TikTok Ads: who really needs it, and how not to waste the budget

Most accounts that fail on TikTok fail for one reason, and it is not the creative being bad. It is the creative running out. A team arrives with the habits of a Meta account, produces four assets, expects them to run for a quarter, and discovers that the channel consumes in a fortnight what they budgeted for a season.

That is not a flaw in the platform. It is the platform. TikTok distributes on the strength of individual pieces of content rather than on the accumulated authority of an advertiser, which means the account is only ever as good as what it uploaded this week. Deciding to buy TikTok is deciding to run a small production line, and the businesses that do well are the ones who understood that before signing.

How the delivery actually behaves

An ad is tested on a small audience. If the early signals hold, watch-through, replays, engagement, the intended conversion, it is given more. If they decay, distribution is withdrawn quickly and quietly, and no amount of budget argues with that.

Two consequences follow, and they are the whole channel.

First, the opening seconds carry more weight than anything else in the ad. Not because of a rule, but because the signal the system reads is whether people stayed, and that is decided immediately. Second, decay is normal rather than a fault. An asset that performed and then stopped did not break. It was shown to the people most likely to respond, and there are fewer of them left. We wrote about how to read that curve and when to act on it in the piece on creative burnout, which is the operational half of this article.

The formats, and what each is actually for

• In-Feed. The default and the one that carries the work. Appears between organic videos, judged against organic videos, and therefore has to survive the comparison. This is where testing happens and where most budget belongs.
• Spark Ads. Paid amplification of a real post, either your own or a creator's with permission. It keeps the original engagement, the comments and the account identity, which is why it usually outperforms the same footage uploaded as a plain ad. It is also the format that needs a relationship rather than a brief.
• TopView and other takeovers. First impression on opening the app. Reach at a price that makes sense for a launch or a national campaign, and rarely for a performance account.
• Creator content generally. Not a format so much as a supply strategy. In practice the accounts that keep production sustainable are the ones that stopped producing everything in-house.

Who it suits

Products that are better shown than described. Beauty, fashion, food, home, gadgets, anything with a visible before and after or a satisfying demonstration. If the pitch is a thing happening on camera, this is the cheapest attention available anywhere.

Consumer apps. Low friction, decision on the same device, and an audience that is native to the format. The usual caveat applies harder here than elsewhere: optimise towards installs and you will get installs from people who never open the app.

Impulse-compatible price points. The decision is being made mid-scroll. A price that requires a household conversation is being asked to survive a context that will not allow one.

Brands willing to sound unlike themselves. This is the real filter, and it is not about age. Polished brand film underperforms consistently, because it announces itself as an ad in the first half-second and the audience has been trained to skip exactly that. Businesses that cannot tolerate looking amateur should spend the money elsewhere rather than spend it here badly.

Who should not bother

• Considered B2B purchases. Reach exists, the decision-maker is in there, and the buying process is not. Judged on pipeline, it loses to search every time.
• Immediate-need services. Same logic as everywhere: an emergency is a search query.
• Businesses that cannot sustain production. If four assets a month is the ceiling, the channel will stall by design. Better to run one channel properly than two at half strength.
• Anyone whose audience is not there. The user base skews younger than Meta's, and while it has broadened considerably, "our customers are on TikTok" should be a finding rather than an assumption.

Two things worth knowing before you plan a year around it

The audience number quoted everywhere deserves scrutiny. TikTok announced one billion monthly active users in September 2021 and has not published an updated global figure since; higher numbers circulating in marketing posts are third-party estimates rather than platform reporting. The channel is large either way. It is worth knowing which of those two things you are quoting when you build a plan on it.

The other is regulatory. TikTok's status has been the subject of legislation and restriction in several markets, and that is a live situation rather than a settled one. It does not argue against using the channel. It argues against building a business whose only demand source is a single platform, which is the same argument we make about ad accounts generally on our SEO page: paid media is rented, organic is owned.

What a serious start looks like

Enough hooks to test the ideas rather than the executions, produced cheaply. A budget that can produce enough conversions to read a result rather than noise. Nothing edited while it runs. A written record of which hooks died and why, because that list becomes the production brief for the following month.

And a decision made honestly at the start about whether the business can feed the channel. TikTok rewards volume of ideas, not size of budget, and an account with money and no supply spends it faster on worse placements. If that sounds like a reason to stay out, it is a reason to stay out, and we would rather say so before the first invoice than after the third.