How long should you run a TikTok ad before killing it?

The question behind the question

“How long should I run this ad?” is really asking “when do I have enough information to decide?” - and the honest answer is measured in money and events, not days. A day of spend at $20 tells you almost nothing; the same day at $100 might tell you plenty. So the rules below are framed the way media buyers actually think: spend thresholds relative to your product’s economics, checked at set moments, with one number - your breakeven - decided before you launch.

Before launch: know your kill number

Everything depends on one figure: your profit per sale, which sets your breakeven cost per acquisition. If you sell at $40 with $15 landed cost and $2 fees, you make $23 per sale, so any ad costing more than $23 per sale loses money. Work this out with the breakeven ROAS calculator before spending a cent - an ad can only be judged against a number you actually know. All the timing rules below are multiples of this breakeven figure.

The first 24-48 hours: judge the creative, not the product

TikTok’s algorithm needs a learning period, and early results are noisy - but the creative signals are readable almost immediately. After roughly $20-30 of spend per ad, look at: hook rate (are people stopping in the first two seconds?), watch time, and click-through rate. A CTR around 1% or better with decent watch time means the creative works; keep spending. A CTR far below that with people scrolling past in under two seconds means the ad is dead on arrival - kill the creative and test another angle, but don’t condemn the product yet. Most “failed products” are actually failed hooks.

The kill checkpoints that actually matter

• At ~1x breakeven spend with zero add-to-carts: kill the ad. If your breakeven is $23 and you’ve spent $25 with nobody even carting, neither the creative nor the audience is connecting.
• At ~1.5-2x breakeven spend with carts or checkouts but no sale: the ad is doing its job and the store is dropping the ball. Pause spend, fix the product page, price presentation, shipping clarity or trust signals (our store guide covers the usual leaks), then relaunch.
• At ~2-3x breakeven spend with a sale or two but above breakeven cost per sale: judgment territory. If the trend is improving day over day, give it one more breakeven’s worth of spend; if it’s flat or worsening, kill it and iterate the angle.
• Profitable at any point: don’t touch it for at least 3-4 days. Editing a winning TikTok ad resets learning; scale by raising budget gradually (roughly 20-30% at a time) or duplicating, not by editing the live winner.

Why TikTok specifically punishes patience

On Facebook, ad longevity is the signal - winners run for weeks and the auction rewards stability, which is why long-running Facebook ads are such a strong profitability tell on our Facebook product lists. TikTok burns creative much faster: the same audience sees your ad repeatedly within days, frequency climbs, and performance decays even on winners. Plan for a winning TikTok creative to have a lifespan of one to a few weeks, keep two or three fresh angles ready to rotate, and treat creative production as an ongoing cost of the channel, not a one-time task. This is also why what’s being advertised on TikTok right now changes faster than any other surface we track.

The mistakes that waste the most money

Killing at $15 of spend because there were no sales yet (you bought noise, not an answer). Running to $300 on hope with zero carts because “the algorithm needs time” (the learning-phase excuse has a limit, and it is roughly the checkpoints above). Judging the product when the hook failed. Editing a profitable ad. And launching without a breakeven number, which makes every other decision a guess. The whole checkpoint system is just validation continued into the live phase: decide the numbers first, then let the numbers decide.

FAQ