Competitor research is the cheapest source of test ideas. Instead of inventing from scratch, you look at what competitors have been running for 2-3 months and start from there. If an ad has held a budget consistently, it worked.
TikTok Creative Center: starting point
Top Ads Dashboard in TikTok Creative Center is the most underused tool in marketing. Free, official, filtered by industry, region, period and objective.
You go in, pick the industry (Education, Gaming, E-commerce etc.), pick the country. You see top 100 ads by CTR, CVR or likes, with public metrics. At Mondly we check Education + US weekly to see which patterns are scaling.
TikTok Ad Library: all active ads
TikTok Ad Libraryis the equivalent of Meta Ad Library for TikTok. You search a brand (e.g. “Creative Tim”), see every ad currently running, in which countries, for how long.
What you don’t see: exact budget, audience, performance. But you see start date. If an ad has been running 40+ days, it’s clearly performing. That’s the one you analyze in detail.
The “sponsored content search” method
On TikTok, search directly for “@brandname” or a keyword from the competitor’s product. Filter by Latest. Ads have the “Sponsored” tag under the name. You see real engagement (comments, saves), not just views.
Trick we use often: open a fresh TikTok account, behave like the target audience (follow 20-30 accounts in your niche, like relevant content). In 48 hours, the algorithm starts serving you competitor ads in the FYP. You save them.
What you concretely look for in a competitor ad
You don’t copy the ad. You extract components: hook structure (question? counterintuitive claim? POV?), angle (price, result, identity, FOMO?), format (UGC, talking head, screen demo, meme-style?), CTA.
Build a spreadsheet with these columns, analyze 30 ads. The patterns surface on their own. At Orient Ceramic we found 7 of 10 competitors used the same angle. We picked a different angle, CPA was under half.
Third-party tools that help
For faster analysis: Foreplay, Minea, AdSpy, Pipiads. They cost between 30 and 200 euros/month. Advantage: bigger libraries, advanced filters, save to collections. Disadvantage: derived data, not official. I use them as accelerators, not primary source.
Research cadence
1 hour a week, consistent. Pick 5 main competitors, scan Ad Library, save new ads. Once every 30 days, audit: which formats scaled, what disappeared, which angles appear at multiple competitors at once (trend signal).
Competitor research doesn’t replace your own testing. It gives you better starting hypotheses. You start from “this has been running 60 days at a competitor”, not from “I think this might work”. The budget difference shows.
