Meta catalog video: the ad spend per SKU that pays for the clip

A catalog product video pays for itself on a SKU once that SKU's ad spend reaches about five times the clip cost, scaled by your cost per acquisition (CPA) over your margin per order. The formula comes from one number on Meta's catalog ads page, read 2026-10-05: "Ad sets with catalog product video saw 20% more conversions per dollar." Meta does not promise that result for your account, so treat the table below as a threshold to test, not a forecast.

Where the 5 comes from

If a SKU's ad spend is X and CPA is C, it earns X / C conversions. A 20% gain per dollar adds 0.2 x X / C conversions. Each conversion is worth your margin per order, m, so the gain is 0.2 x X x m / C. The clip pays for itself when that gain reaches its cost V, which gives X = 5 x V x C / m.

Meta lists the setup needs: a synced catalog and the Meta Pixel or Conversions API. Without them there is no catalog ad to attach the video to, and no measurement of the gain.

Break-even spend at two Sume clip costs

Clip costs use Sume prices read 2026-10-05: a 5 s, 720p Wan 3.0 clip at $0.625 plus $0.20 captions and $0.10 Timeline is $0.925; Seedance 2 at $1.89 gives $2.19. Rows vary CPA relative to margin, since a SKU whose CPA is half its margin recovers the cost on far less spend.

How to use the threshold

Only SKUs that already carry ad spend above their row qualify. A SKU spending $4 a month with CPA equal to margin will not repay a $2.19 clip at a 20% gain, so give it a still image or a cheaper clip. A SKU spending $200 repays any of these rows many times over if the gain holds.

Because the 20% is a ceiling you have not yet measured, the safe approach is a test: attach video to half of a SKU group, keep the rest on images, and compare conversions per dollar after a full budget cycle.

• Rank candidates by spend first, then by margin; see the profit-ranking script.
• Count retakes in V: a clip you regenerate once costs double.
• Re-run the math if the gain you measure is 10% rather than 20%: the break-even spend doubles.

For making a clip for every SKU in a feed, see product video for every catalog SKU as a bulk run, and for the full cost of a ten-SKU test see the 10-SKU pilot.