 2022. Another developer drops in now and then to deal with an individual ad, but day to day it's been me on my own.
The result
The best numbers I have come from BMW's Nordic Christmas campaign, which ran the format across Norway, Sweden, Denmark and Finland. It pulled around 510,000 loads and about 285,000 interactions, so roughly 56% of the people who landed on it did something with it.
I'd take that 56% with a pinch of salt. These are people who already tapped an ad to get there, so they've self-selected in a way that someone scrolling past a banner hasn't. It isn't the same kind of number as a display interaction rate and I don't want to present it as one. What it does tell me is that once you get someone to the creative, the thing holds up.