How to apply brand science to social media

Brand science applies to paid social as written, with one change to the mechanics, and that change is the reason a creative system exists. Byron Sharp published How Brands Grow in 2010, Jenni Romaniuk's work on distinctive assets and category entry points has been in print since 2018, and the opportunity is in applying them to paid social, the largest media line for most consumer brands and the one still largely run on assumptions the science retired fifteen years ago. This essay does the application, then says exactly where we agree with Sharp and where the medium makes us differ.

Penetration over loyalty, and who follows you

Sharp's central result, from the Ehrenberg-Bass Institute, which is sponsored by large advertisers and sells research, is that brands grow mainly by being bought by more people, not by being bought more often by the people who already buy them. The buyer base of any brand is dominated by light buyers, the How Brands Grow illustration from UK panel data is that most Coca-Cola buyers bought it once or twice a year, and the heavy, loyal buyers every dashboard celebrates already buy you.

Hold that against the audiences a social account accumulates, followers, people who interact, retargeting pools, CRM lists. Every one is built by the same filter, prior interest, and light buyers mostly fail it, because the person who buys your category twice a year rarely follows anyone in it. The people in the warm circles were never the growth, they were the evidence of past growth. The target is the market, and most of it has never flagged itself to anyone.

Retargeting deserves a fair reading, its numbers are accurate. Warm audiences convert at multiples of any cold audience, the interpretation is the problem, because by construction the pools contain people who already moved toward the brand, and where incrementality is measured against a holdout the gap is usually sobering. It is loyalty marketing conducted through paid auctions at acquisition prices, and its real cost is the allocation it justifies, because every review where warm efficiency beats cold cost moves budget away from the light-buyer reach where penetration is built. Facebook's 2012 analysis with Datalogix across 50 CPG campaigns, dated and platform-funded, found more than 90% of people who saw an ad and then bought in store never clicked, and Binet and Field's brand effects, from the IPA's databank of agency-submitted cases, build slowly and last. The click measures the harvest, and a seven-day window held against an effect on that clock concludes the effect does not exist.

Distinctive beats different, in two seconds not two metres

Buyers of competing brands barely differ, a brand shares its customers with every other brand in the category roughly in proportion to those brands' sizes, so there is no secret segment that is yours by temperament, and what predicts performance is distinctiveness, being recognisable as yourself instantly, not being different in a way buyers must evaluate. The feed's viewer is Kahneman's fast mind, deciding watch or skip on recognition and feeling in under two seconds, so what gets through is whose this is, the colour, the character, the cue, plus a feeling, and where real product difference exists the feed carries it shown rather than argued.

Romaniuk's distinctive assets earned their place on the shelf, built to win the last two metres of the purchase. The feed is the harder venue. Nobody is shopping, the brand has perhaps two seconds, often with the sound off, and the asset is the mechanism by which an exposure gets attributed to the brand at all. A branded first second can refresh a memory that already exists, which is Sharp's reminding job and a real one, but it cannot build an association that is not there, and branding that arrives at second twelve is branding the median viewer never meets. Logos pass weakly because they need reading, colour bold enough to be a signature and characters that recur pass. The fame and uniqueness lens is unchanged, the bar moves, and an asset that needs three seconds of full attention is, for feed purposes, not yet an asset.

Category entry points are the targeting layer nobody briefs

Mental availability is the goal all of this serves, the propensity of the brand to come to mind in a buying situation, and Romaniuk's category entry points are the route to it. The targeting section of most social briefs is demographics and interests, and what almost none of them specify is the thing buying pivots on, the moment, the situation, feeling or occasion in which a person enters the category, the long shift ending, the guests arriving in an hour. It is a network of links between those entry points and the brand, and a brand linked to one entry point fires when that cue arrives. Linked to seven, it fires seven ways.

KitKat is the cleanest worked example, because the brand owns one of the great entry points in confectionery, the break, and the break is not one moment. The nurse pulling off scrubs after a ten-hour shift, a break that means decompression. The bar in the backpack at the summit pause, a break that means reward. The 3pm desk break, two minutes reclaimed between meetings. One brand, one distinctive asset system, three entirely different ads reaching three audiences through three lived moments, each encoding the same link. KitKat, a brand everyone already knew, saw two times the purchase intent of its existing creative in lift testing, the same brand saying the same thing through moments instead of at a demographic.

