How to work out what your creative really costs

A creative fee is the wrong number to argue about, because the cost of creative on paid social is what each watched impression costs once the fee is inside the number, set against what the brand's prior creative was already paying for a watched impression on media alone. That comparison has a name in How to read paid social results, marketing efficiency, and it is the one line in the standard that can come out against the fee, which is why a procurement team should ask for it.

Where the split came from

For as long as media was negotiated, the plan came pre-divided. Working spend bought the impressions, non-working spend was production and fees, and procurement's job was to push the working share up, which was correct, because a rating point cost what it cost whatever ran in it. Why creative is your largest working media lever makes the case for what changed, and the short version is that on an auction every impression is priced partly by the ad itself, Meta ranks each ad on bid, estimated action rate and ad quality, two of which are properties of the creative, so an ad people watch wins impressions cheaper. The work is inside the price. The split did not become wrong, its line moved.

The test

The marketing efficiency test takes two numbers. The baseline is the brand's prior creative on media alone, media spend divided by the watched impressions it produced, ThruPlays on Meta or focused views on TikTok, one platform, one threshold, the objective stated. The comparison is the new creative with everything it cost, media spend plus the fee, divided by the watched impressions it produced on the same media plan under the same objective. Media efficiency, the CPM, stays on the media team's scorecard. Marketing efficiency is the all-in figure, and it answers the procurement question directly, did the whole arrangement deliver a watched impression for less than the old arrangement did without any fee at all. Whether the brand moved stays in the effectiveness tier and is read from a lift study on its own clock, because a watched impression is not a memory, but on a medium where continuing to watch is a choice it is the unit a brand effect needs to be possible, and its all-in cost is the honest price of the work.

A worked example

The numbers are illustrative, chosen for easy arithmetic, not drawn from any account. Take a brand paying a CPM of $10 under a reach objective. Its existing creative is watched, to a ThruPlay, on four impressions in every hundred. A thousand impressions cost $10 and produce forty watched impressions, so a watched impression costs 25 cents on media alone. That is the baseline.

Now suppose the new creative, on the same media plan at the same CPM, is watched on six impressions in every hundred, a rise of half. The same thousand impressions produce sixty watched impressions and the media-only cost per watched impression falls to about 17 cents. Then the fee goes in. Call the fee f, a share of the media it is quoted on. The all-in cost of those thousand impressions is then $10 multiplied by one plus f, and the all-in cost per watched impression is that figure divided by sixty.

The arrangement beats the baseline when $10 × (1 + f) ÷ 60 comes in under 25 cents, which is when one plus f is under one and a half, which is when f is under one half. In words, a creative that lifts the watched rate by half supports a fee of up to half the media cost before the all-in cost climbs back to the baseline. That threshold is the example's, not any fee's, and it moves with the watched rates.

The break-even rule

Strip the example away and the rule is one line. A fee is justified when the watched rate rises by a larger proportion than the fee adds to the media cost. Up by a fifth supports a fee of up to a fifth of media, up by half supports up to half, doubled supports up to the media cost itself, and any rise smaller than the fee's share means the arrangement delivered watched impressions dearer than the old creative did. A fee quoted on the media plan scales with the media, so f is the same at any budget and so is the rule. The rule is also conservative. It holds the CPM constant, and under reach and awareness objectives creative that gets watched more tends to clear cheaper, so if the CPM also falls the test gets easier and the fee does not need the help.

The split, from procurement's chair

The counter-position is procurement's and it is a good one. The split gave finance a clean, comparable line, it stopped every supplier from calling its costs working, it travelled across years and agencies, and on negotiated media it was exactly right. Its defenders will add that the test trusts a platform's own reporting, that a baseline can be chosen flatteringly, and that a six-second ad will post a higher watched rate than a thirty-second one without anyone being better at anything. All three are fair, and the standard's rules answer them. The watched rate comes from the brand's own account, not the supplier's deck. The baseline is the brand's prior creative in the same account, on the same platform, threshold and objective, stated on the card. Asset lengths are matched, or the comparison says so. What the test keeps from the old split is a number the supplier cannot talk its way around, and what it changes is the unit, from the share of budget the creative consumed to the price the creative earned.

What to do with this

Pull last quarter from Ads Manager for one platform and compute the baseline before any conversation about fees, media spend divided by ThruPlays on Meta or focused views on TikTok, with the objective written beside it. The number belongs to the brand, and every proposal is judged against it. Then write the fee into any proposal as a share of the media it is quoted on and ask by what proportion the watched rate has to rise for the all-in cost per watched impression to beat the baseline, and whether that rise is plausible given the spread between the best and worst ads already in the account. Ask for the test monthly with the fee in the numerator, including the months it comes out against the fee, because a supplier willing to show that number is the one worth keeping. Media creates the opportunity to see your ads. Creative determines whether they work. How to audit your own paid social account pulls the baseline and the spread from your own account in ten moves, and it is the natural next read.

Questions

Isn't a creative fee non-working spend by definition?

On negotiated media it was, because the price of an impression was fixed before the creative existed and nothing in the work could change what a rating point cost. On auction media the price of every impression is set partly by the ad, so a fee that improves the ad changes what the whole media budget buys, and the honest classification is to stop arguing about the label and run the test.

If the all-in cost per watched impression comes in under the brand's media-only baseline, the fee paid for itself out of media the brand was spending anyway. If it does not, the fee was non-working in the only sense that matters, and the number says so.

What baseline do we use if we have never measured watched rate?

You already have it. Any account that has run video on Meta carries impressions and ThruPlays for every ad, and TikTok carries focused views, so the baseline is last quarter's media spend divided by last quarter's watched impressions, one platform at a time. Where the account ran several objectives, compute the baseline under the one the new work will run under, because the CPM and the watched rate both move with the objective.

Where asset lengths varied, say so beside the number. A baseline with its conditions stated is worth more than a benchmark, because the benchmark rates circulating for hook and hold now mostly cite each other.

What happens to the test when the CPM moves?

Under reach and awareness objectives the CPM tends to fall as the creative gets watched more, because the auction rewards ads people respond to, and that makes the all-in figure better still. The rule in this essay holds the CPM constant so the fee is judged on the watched rate alone, which is the conservative reading.

Under conversion objectives the CPM tends to rise with stronger creative while the cost per outcome falls, so the test is read on cost per outcome instead. Either way the test is never read without the objective beside it.

Sources

• How to read paid social results, the measurement standard published in this section, September 2026. Marketing efficiency defined as all-in cost per watched impression, media plus fee, against the brand's prior creative on media alone, and the rules on baseline, objective, platform, threshold and matched asset lengths. Our own standard, we sell the creative it measures.
• Why creative is your largest working media lever, in this section. The working versus non-working split and the direction of the CPM effect by objective. Our own essay.
• Meta Business Help Centre, "About ad auctions" and video metric definitions (3-second video plays, ThruPlay). Ranking on bid, estimated action rate and ad quality. Platform documentation, Meta sells the media.
• TikTok Ads Manager Help Centre, "About the video views objective". The focused view definition. Platform documentation, TikTok sells the media.