Ad Creative Testing for Kenyan Brands: What to Test and When to Kill It
By Joseph Maina · · 5 min read
Every brand I work with asks the same question sooner or later: why did that one ad blow up while three others just sat there eating budget? The answer is almost never luck. It's a testing process, run properly, in a market that behaves differently from the case studies you'll find on YouTube.
Test one thing at a time, not everything at once
Hook, visual, and offer are the three things worth testing on their own. Change all three between versions and you'll have a 'winner' with no idea why it won, which means you can't repeat it.
Start with the hook. In a market where people are scrolling TikTok on a Safaricom bundle they're watching the clock on, the first two seconds decide everything else.
What actually performs with Kenyan audiences
Sheng and everyday Swahili phrases used naturally, not forced, consistently outperform stiff corporate English in the comments and shares, especially for anything aimed at under-35s in Nairobi.
Real people using the actual product on camera, in a real setting like a matatu stage, a salon chair, or someone's kitchen, beats a studio shoot almost every time for engagement rate. Polish reads as an ad. Realness reads as a recommendation from a friend.
Give a test enough data before you judge it
A day and a half of spend with three clicks is not a result, it's noise. Let each variant spend enough to get a meaningful number of link clicks or add-to-carts, usually a few thousand shillings per version, before comparing them.
Judging too early is how good creative gets killed by accident and mediocre creative gets scaled by mistake.
Know your kill criteria before you launch, not after
Decide upfront what 'this isn't working' looks like: a cost per result above a set number, or a click-through rate under a set percentage, checked after the variant has spent enough to be judged fairly.
Without a kill criteria set in advance, it's tempting to keep a creative running because you personally like it, even after the numbers have made the call.
Build a refresh habit, not a one-off shoot
Budget for a new batch of creative every three to four weeks, even for campaigns that are performing well. Frequency climbs faster in Nairobi's smaller, more connected online audience than it does in bigger markets, so fatigue sets in sooner than most brands expect.
Keep a simple swipe file of everything that's ever worked, hooks, formats, offers, so testing next month builds on what you already know instead of starting from zero every time.