The question of starting budget comes up systematically when launching a first TikTok Shop Ads campaign. A budget too small does not let the algorithm learn properly, a budget too high without a sufficient test base wastes resources on unverified assumptions.
Size the budget around the target acquisition cost, not an arbitrary envelope
Rather than starting from a randomly chosen amount, it is more useful to estimate a reasonable acquisition cost for the product in question, then plan a budget that can generate several dozen conversions during the test phase. That volume of data is what allows the algorithm to start optimizing delivery.
Spread the budget across several creatives rather than one
Concentrating the entire test budget on a single creative limits your ability to identify the best performing angle. Spreading that same budget across three to four different creatives, even modestly, gives a far more useful comparison base for the rest of the campaign.
Plan a margin for the learning phase
The first days of a campaign serve the algorithm's learning phase, during which results remain unstable. Planning a budget that covers this phase without spending the entire available budget on it avoids cutting the campaign before it has had time to stabilize.
The starting budget is not meant to be profitable immediately. It exists to generate the data needed to make the next campaign profitable.
Key takeaways
- Size the budget around the target acquisition cost and the conversion volume needed for learning
- Spread the test budget across several creatives rather than one
- Reserve a margin for the learning phase before judging profitability
We define this starting budget with each brand based on their catalog and margins. Let's discuss yours.