TikTok campaigns, read and drafted at the gate

TikTok is the platform where the levels look familiar and behave differently. The three tiers have the same names they have elsewhere, and then the schedule turns out to live one level down, targeting is not a campaign-level concept at all, and the budget may or may not be yours to change depending on a setting made when the campaign was created. A tool layer earns its keep by knowing which of those is true for the entity in front of it.

A connection is a grant against the TikTok Marketing API, scoped to the advertiser accounts you pick. Tempio syncs campaigns, ad groups and ads with the daily numbers behind them, and normalises the lot into the shape every connected platform lands in: integer minor units with an explicit currency, base counts rather than stored ratios, one canonical id per entity for a change request to point at.

What an agent can then ask is platform-agnostic by design. Blended KPIs and a per-channel breakdown over a date range. Per-campaign rows highest spend first, each in its own account's currency, with a marker saying whether the order is a genuine ranking. The structure below any campaign, each row carrying the operations that level can express. The media library, each asset with the id the platform holds those bytes under, so a creative references an upload that already exists rather than sending the same file twice. And the account brief, which is the one that changes how an agent answers: ninety days of coverage, totals, a twenty-eight day comparison, objective mix, spend concentration and weekday rhythm, so a number is read against what is normal for the account rather than against a feeling.

Every ratio is computed when it is read. A campaign with impressions and no clicks has no CTR worth reporting as a zero, and reach is a sum of each campaign's daily reach rather than a count of people, so an agent quoting it calls it what it is.

Three levels: campaign, ad group, ad. The differences from the platforms beside it are where the care goes.

Money can sit at the campaign or at the ad group, and which one is writable is a fact about the campaign rather than about the platform: when campaign budget optimisation owns the money, the ad groups beneath it do not. The connection reports that both levels are writable somewhere on this platform; the entity in front of you decides which, and a draft aimed at the wrong one is refused rather than quietly applied to the other.

Schedule and targeting are ad-group concepts here. There is no campaign-level schedule to set and no campaign-level targeting to change, so an instruction that would be a one-line campaign edit on another platform is a set of ad-group edits on this one. The tool layer refuses the campaign-level version outright instead of picking an ad group on your behalf.

Removal is terminal. There is no archive state: the third status value is a delete and it cannot be undone. As on Search, that makes a reversible status change the instruction of choice and a removal something no rule clears on its own.

What the API withholds is a long reporting window. Daily insights come back thirty days at a time, which is a hard limit on the platform rather than a preference, so a ninety-day question is three requests and the sync plans it that way. Today's figures are fetched and marked partial rather than presented as final, and a trailing fortnight is refetched on every sync to absorb the platform's own restatements.

One call writes a whole batch of changes as change requests, in the order they apply, each entry able to name the earlier entry that creates its parent. The reply is the batch, a review link for the group, and the authority that cleared it — waiting for a person, or cleared on the way in by a rule the workspace wrote. Each row carries its place in the run and, once cleared, whether a person, a key or a policy did it.

There is no server-side dry run on this platform, so validation is what the tool layer can check on its own: the operation has to be one the platform can express, the level has to be one that holds the thing being changed, the currency is stamped from the target account rather than taken from the request, and the guardrails run before the row can wait at the gate. The platform's own verdict arrives when the change is applied, and the row records it either way.

Creates land paused, so approving a batch builds structure without spending. Starting delivery is a second change through the same gate, which is a deliberate seam rather than an inconvenience: it is the difference between approving a structure and approving a spend.

Ad review on this platform is its own queue and its own timetable. A new ad enters it, and an edited creative re-enters it; a change can be applied and verified as landed while the ad sits in review and delivers nothing. The record separates the two so nobody reads a flat first day as a failed change.

Pacing follows the ad group's budget and its schedule, and a change to either is an instruction from the moment it lands. The most common surprise is not pacing at all — it is the thirty-day reporting window meeting a question about a quarter, where the answer is assembled from three requests and the oldest of them may have been restated since. That is why every figure carries the day it was read and why an account brief older than a week says so before it is quoted.

Applying a change to a live TikTok account is in private beta, behind the platform's own app review. Everything before it is built: the reads, the numbers, the plan, and the gate every change waits at — approved by you, or by a rule you wrote.