Why creative is your largest working media lever

Creative is the largest working media lever most brands have, because on auction-priced media the work sets the price of every impression and decides what each one converts into. For decades the media plan came pre-divided into two categories. Working spend bought airtime and impressions, the delivery of the message to a person. Non-working spend was everything else, production, fees, the cost of making the thing. Procurement's job was to push the working share up, and the logic was sound, because when the price of an impression was fixed in a negotiation months before the ad ran, creative could not change what a rating point cost. That split made sense for fifty years. On paid social it has inverted, and most budget conversations have not caught up.

How the auction prices the work

Every impression on a social platform is sold in a real-time auction, and the auction does not clear on the bid alone. Meta describes each ad as ranked on a total value made of three factors, the bid, the estimated action rate and ad quality, and it is worth taking the three apart because two of them belong to the creative.

The bid is what the advertiser is willing to pay for the outcome the campaign is optimised for, and under the automated bidding most accounts run, the platform sets it on the advertiser's behalf to spend the budget. It is the only one of the three the media plan controls directly. The estimated action rate is the platform's prediction, for this person and this ad in this auction, that the person will take the optimised action, watch the video through, click, convert, and the system learns that prediction from how people have responded to the ad so far. It is a property of the work meeting an audience. Ad quality is drawn from how people react to the ad itself, whether they hide it, report it or respond to it, and whether it carries the attributes the platform's own guidance treats as low quality. It is a property of the work alone.

Put two ads from two brands into the same auction for the same person with the same bid, and the one the system predicts this person is likelier to watch carries the higher total value and wins the impression. Run that across the millions of auctions a campaign enters in a week and the effect shows up in the account as a lower CPM for the same delivery, or more delivery for the same budget, without anyone negotiating anything. The platforms are not selling space, they are selling their users' time, and their business depends on people staying in the feed, so an ad people watch, finish and respond to wins impressions it would otherwise have lost and wins them at a lower effective price. An ad people skip pays more for what it still gets, and the difference goes to whichever competitor the auction preferred.

This is the mechanical fact that collapses the split. Two brands with identical budgets, identical audience settings and identical flight dates will pay different CPMs, reach different numbers of people and earn different amounts of attention, and the only variable separating them is the work. Meta's test data, from the platform that sells the media, puts numbers on the spread. Across 235 brand lift tests the winning creative delivered a 13% lower cost per ad recall than the average of the other creatives in the same test. Across 14,780 A/B tests it delivered a 21% lower cost per website conversion and 27% lower cost per app install. Same media, same audiences, different work. The direction of the CPM effect depends on the objective, downward under reach, and under conversions often upward with a lower cost per outcome, which How to align your media and creative strategy works through.

One measured account

The effect compounds when the loop is running. In one measured account, same budget, same media plan and targeting, across three weekly cycles, the brand's existing creative was buying a completed view at $0.50. The portfolio built for the feed and optimised weekly on the same plan beside it reached $0.09 per completed view by the third week, with five times the completion rate of the existing creative. One account, so it is an illustration rather than a law, and part of the gain is the loop retiring the weakest work each week and extending the strongest, so the figure describes a managed portfolio and not a single ad. But the decomposition is the point. Most of the movement is the completion rate itself, more of every thousand impressions becoming watched ads, and the remainder is the auction delivering the stronger signals a little cheaper. Both are creative levers. The media buying never changed.

A large global CPG company showed the second lever on its own in another account, a 48% lower CPM against the brand's existing benchmark on the same media plan, before counting a materially higher completed view rate on what was delivered. The media agency did not negotiate that discount. The creative earned it, impression by impression.

Creative's share of the outcome

The industry has spent a decade arguing about how much of an ad's effect belongs to the creative and how much to the media, and most of the figures in circulation cite each other. One has a primary source. NCSolutions' 2023 analysis of nearly 450 CPG campaigns attributes 49% of sales lift to the creative, ahead of brand, targeting, reach and recency, consistent with its 2017 finding. NCSolutions sells sales-effect measurement and the figure is a decomposition of sales lift, not a split of creative against media, so read it as a statement about which lever moved the outcome most, not as a budget ratio. Read that way it says what the auction says. The work is where the variance is.

The account-level version is a lift study with the media held constant. Whiskas ran a portfolio built for the feed against the brand's existing creative, same brief, same media, and the portfolio delivered five times the brand favourability lift of the existing creative in brand lift testing. Nothing about the plan changed. The outcome moved with the work, which is the only place it could have moved from.

The lever comparison

Suppose you want more watched reach next quarter and you have two levers. Watched impressions are the budget divided by the CPM, times a thousand, times the watched rate. Three terms. The media lever moves the first. A 20% increase in budget buys at best 20% more impressions at your current prices, and in practice fewer, because reach curves are concave and the marginal impression costs more than the average one, and every one of those impressions converts into a watched view at the same rate as before. If your creative is being skipped you are buying 20% more skips at rising prices.

The creative lever moves the other two terms, and it moves them on the whole budget rather than the increment. Better work raises the watched rate, so more of every thousand impressions already being bought becomes a watched ad, and the auction rewards the same work with a lower CPM, so each thousand costs less. The two effects multiply. In the measured account above, the same media budget delivered more than five times as many completed views as the brand's existing creative, across the full spend, and no plausible media increase produces that, because no finance function approves a budget increase of several hundred percent to buy the same ad more skips. The cheapest watched reach available to most brands is in the creative line, because the operating model they inherited was built when the levers were the other way around. Kantar's cross-media database for Meta, 1,001 campaigns and platform-commissioned so read with that interest, indexes Meta's cost effectiveness on brand motivation at 103% above the average channel, and the deck that carries the figure spends its final chapter arguing that creative is the number one driver of the attention that produces it.