The delivery mechanics reward it. The platforms match each ad to the people who respond to it, Meta now describes creative diversification as the best lever for finding the most relevant audiences, so a shift-end ad finds shift workers and a trail ad finds hikers, and the entry point layer is a targeting layer, the one the settings menu does not contain. Briefing it is three questions in order. Which moments bring buyers into this category, listed concretely, and the first draft usually surprises the room. Which of those could this brand credibly own or share. Then one distinct creative hypothesis per chosen moment, with the entry point named in the brief so the weekly data can be read against it. Why you need a creative portfolio, not more content carries the delivery mechanics.

Where we agree with Sharp, and the one place the medium changes the mechanics

Many of the people we work with have How Brands Grow on the shelf and deserve to know where a creative system sits against it. We agree on the goal and on nearly all of the method: penetration over loyalty, reach over targeting because the light buyer is the growth and no interest filter finds her, continuity over bursts because memory decays, distinctive assets, mental availability and category entry points as the structures advertising builds, and scepticism of engagement metrics, which is where people assume we differ.

Engagement, likes, comments, shares, clicks, is an action a minority takes, and the minority skews to fans and heavy buyers. Ehrenberg-Bass is right to dismiss it, and we do the same. Watching is a different kind of thing. It is the absence of an action, not skipping, and it is measured across every impression served, including the light buyers who never follow or tap anything. A watched second with the distinctive assets on screen is a branded second, so watched rate across a portfolio is branded reach, and the served reach on the media report was only the opportunity for it. Attention, read this way, is not an engagement metric wearing a new name, it is the closest thing the medium offers to the processed exposure Sharp's reach arithmetic assumes.

That is also the one place the medium changes the mechanics. Sharp's reach was built on media where the ad break held people in their seats and a rating point was a fair proxy for a memory refreshed. On a skip medium the swipe is the default and inattention is a recorded event. Amplified Intelligence, which sells attention measurement, puts the memory threshold at about 2.5 seconds of active attention and finds more than 80% of ads do not reach it, so an unwatched impression is not reach in the memory sense. It can remind, if the assets were on screen in the first second, it cannot build. Buying broad, continuous reach cheaply is still right, it is just not sufficient on its own, because the cheapest served reach on the feed can be a large number of people who swiped, and the portfolio and the weekly loop are how you get the reach Sharp actually wants, different openings stopping different kinds of people until watched reach approaches the served reach the media bought. Where we differ is not on the goal, it is on whether the impression counts before anyone watched it.

The second place people expect a difference is wear-out, and here we are closer to Ehrenberg-Bass than the industry habit is. The wear-out research is a warning against retiring working ads out of boredom, teams tire of their own work long before audiences do. We agree, and we hold the line harder than most, an execution in a creative system is retired because it was not watched, never because it is old. What the research does not argue is one execution forever. Romaniuk's own position is consistent assets and fresh executions. Consistency lives in the distinctive brand assets and the campaign idea, variation lives in the execution, and variation in execution is how you reach different people, because different openings stop different people. Portfolio for reach, not refresh for novelty. Red Bull is the public example, one idea, one asset system, decades of it, endless executions, none of which anyone would mistake for another brand's.

The last point is operational. We read attention, we do not buy it. Once watched rate is on the dashboard the temptation is to set the campaign objective to ThruPlay or video views so the platform optimises for it, and the result is a system that goes looking for the people who watch anything, reach narrows to the already attentive, and the light buyer is gone again. The creative is optimised for attention and the platform objective stays set for the brand's goal, reach for brand building. The goal is the brand's, the objective is only the platform setting, and How to align your media and creative strategy keeps the two words apart.

What to do with this

Audit last quarter's spend by one question, what share was structurally pointed at people who already knew us or their statistical twins, and rebalance toward broad reach with creative built for people with no existing intent. List the category's entry points and mark which ones the account has ever addressed in an ad, because the gaps are the next brief. Check the first second of every live ad for the distinctive assets, since that second is the only part of the ad the skipper sees. Pull served impressions and watched impressions per ad for the quarter and read the second column as the reach. Size the retargeting line to its incremental contribution, and measure the brand effects on their own clock, in the order How to read paid social results sets out.

