THE CURRICULUM / CH 3 — Marketing Mastery

3.2TikTok Setup

The same spine, built fast — then the differences that actually decide the outcome.

5 min read

The ad that carried your Meta account for a month gets re-cropped to vertical, uploaded on a Monday, and by Wednesday it has spent its budget and sold nothing. The pixel is fine. The account is fine. It is a commercial, playing in a feed where the video directly above it was shot at arm's length over somebody's kitchen bench.

THE SPINE

One purchase reported twice and counted once. The browser pixel is JavaScript in the buyer browser and sees on-site actions your backend never sees, but it loses blocked scripts, tracking prevention, a tab closed early and a session that moved. The server stream sends the same events from your own backend and survives blockers, but it loses anything the backend never learns about. Both reports carry the same event name and the same event ID, which is what deduplicates them, so the platform counts one. Without the shared ID it counts two. No count or rate is shown.
The same two streams as Meta, deduplicated by the same shared event ID.
  1. Business Center under your own login. Two-factor authentication, second admin.
  2. Ad account inside it. Prepaid balance or card billed afterwards — find out now.
  3. Set the identity the ads run under: the name a stranger sees.
  4. Install the pixel and confirm events fire in order.
  5. Add the server-side stream, deduplicated with a shared event ID.
  6. One test order. Confirm one purchase is recorded, not two.
  7. Verify your website.
  8. For Spark Ads, secure the post first, with an expiry you wrote down.
Three tiers of number ranked by how much you can trust them. Measured: money that landed in the bank, orders in your admin, invoices you paid. Modelled: platform-attributed conversions, anything with a lookback window, blended estimates. Unknown: what you assumed, what you were told, what you hope. This is a rank, not a score.
Three tiers of number, ranked by how much you can trust them.

The pixel under-reports here for the reasons it under-reports on Meta, and the server stream fills the same gaps. What is platform-specific is the frame: attribution defaults differ, so two dashboards can be counting the same day over different windows.

NEITHER DASHBOARD IS YOUR BANK.

settle it on your own blended revenue and spend

One purchase, counted by two different ledgers. Two things decide whether the platform records it. First the window: how long after the ad a purchase still counts, so a purchase made inside the window is counted and one made after it closed is not. Second the device: whether the half that saw the ad and the half that bought can be joined up, so a purchase on the same device is joined up and one on another device may or may not be. The platform counts events; your bank counts money. No figures appear.
One purchase, two ledgers, and two things that decide whether it counts.

SPARK ADS

A real post, running as an ad

What it gives you

  • A real handle, and a profile tap that leads somewhere real.
  • Comments, likes and sound stay with it, so engagement accumulates.
  • It reads as the platform, not as an interruption.

What you depend on

  • An authorisation that expires, on an account staying in good standing.
  • Terms you agreed verbally, if you agreed them verbally.
  • So write it down: what, from which account, which markets, until when.
The seven things a usage agreement has to name before anyone films. Term: how long the footage may be used. Territory: which countries it may run in. Channels: paid, organic, email, the page itself. Exclusivity: whether they may also film for a rival. Whitelisting: whether ads may run from their handle. Revisions: how many redos are included, and when. Disclosure: who marks it as an ad, and how. Every box is drawn open, because the values are yours to agree and write in.
The seven things the authorisation has to name, before anyone films.
Statement poster: "It should be working is not a metric."

WHAT ACTUALLY DIFFERS

Where the two feeds diverge

What differsMetaTikTok
Starting contextA relationship carries the ad.A stranger. Every ad starts at zero.
FormatA clean product shot can carry it.Vertical, fast, phone-shot. Polish reads as an advert.
SoundAssume many viewers are muted.The audio frequently carries the hook.
CommentsA channel to moderate.Part of the ad, under what you paid for.

Meta is a feed of chosen relationships. TikTok is a recommendation engine of strangers, so the creative does the selecting.

Targeting drawn as a ladder from broad to narrow: no definition beyond the basics, one interest or behaviour, several stacked, modelled on your own customer list, and retargeting people who already touched the funnel. As you move down, the pool gets smaller, frequency builds faster on the same spend, selection moves from the creative to the targeting box, your own ad sets begin overlapping, and a win says less about the hook. Neither end is the correct one. The bar widths are illustrative and no pool size or cost is shown.
Narrow it and the settings do the selecting. Leave it broad and the hook does.
The first three seconds of a video ad on a vertical time axis. Zero to one second: stop the right person, not everyone; it fails when it stops everybody and you pay for that. One to two seconds: earn the next second by making staying obviously worth it; it fails when the ad slows down to introduce itself. Two to three seconds: the same job continues; it fails when the stop never connects to the product. Everything after the three seconds is the argument: context, mechanism, objection, proof, offer and one call to action.
The first three seconds, and what each one has to do.
  • One naming convention, with the tested variable in the ad name. In six weeks the names are the only record of the question.
  • One event taxonomy, or every cross-platform comparison afterwards is fiction.
  • Measure how long your own assets last here rather than assuming Meta lifespans.
  • The rented account market exists here too, on the same trade. Own the Business Center.
The shape of a creative dying and the order the numbers move in. Frequency climbs first. Link click-through drifts down. Cost per result drifts up. On-site conversion rate stays unchanged, which is the tell: the page is fine and the audience has simply seen it. Below, three patterns that are not fatigue: weak from day one with no peak to decay from is a bad ad; every asset moving at once is the auction or the season; clicks arriving as normal with add to cart collapsing is the page, the price or the offer. The traces show direction only and neither axis carries a scale.
The shape you are measuring when you ask how long an asset lasts here.

TAKEAWAYS

  • The infrastructure ports between platforms almost unchanged. The creative habits do not.
  • Meta is a feed of chosen relationships. TikTok is a recommendation feed of strangers.
  • Spark Ads keep the handle and the comments, and depend on an expiring authorisation.
  • One naming convention and one event taxonomy across both, or comparison is guesswork.

CHECK YOURSELF

Why does a re-cropped Meta ad tend to underperform on TikTok?

Two platforms report a different number of purchases for the same day. What should you check first?

DO THIS BEFORE THE NEXT MODULE

  1. 01Find out which billing model your market uses before anything is scheduled.
  2. 02Write the attribution settings of every platform you run side by side on one page.
  3. 03Watch fifty ads in your category and write down the first line of the ten that stopped you.
Enrol free