In 2010 Comscore acquired Meg Blair’s RSC, a gem of a copy testing company, which Comscore later divested. Soon after the acquisition they developed a market mix model (MMM) based on 35,000 copy tests showing that creative was 80% of sales effect.
However, a year earlier, TRA and Mars presented results at ARF showing that the creative percentage of sales effect was 65%. Apollo and NCS also found it to be 65% in that same time period. These studies all used singlesource – based on panels for Apollo and Mars, based on big data for TRA, and based on both for NCS. Singlesource is household-level MMM. Both methods traditionally use “tonnage” i.e. ad spend or its proxy GRPs (and in rare instances nowadays reach/frequency). Quantitative data, having no human measures of motivation or creative.
However, the right number might actually have been 80%, to the extent that human factors might play any role in sales effect.
RSC’s data were data about creative, persuasion scores, and the like.
MMM runs on media GRP/dollars. The singlesource studies used multiple regression analysis similar to MMM however at the household level rather than at aggregate levels. Still, the singlesource estimates of the creative contribution to sales utilized dollar spend rather than creative parameters.
Use of data about creative and use of data about media dollars are two variations of the same method. One might be more accurate than the other. The two combined together might be more powerful than the media-only MMM/singlesource, without creative.
Is there any evidence of that? Yes, the ARF Cognition Council study found that the creative, as quantified through RMT’s AI lens, accounts for 48% of sales effect. MMM based on media dollars generally finds that all of advertising – both the creative and the media – account for only about 7% of sales effect. The ARF study using just the creative data based on RMT methods showed that advertising’s creative effect on sales was about 7X advertising’s media effect. It’s possible that MMM or singlesource using both media dollars (or TRP or reach/frequency) plus RMT creative parameters (essentially the dimensions of motivation) would show that advertising drives 55% of sales effect (48% creative and 7% media).
However, creative is only one part of advertising, so if creative is 48% of sales, and media, promotion, and all other factors the other 52%, then marketing accounts for 100% of sales, which is impossible, because habitual behavior is so strong, and looked at on a very long range basis, way back in the beginning, marketing, along with product experience, did create those habits. So we might say that ultimately marketing including word of mouth does drive 100% of trial sales, although when habits form, their having been originally formed by marketing and product experience drops away as perceived by advertiser C suites.
So what is the fair answer to the question of creative percentage of sales?
Soon, someone will answer that question by being the first MMM and/or singlesource to utilize both the usual plus the RMT factors. Or some other data representing the creative, or several sets. This method will then become the standard.
In the meantime, we notice that the creative percent of sales effects is trending down.
Here are the three data points from which I draw that potential conclusion. They all use similar econometric modeling based on the dollars (or GRPs) allocated by media types.
In 2009 or thereabouts, four companies using singlesource agreed on 65% (Mars, TRA, Apollo, NCS).
In 2017, using panel plus big data singlesource, NCS reported 47%.
In 2023, NCS using panel plus big data singlesource, found 39% for TV and 26% for digital.
Why would the power of the creative be going down?
It would not seem to be about the quality of the creative, but more about the changes in the media environment and in the audience.
The audience working longer hours than their parents and soured out. Levels of skepticism and cynicism greatly increased, brand trust in many cases shrinking away, more pessimism than optimism.
A main driving factor I feel is that digital is the next generation of promotion. Advertising increases value perception but promotion is focused on triggering immediate sales generally utilizing reminder and price effects.
As documented here, the digital standard 1-2 seconds of attention can accelerate sales among existing customers, but not win new hearts and minds. Jim Donius foretold it five years ago: digital is actually part of the promotion budget or should be. If only 26% of its sales effect comes from creative, it must be more like promotion and less like advertising.
As more dollars shifted (plus new dollars) the emphasis of the new lower-cost mass medium, digital, was on short-term sales performance.
The biggest winners being walled gardens giving themselves report cards on everything from audience sizes to sales effects and everything in between. Most often their self-studies did not include TV or other non-digital media, so any sales produced by TV was attributed to some form of digital. Because of the shininess of the new object, its fast device level data reporting, and especially those lovely low CPMs, not to mention what one’s kids advise their marketer parents, many C suites have hypnotized themselves into believing that they can cut TV – “no one watches it any more except poor old people” – with impunity, as long as they continue to pay for digital shelf space in amounts dictated by the big walled gardens.
Now groupthink herd mentality has convinced marketers that they can cut marketing year after year and simply make up for it with greater efficiency – i.e. those attractively low CPMs in digital.
In parallel to that, TRA’s introduction of big data cross-platform ROI systems 2005-2014 had the same effect on increasing C suite focus on short-term performance once the new systems brought a greater sense of control. You actually could see a glimmer of cause and effect. Everything suddenly felt like direct response. A new level. That has translated today into programmatic.
Quarterly earnings reports and management compensation are of course linked.
Everything points to the wisdom and appropriateness of focusing on the short-term present.
The only flaw in the ointment is the pressure to grow brands.
Brand thinking is longer term in its vision. The apparent debate between branding and sales for a short time appeared to be intelligently converging on an emphasis on both, in an integrated way, without silos separating them. That illusion has fast faded in the U.S. Perhaps in other countries there is still a respect for branding effects but all that has been flushed away here.
To grow brands, you need to sustain attention considerably longer than most digital. The arrow is back toward the spurned bride of television. But few if any marketers have noticed this so far.
It’s best to not starve a symbiote. If something grows along with your growth, that is a signal that it might be a symbiote. For many years TV and brands had a grand symbiotic relationship which enabled them both to grow together.
The FOX BHC ROAS study based on MMM methodology has found that since 2014, fewer than one in five brands studied have gained market share.
Starving TV has hurt the buy side even more than the sell side. As shown in the FOX BHC ROAS study, for example, if consumer electronics and technology marketers in the U.S. utilized the optimal allocation by media types – involving a shift back to TV and a reduction in digital ad spend – the U.S. incremental sales made in the 12 months ending September 2023 would have increased by a factor of 2.6X. Across Automotive, CPG, and Consumer Electronics/Technology, the additional U.S. sales that would have been made in that same 12-month period with its optimal allocation shifting some digital dollars to television, would have been $55 billion. Over the period 2014-2024, the cumulative lost sales dollars across all verticals is roughly estimated to be in the neighborhood of nearly a trillion dollars, caused by overinvesting in short-duration-attention digital at the expense of under-investing in the world’s most powerful in-home advertising medium, as measured objectively by ROAS.
A Nielsen report of industry changes based on 150,000 observations notes that half of all brands studied are underspent and only 25% are overspent – half of all brands studied not achieving the ROI they could if they spent more.
Takeaway: Test hypotheses using random control trials, singlesource, and MMM, instead of rolling the dice based on herd mentality. Use digital too, but consider it mostly as part of promotion (branded content and influencers are more like advertising than like promotion), and do not cut advertising that is working. Measure the effects of cutting advertising without having to test the whole carpet. Measure the effects of increasing advertising too. Open your minds to evidence-based decision making and resist going along with popular street wisdom black-and-white proclamations.
Posted at MediaVillage through the Thought Leadership self-publishing platform.
Click the social buttons to share this story with colleagues and friends.
The opinions expressed here are the author's views and do not necessarily represent the views of