Adding another interest to a TikTok ad group can make the setup look more selective while actually broadening the pool under consideration. The important question is whether the selections mean “any of these” or “all of these.” A longer list is not necessarily a tighter definition.

TikTok's targeting documentation, updated in October 2026, states that multiple selections within one dimension use OR logic, while conditions across dimensions use AND logic. Marketers planning B2B campaigns need to understand that distinction before interpreting an audience estimate or explaining why they changed a campaign.

Read the settings as a sentence

Describe the ad group in ordinary language. If several interests are connected by OR, a person can qualify through any one of them. Adding an interest brings another set into the candidate pool. It does not require someone to share every selected interest. This is a statement about the selection logic, not a forecast of delivery, pricing or business results.

Conditions from different dimensions work differently. A location condition and an interest condition must both be satisfied under the documented AND relationship. Additional available demographic conditions can constrain the intersection further. Similar-looking controls can therefore perform opposite jobs: one adds alternatives, another adds a requirement.

Check the controls actually available in the specific campaign and market. Automated expansion settings and campaign modes deserve their own review; do not assume a sentence that describes one setup will describe another. A useful campaign note includes the operative conditions rather than only a screenshot full of checked boxes.

This matters when a team asks an agency to target a professional audience. “Interested in furniture or interior design” is not equivalent to “buys furniture for a distributor.” The first describes platform signals. The second describes a commercial role and task. No number of related interests turns the former into verified evidence of the latter.

Define the buying task outside the checkbox list

An international distributor and a company purchasing furniture for a new office may both encounter the same category. They still need different information. The distributor may want a coherent range and a supply arrangement. The office buyer may need dimensions, intended use and the information required for a quotation. The ad should help those readers recognize which discussion is relevant to them.

Chinese suppliers preparing material for an overseas campaign can make that distinction before an agency touches the audience controls. Specify the offer's intended buyer and the question the landing page will answer. If the business cannot explain those two things, an elaborate targeting combination is unlikely to clarify them.

A lead form can then ask about intended use or the purchasing task, where appropriate. Such questions give the sales team something concrete to evaluate. They should not be added merely to make the form look professional, and they should not request unnecessary personal information. Interest selection and qualification questions have different jobs.

Where budget permits, separate genuinely different audience hypotheses into tests with matching creative and follow-up. That makes comparison more interpretable. It does not mean every new interest deserves its own ad group. Excessive fragmentation can leave a small team maintaining many combinations without enough evidence to distinguish them.

For a limited pilot, a single intelligible hypothesis is preferable to a collection of unexplained settings. Write it before launch: this offer is intended for this purchasing situation, these platform conditions approximate that audience, and this question will help us assess relevance after the click. The approximation should remain visible in the team's reasoning.

A campaign history entry reading “optimized targeting” tells the next marketer very little. Record whether a change added an alternative or added a restriction, and why the new audience is worth testing. That description lets someone compare the intended pool before and after the change.

Changing location, interests, creative and the form at the same time may be operationally necessary, but it makes attribution harder. If the result changes, the team cannot confidently assign the change to one setting. Keep the major comparison clear when feasible, and acknowledge combined changes when isolation is not practical.

Audience construction is also separate from campaign health diagnostics. The former describes who is considered under the configured conditions; the latter concerns platform-side conditions affecting delivery. Review the audience sentence, the delivery diagnosis and the buying task reported in the form separately before deciding which part to change.

The next time someone proposes selecting more interests to make a B2B campaign precise, ask them to read the resulting audience sentence aloud. If the sentence has become broader, that may be a reasonable experiment. It should simply be called a broader experiment, with a buyer question that makes its results useful.

Sources