Is it my creative or my targeting?

The split the question assumes no longer exists

There was a time when targeting was a separate job. You picked interests and behaviours, and the creative spoke to whoever you had picked.

Under broad targeting, you have handed audience selection to the algorithm. What you still control is the signal you give it — and that signal is the ad. The desire it names and the person it describes are what determine who Meta decides this ad is for.

Which means creative and targeting have collapsed into the same thing. The ad *is* the targeting. Asking which one failed is asking which half of one object broke.

The question that replaces it

Every angle is built from three layers:

1. The core desire — the primal drive behind the purchase, not the surface want.
2. The shared belief system — the worldview that shapes how this group evaluates products.
3. The specific avatar — the person who lives in that pocket.

When something fails, one of those three was wrong. Finding out which is a far more useful afternoon than arguing about creative quality, and it does not require scrapping the concept.

Reading the funnel to find the failing layer

Compare each new angle against your own account baseline. An angle is working if it holds on at least three of four:

• CPM within 10–15% of baseline
• Outbound CTR at or above baseline
• Click quality — landing page views divided by link clicks — at or above baseline
• ATC rate — add-to-carts divided by landing page views — at or above baseline

The pattern of which ones fail is the diagnosis.

Strong CTR, weak landing page views. Curiosity clicks. The promise was interesting enough to click and too vague to act on, or the landing page does not continue the sentence the ad started. Fix the promise and the page before you kill the angle — this is the single most common case of a good angle being thrown away.

Landing page views fine, weak add-to-carts. The message brought the right people and then failed to convince them. Usually the desire is right and the after-state is wrong: you described a problem they recognise and a resolution they do not want.

CPM and CTR both below baseline. The pocket does not exist in meaningful volume, or nothing in the ad resonated with anyone. Kill it. This is the one clean kill signal.

Everything at baseline but no purchases in the first window. Normal, and frequently mishandled. People in a genuinely new pocket are learning that you exist. Judging them against warm retargeting is judging a first conversation against a fifth. Check whether blended MER, retargeting conversion rate or overall conversions improved while the angle was live — new top-of-funnel angles feed the retargeting pool before they show up as direct sales.

Why "make it more relatable" is not a brief

A vague brief produces variations. Variations are not tests, because when one does better than another you cannot say why.

A three-layer brief hands the strategist a formula: this desire, this belief system, this person. Handoff becomes mechanical. Results become legible. And when an angle wins, you can generate five more by changing exactly one layer, which is how one winner becomes a set instead of a lucky asset nobody can reproduce.

The one case where it really is execution

Test new angles as statics first. Not because video does not work — it does. Because statics strip out everything that is not the positioning: no retention editing, no trending audio, no creator charisma.

If an angle works as a static, it usually works harder as video.

If it only works as video, you do not have a winning angle. You have a winning performance. That distinction matters because performances cannot be scaled or handed to someone else, and it explains why some accounts have "winners" that were never successfully rebuilt.