Most of what I write on this blog is strategy and opinion, backed by what I've seen across brand partnerships. This one's different. This is one campaign, one ad, real Ads Manager numbers, no rounding up.
DefensivX, a UAE-based app, came to me wanting a single street interview ad to test. Here's exactly what happened when it went live.
The setup
One street interview video. Filmed on location in Dubai, in the same minimal style I use for every shoot: no crew, no script, just a mic and real people giving real reactions. It ran as a single ad, targeted the same way you'd target any other UGC creative on Meta.
Week one: 6.29x ROAS
In the first seven days:
- 6.29x Purchase ROAS
- 53 website purchases
- AED 22.78 cost per result
- 23,244 impressions, 17,775 reach
- 3.47% CTR
For context, a "good" ROAS on Meta for most e-commerce and app campaigns sits somewhere between 2x and 4x. This ad cleared that in week one, off a single piece of street interview content.
Month one: still holding at 4x
The real test of any ad isn't week one. Plenty of creative gets an initial spike and then fatigues fast. A month in, with spend scaled up 5.1x from where it started:
- 4.00x Purchase ROAS
- AED 24.5K revenue generated
- 172 new customers acquired
- 42.11% hook rate
- 3.32% hold rate
A hook rate above 40% is strong for any format. It means more than 4 in 10 people who saw the first few frames kept watching. That's the street interview format doing what it does best: the opening reaction is interesting enough that people don't scroll past.
Scaling spend 5x and only losing a third of your ROAS is the sign of creative that isn't fatiguing. A lot of studio-produced ads can't say the same after week one, let alone at 5x the original budget.
What the client said
Ahmed from DefensivX put it simply afterward: the lowest CPC they'd ever seen, a CPA around 25, and results they said they'd never had before from any other creative.
Why this worked
Three things, consistent with what I've seen across every brand I've worked with:
The hook was real, not written. The opening reaction wasn't scripted copy, it was someone's actual response. That's what got the 42% hook rate.
It didn't fatigue at scale. Because the content doesn't look like an ad, the algorithm doesn't treat it like one, and viewers don't burn out on it the way they do with obviously branded creative.
One piece of content, sustained results. This wasn't a content calendar of 20 videos. It was one ad that kept performing a month later.
You can see the actual ad and full Ads Manager screenshots in the case study section on the homepage.