For the light buyers who are the market, working means being effortlessly recognisable, twice a year, for years. Media creates the opportunity to see your ads. Creative determines whether they work. Why you need a creative portfolio, not more content is the next read, because on this medium the portfolio is how the reach gets watched.

Questions

Byron Sharp says reach everyone, so why not just buy the cheapest reach we can?

Buy it, broadly and continuously, that part of Sharp holds on the feed exactly as it held on television. The difference is what the impression is worth once it is bought. On a skip medium a served impression is the opportunity to be watched, and the cheapest served reach can be a large number of people who swiped before the brand appeared. An unwatched impression can refresh a memory if the distinctive assets were on screen in the first second, it cannot build one.

So the cheapest reach is the right media plan, and the creative portfolio is how that reach becomes reach in the memory sense, different openings stopping different kinds of people until watched reach across the portfolio approaches the served reach the plan bought. Read watched rate per ad, keep the objective set to reach, and judge the plan on both columns.

Doesn't changing the creative every week contradict what the science says about consistency?

Consistency and variation live at different levels, and the science asks for both. Consistency belongs to the distinctive brand assets and the campaign idea, which never move, and Romaniuk's own position is consistent assets with fresh executions. Variation belongs to the execution, who is in it, where it is set, which entry point it opens on, because different openings stop different people and that is how a portfolio reaches more of the market.

The wear-out research warns against retiring working ads out of boredom, and a creative system agrees, an execution is retired because it was not watched, never because it is old. What the research does not argue is one execution forever. Red Bull has run one idea and one asset system for decades across endless executions, and nobody would call that inconsistent.

Should we stop retargeting altogether?

No, size it to what it adds. Retargeting is loyalty marketing conducted through paid auctions, and loyalty marketing is a legitimate job with its own economics. The problem is the attribution, not the activity, because the pools contain people who had already moved toward the brand and most of the conversions the dashboard credits would have happened anyway.

Measure the retargeting line against a holdout and fund it to its incremental contribution. Whatever that releases goes to broad reach against light buyers, where penetration is built and where no retargeting pool will ever find them.

Sources

• Byron Sharp, How Brands Grow, Oxford University Press, 2010. Penetration over loyalty, double jeopardy, duplication of purchase, the light-buyer distribution and the Coca-Cola panel illustration. Ehrenberg-Bass Institute, sponsored by large advertisers, sells research.
• Jenni Romaniuk and Byron Sharp, How Brands Grow Part 2, 2016. Mental availability and category entry points. Ehrenberg-Bass, as above.
• Jenni Romaniuk, Building Distinctive Brand Assets, 2018, and Better Brand Health, 2023. Fame and uniqueness of distinctive assets, breadth of category entry point coverage, consistent assets with varied executions. Ehrenberg-Bass, as above.
• Les Binet and Peter Field, The Long and the Short of It, IPA, 2013, and Media in Focus, IPA, 2017. Brand effects build slowly and last, activation spikes and decays. IPA Databank of agency-submitted award cases, trade body.
• Daniel Kahneman, Thinking, Fast and Slow, 2011. System 1 and System 2.
• Amplified Intelligence, "Why does the attention-memory threshold matter?", 27 June 2022. The 2.5-second threshold and the share of ads that miss it. Amplified sells attention measurement.
• Meta for Business, "Demystifying Creative Diversification", April 2025, as quoted by Jon Loomer, 2025. Creative diversification as the best lever to find relevant audiences. Platform wording via an independent consultant, Meta sells the media.
• Facebook internal data based on Datalogix analysis of 50 CPG campaigns, Q3 2012, as presented in Facebook, "Everything competes with everything", circa 2016. Over 90% of viewers who bought in store did not click. Platform-funded, dated.
• KitKat, brand lift testing against the brand's existing creative on the same brief and media. Cleared proof point from our own accounts.
• Red Bull, cited as a public example of one idea and one asset system across decades of executions. Qualitative observation, no figures claimed.