Where the fee sits

The classification question comes up in every procurement conversation. Where does a creative fee sit, working or non-working? The direct answer is that the question has stopped meaning what it meant. When creative sets the price of media, a fee that improves the creative is a fee that lowers the CPM and raises the watched rate on the whole media budget, which is what working media was always supposed to describe. A fee quoted on the media plan, which is how a creative system prices itself, sits inside the media line by construction and scales with it, and the test of whether it is working is the marketing efficiency number in How to read paid social results, the all-in cost per watched impression, media plus fee over watched impressions, against the brand's prior creative on media spend alone. It is the working versus non-working question answered in one number, it is the one card that can come out against the fee, which is exactly why it belongs in the standard, and How to work out what your creative really costs runs the arithmetic.

None of this repeals the oldest law in media planning. Les Binet and Peter Field's excess share of voice relationship still holds. What the auction changed is what produces voice. Voice was a function of spend at a common price. It is now spend multiplied by creative efficiency, and a brand whose creative earns half the CPM of its category has doubled its voice budget without a finance conversation, while a brand whose work gets skipped pays more for every point of voice and donates the difference to whichever competitor the auction prefers.

The split, defended

The counter-position is procurement's and it is not unreasonable. The split gave finance a clean way to hold agencies to account, and on negotiated media it was correct. Its defenders will say a fee is still a fee and an impression is still an impression, and that blurring the line invites every vendor to call its costs working. The answer is not to blur the line but to move it to where the auction has put it. An impression's price is now set partly by the work, so the work is inside the price, and the accountability procurement wants is better served by asking what price the creative is earning in the auction than what percentage of the budget it consumed. Media creates the opportunity to see your ads. Creative determines whether they work. On this medium the work also determines what the opportunity costs.

What to do with this

Export last quarter's ads from Ads Manager under one objective and one platform, with impressions, spend and ThruPlays for each. Compute two numbers per ad, CPM and cost per thousand watched impressions, which is CPM divided by watched rate. Rank by the second. The gap between the best and the worst ad is the size of the lever, and it was there under one media plan with the same team buying it. Then look at where the spend went, because if most of the budget sat behind ads in the bottom half of the ranking, the media was bought well and the work was setting its price badly. Take the comparison to the next budget conversation as the alternative to the increment, and ask what a 20% media increase would have delivered against moving the account's watched rate to the level its best ad already reached. A lever this large needs someone moving it every week, which is the argument of Why creative needs managing as actively as media.

Questions

Our media agency already negotiates strong CPMs, how does creative change that?

On auction-priced media the CPM is not negotiated, it is earned in every impression, and the auction ranks ads partly on the predicted response to the ad and on its quality. A strong media setup is a necessary condition and the agency's remit is unchanged, but the price it achieves is set in part by the work it is given, so two accounts with the same setup and different creative pay different CPMs and reach different numbers of people. The media team sets the ceiling, the creative decides how far below it the account runs.

Is a creative fee working or non-working media?

The question was built for negotiated media, where the work could not change the price of an impression. On social it can, so a fee that raises the watched rate and lowers the CPM across the whole media budget is doing what working media was defined to do. The honest test is the all-in cost per watched impression, media plus fee, against the brand's prior creative on media alone, and that number can come out either way, which is why it belongs in the reporting rather than in the pitch.

Wouldn't more media budget get us the reach faster than new creative?

More budget buys more impressions at the same watched rate and, on a concave reach curve, at a rising price, so the increment delivers less than its face value in watched reach. Better creative changes what every impression already being bought converts into and lowers what each one costs, and it does so across the full spend. The media lever is still real and weight still matters, but once the account's own data shows a wide spread between its best and worst ad under one plan, the creative lever is the larger of the two.

Sources

• Meta Business Help Centre, "About ad auctions". Ranking on total value from the bid, the estimated action rate and ad quality. Platform documentation, Meta sells the media.
• Meta, "Creative and Media Excellence", circa 2023. Winning creative against the average of other creatives, 235 brand lift tests (April 2021 to 2022) and 14,780 A/B tests (April 2021 to April 2022). Meta internal test data, platform interest.
• NCSolutions, "Five Keys to Advertising Effectiveness", August 2023 update of the 2017 study. Creative at 49% of sales lift across nearly 450 CPG campaigns on TV and digital, a decomposition of sales lift. NCSolutions sells sales-effect measurement to CPG advertisers and media owners.
• Les Binet and Peter Field, The Long and the Short of It, IPA, 2013, and Media in Focus, IPA, 2017. Excess share of voice and share growth. IPA Databank of agency-submitted cases.
• Kantar LIFT+ cross-media database, 2019 to 2024, 1,001 campaigns, as presented in Meta, "Demystifying attention", 2025. Meta's indexed cost effectiveness on brand motivation. Kantar sells the database, Meta commissioned the cut, modelled.
• Whiskas, brand lift testing, brand favourability at five times the brand's existing creative, same brief, same media. Cleared proof point from our own work.
• One measured account, anonymised. Cost per completed view from $0.50 to $0.09, same budget, same media plan and targeting, across three weekly cycles, with the weekly loop retiring weak work as part of the gain.
• A large global CPG company, anonymised. CPM 48% below the brand's existing benchmark on the same media plan.