Hook Rate And IPM: What Account Reporting Cannot Tell You

The short version

Campaign-level reporting tells you the account is underperforming. It cannot tell you which creative caused it, and it cannot tell you whether the problem is that people did not watch or that they watched and did not buy. Two creative-level metrics separate those cases, and most teams never look at them.

Two different questions

Account reporting answers: is this working?

Creative reporting answers: which specific assets are responsible, and how are they failing?

Those are different investigations, and running the second one on account-level data is how teams end up refreshing creative that was fine and leaving the actual problem in place.

The metrics that only exist at creative level

Most creative comparison uses the same columns as account reporting: spend, CPC, CPM, CPA, conversions, revenue, ROAS, CVR, impressions, clicks, CTR. Useful, but they describe outcomes without describing why the outcome happened.

Two columns change the diagnosis.

Hook rate reads early retention. It answers whether anyone stayed past the opening. This is where most video creative actually fails, and it fails silently, because a video nobody watched still produces a full set of downstream metrics that look like a conversion problem.

IPM reads installs per thousand impressions. It is the standard creative efficiency measure in mobile user acquisition, and it isolates creative pull from bidding and budget effects in a way CPA does not.

Why hook rate reorders the diagnosis

Consider a video ad with a weak conversion rate. The obvious reading is that the offer or the landing page is not converting.

Now add hook rate. If it is poor, almost nobody saw the ad's argument. The conversion rate is not measuring your offer; it is measuring the small self-selected group who stayed through a weak opening. Fixing the landing page will not move it. Fixing the first two seconds might.

This is the specific reason to start at hook rate for video rather than at the outcome column. A weak hook and a weak offer produce similar-looking CPAs and require completely different responses.

Sorting by the decision, not by habit

Most people sort by spend because it is the default. Sort by the question instead:

Deciding what to scale. Sort by the outcome metric that matters to the business, then check whether the sample is large enough to believe before acting on it.
Deciding what to retire. Sort by spend and look for high spend against weak outcomes. This is the one case where spend-first is right.
Diagnosing video. Start at hook rate. Answer the retention question before the conversion question.

The comparison mistakes that produce confident wrong answers

Comparing creatives with very different spend. A creative with 40 conversions and one with 3 are not two data points of equal weight, and ranking them in the same table invites treating them that way.
Judging too early. A creative that has not accumulated enough impressions has a metric, not a result.
Mixing objectives in one comparison. Creatives optimized toward different events are not comparable on the same outcome column, regardless of what the table shows.
Mixing date windows. Two creatives that ran in different weeks were competing in different auctions.

Filter before comparing. The narrowing is not a convenience feature; it is what makes the comparison valid.

Practical rule

Use account reporting to find out that something is wrong, and creative reporting to find out what. For video, read hook rate before you read anything downstream of it.