Campaign Testing Structure
Posted: Sun Apr 20, 2025 5:53 am
We always create a test campaign at the start, which runs continuously in parallel with the scaling campaigns . In addition to the ads themselves, we also test different optimization strategies and various ad sets with different targeting settings :
Broad (no restrictions)
Automatic Targeting
Cluster of interests
Hashtag-Interactions
Creator-Interactions
Video-Interactions
We start with a small budget and an initial selection of creatives, ideally 3-5 tunisia phone number data creatives per ad set . We pay particular attention to using features specific to the platform such as current trends, dialogues, POV (point of view) or question stickers . It is also particularly important to use the platform's original fonts. The easiest way to do this is to edit the video directly in the TikTok editor or using the “CapCut” app, which also belongs to TikTok's parent company ByteDance. In the testing campaign, we ensure that each ad set receives the same amount of budget so that we can test them against each other effectively. The creatives within an ad set should be structured so that they have the same angle (basic structure) and only differ in the hook so that we can decide which creative performs best based on this starting point.
Testing campaign structure and design
Testing campaign structure and design
To determine the success of a creative, we look at the following values:
CTR, CPA, CVR and Average Time Per Video View .
These metrics should never be considered individually, but always together. A high CTR does not necessarily mean a low CPA, and a high average time per video view does not necessarily mean a high CVR. We therefore decide which creatives to send to scaling based on the combination of these metrics . Here, too, it is important that not only the ads themselves are right, but also what happens afterward. A good ad can bring a lot of potential customers to the landing page or shop - but if the user experience is not right, then probably no purchases will be made and the CPA will rise, even if the ad performs well in terms of CTR, CVR and average time per view.
Broad (no restrictions)
Automatic Targeting
Cluster of interests
Hashtag-Interactions
Creator-Interactions
Video-Interactions
We start with a small budget and an initial selection of creatives, ideally 3-5 tunisia phone number data creatives per ad set . We pay particular attention to using features specific to the platform such as current trends, dialogues, POV (point of view) or question stickers . It is also particularly important to use the platform's original fonts. The easiest way to do this is to edit the video directly in the TikTok editor or using the “CapCut” app, which also belongs to TikTok's parent company ByteDance. In the testing campaign, we ensure that each ad set receives the same amount of budget so that we can test them against each other effectively. The creatives within an ad set should be structured so that they have the same angle (basic structure) and only differ in the hook so that we can decide which creative performs best based on this starting point.
Testing campaign structure and design
Testing campaign structure and design
To determine the success of a creative, we look at the following values:
CTR, CPA, CVR and Average Time Per Video View .
These metrics should never be considered individually, but always together. A high CTR does not necessarily mean a low CPA, and a high average time per video view does not necessarily mean a high CVR. We therefore decide which creatives to send to scaling based on the combination of these metrics . Here, too, it is important that not only the ads themselves are right, but also what happens afterward. A good ad can bring a lot of potential customers to the landing page or shop - but if the user experience is not right, then probably no purchases will be made and the CPA will rise, even if the ad performs well in terms of CTR, CVR and average time per view.