Hook Rate in Meta Ads: Diagnose Attention Without Ignoring Sales

For this worksheet, hook rate means three-second video views divided by impressions. Confirm the exact export definitions, including repeats and short-video handling, before comparing rows. Attention is a diagnostic, not proof that the ad acquires qualified customers; a viewer may click without watching three seconds.

Seven openings for one truthful offer

Hypothetical product brief: a backpack with dimensions and contents that the seller must verify before publication. Keep the remainder of the video, destination and offer the same. None of these copy examples claims a customer result.

Compare allocations before ranking openings

1. Declare net paid-order CPA or qualified-lead CPA as the primary outcome and hook rate as a secondary diagnostic.
2. Keep body, CTA, offer, event and eligibility comparable. Record length/placement differences.
3. Seven ads in one ad set need not receive equal delivery. Use an available controlled design or test fewer prespecified contrasts when cash cannot support seven independent arms.
4. Log spend, impressions, view counts, mature qualified outcomes and rejected/duplicate orders for every version. An underdelivered arm is not automatically a loser.
5. Approve spend and cash-loss limits, expected conversion maturity and useful effect size before running. An impression count alone does not establish sufficient business evidence.
6. At the cap/deadline report inconclusive if outcomes or assignment are insufficient. Repeatedly choosing the largest attention percentage among seven arms can select noise.

Low attention can coexist with useful customers

Synthetic arithmetic: opening A has 200 three-second views from 1,000 impressions (20%) and five mature qualified orders from $100 spend (CPA $20). Opening B has 400/1,000 views (40%) and two orders from $100 (CPA $50). These are invented inputs and calculations, not a measured winner. They show why attention alone cannot decide; order quality and uncertainty still need review.

A drop in attention is a hypothesis about relevance or delivery, not proof that the opening caused poor sales. Check the offer, page and qualified outcomes before replacing it.

Prepare and approve the variant

In NovaStorm, write the hook hypothesis and prepare a paused creative/campaign draft for user review. Keep approved version IDs and evidence for product claims; do not imply the product randomizes the test or reports hook metrics. The creative testing protocol supplies the next planning step.