Success Story
Jewellery Brand on TikTok: How a Post-Purchase Survey Turned a 'Failing' Channel Into a Profitable One
TikTok showed a jewellery brand under 1.0x ROAS. Blended attribution showed 2-2.5x against a 1.6x breakeven, so it scaled.
Reading time: 2 mins · Published
Challenge
A long-standing client had products that suited TikTok, but had never advertised there. We showed the owner what the platform could do for his range, and he agreed to test it.
For the first two weeks, TikTok’s own reporting showed ROAS below 1.0x: on paper the channel was losing money. We tested several winning audiences, creatives, angles and offers, and none came close to target in the platform’s numbers. The owner started to doubt TikTok could work for his products at all.
We had seen this before. Shopify revenue was rising as TikTok spend rose, which pointed to an attribution problem rather than a demand problem. Once a visitor leaves the TikTok app, the pixel can track only a fraction of them, and since iOS 14.5 many conversions are invisible to the ad platforms.
What we did
- Asked the customers. Rather than evaluating complex third-party tracking tools, we added a one-question post-purchase survey (“How did you find our store?”) with a few multiple-choice answers. It took 15 minutes to set up.
- Read the answers. 43% of buyers responded, and they credited far more sales to TikTok than either Shopify or TikTok Ads Manager did, even before counting the 57% who didn’t answer.
- Built a blended report. We combined Google Analytics UTMs, Shopify UTMs, Shopify revenue by channel, TikTok Ads Manager and the survey answers. It was not perfect attribution. Its purpose was to identify the winning TikTok ads, and to give the owner confidence that more TikTok spend was producing more Shopify revenue.
- Tightened tracking for scale. We recommended server-side (Events API) tracking, UTMs on every ad, the TikTok app on Shopify, and an attribution tool where it made sense.
Results
- Blended reporting showed TikTok returning 2-2.5x ROAS on its own, against a 1.6x breakeven, so the channel was profitable throughout and with a healthy margin.
- With that evidence, the owner kept TikTok, and we moved on to scaling it without losing profitability.
- The survey kept running as a cheap, permanent attribution check, with a 43% response rate.