We live in a fantastic marketing era where the rubrics of advertising effectiveness are known by more marketers, with more data and more certainty than ever before. We’ve never had this level of advertising clarity. I am old and fat enough to write that last sentence. And I put this fantastic age of effectiveness clarity down to a few key factors.
First, the gradual decline of academic institutions and their influence over marketing. Knowledge dissemination was once owned by the big American business schools and that meant it was slow and rarely shared outside very small, elite circles. These days the power of podcasts, YouTube and, ahem, virtual marketing training courses mean a much broader audience gets this information fast, often free and in far more entertaining forms than the olden days.
We are also enjoying a golden generation of effectiveness thinkers. It’s led by Ehrenberg-Bass, and Field and Binet. But riding on their coattails and occasionally now stepping off and ahead of them is a broader group of thinkers, who add to the effectiveness corpus. The Hurmans, Nelson-Fields, Woods, Pauwels, Kites and so on.
This is any marketer’s global faculty. Each brings something different to the effectiveness party. And, following on from the original effective champions, all their work is empirical, applied and instantly useful. We should also shout out to the research companies that have backed up and often furthered these thinkers with broader empirical support. I nominate three agencies – System1, Kantar and Analytic Partners – for their dogged ability to follow the effectiveness train and add new carriages of data every now and again. It’s rare for research companies to be so on top of things and, again, so good at dissemination.
Knowing is not the same as doing, but it’s a big step in the right direction.
Gaze back at the last decade and the effectiveness concepts line up beautifully. Any newly minted CMO who knows their advertising onions is going to follow the script.
They will appreciate the pre-existing size of their brand has a massive impact on current strategy and future impact. They will use excess share of voice (ESOV) to assess the right overall level of advertising investment. They will split it into long and short baskets based on the nature of their category, brand and market. They will brief the correctly selected agencies on a handful of objectives, with clarity on target audiences and position. The power of the creative idea will be paramount. The work will need to be distinctive and studiously apply distinctive brand codes throughout. It will employ emotion, especially for the branding part. The work will spread across multiple media channels in an integrated fashion. It will prioritise reach over extended frequency. It will take into account the varied attention metrics of the different channels. And the marketer will use a blend of econometrics, market testing and brand tracking to optimise the campaign’s effect, assess its ultimate impact and produce learnings ahead of its next iteration.
I love that last paragraph, partly because it summarises so much good work and also because it already sits in the consciousness of many marketers without me having to write it.
Whether all of this stuff then gets to be applied is an entirely different, more practical question. You might know about ESOV and be able to calculate that you need 12% of the category share of spend – about 2 million quid – to have the best impact next year. Getting that money is a different story. You might have followed the recession playbook and understand why you must keep investing in brand through tough times, but defending your budget internally may be impossible. The power of creative in all of this might be obvious, but you still struggle to get great work from your agency. Knowing is not the same as doing, but it’s a big step in the right direction.
A new addition to the marketing canon
And it may be time to propose a further additional leap in our increasingly clear journey towards effectiveness. It’s been banging around for about a year thanks to various studies from the large research firms, but it may be time to add it to the book. I am talking about ‘wear-out’. Or, more specifically, I am talking about the absence of it.
For 50 years, marketing academics have been measuring two things. First, the number of repetitions an ad requires to have a significant positive effect – known as wear-in. Second, the number of repetitions of an ad that transforms that positive effect into either a neutral or negative effect. The use of basic experimental methods in which college students were repeatedly exposed to advertising stimuli produced a generally accepted approach to frequency. Namely that wear-in happens with the first ad but peaks at three exposures. At around the fourth exposure, wear-out begins as boredom and annoyance set in. These negative thoughts grow and eventually undermine an ad’s persuasive impact. This early experimental work spawned both the heuristic of three exposures being optimal and the idea that wear-out unavoidably occurs.
But the devil was in the epistemological detail. Like so much of the early advertising research, the belief that an experimental setting, with invented ads, incentivised college students and a host of crudely manipulated variables were somehow representative of real-world advertising turned out to be nonsense. Viewing ads in a lab was not free of context, nor were the results therefore generalisable to all contexts as claimed; viewing ads in a lab was a bizarre, alien context that made the outcomes entirely inapplicable to anything.
The partial viewing of an ad by an unmotivated (and often un-noticing) actual audience made a mockery of the precision of wear-out research. Similarly, the generalised assumption that all ads followed similar arcs of wear-in and wear-out ignored the very real confounders associated with different creative shown across different media. And finally, given that much of the ‘long’ effect of advertising might take months to manifest, the idea that the effects of an ad could be instantly measured after a sequence of exposures was empirically suspect.
Analytic Partners measure of campaign wear-out
Now, thanks to data from Kantar, System1 and Analytic Partners (those three again) it appears the general advice for most marketers is that wear-out – at least in the real world – really does not exist. Analytic Partners looked across more than 50,000 ads in 2020 and found that only 14 of these campaigns had run their full course and were exhibiting wear-out when they were replaced. The other 51,232 were pulled despite the fact they were demonstrating no signs of wear-out at that time.
That finding not only demonstrates that wear-out rarely occurs for target consumers. It also signals the contrasting reality that wear-out for marketers, who grow impatient with their existing tactics and seek the green-grass smell of new creative, is a very real issue. The problem of wear-out is one of the marketer, not the market.
We’ve seen a similar story from System1. The average rating for an ad on its database during the first month it airs is 2.2 stars out of 5. The average rating for an ad that has been continuously shown for more than two years is 2.2 stars. And that performance, give or take the variance inherent in the sample size, remains constant right across the thousands of ads System1 measures.
So consistent is the effect, Nick Williamson and his team of merry analysts have been trying to find any evidence of wear-out across any of its database, at any time, for any reason. That might seem an odd way to disprove wear-out but the team is so convinced it does not exist they have spent most of 2023 just making sure it never occurs. The only slight wrinkle they have uncovered is with ads that focus on a specific event, for example, a FIFA World Cup ad that is shown after the event has ended. But that’s screamingly fucking obvious, unlikely ever to happen in the real world and, even then, the wear-out effect is not that pronounced.
System1 measure of campaign wear-out
Kantar’s tracking data shows a similar finding. “There is no evident impact of how long the ad has been running on that ads’ performance,” the company concludes. It adds that “wear-out is not some unavoidable destiny for ads, it is possible to create strong ads that can maintain their success”.
The company cites the ‘Tealand’ campaign from Twinings as an ‘old’ ad that continues to perform free from wear-out issues. Originally conceived in 2013 in Chile, the ad continues to be enjoyed, have impact, drive demand and be distinctive almost a decade later across a range of global markets.
‘Tealand’ from Twinings, data from Kantar
Takeaways for marketers
So what does this all mean for marketers? Aside from an interesting and important addendum to the effectiveness literature, there are several crucial implications of applying the apparent absence of wear-out to your own work.
1. Learn patience
The most obvious implication of the lack of wear-out is the need for greater patience on the part of marketers with respect to their creative execution. As noted above, while wear-out does not exist for most consumers, it is alive and well in the world of most marketing departments.
In another implication of market orientation, marketers must learn the patience of the market. They look at their ads with such focus and passion that – inevitably – they grow tired of the work long before most of their target market has even become accustomed to it. This results in campaigns being pulled long before they have had time to develop and impact the market fully and leads marketers to replace these high-potential ads with new campaigns that – if the hoary old data is to be believed – are likely to be less effective four times out of five.
I remember Jon Evans, the storied chief customer officer at System1, telling me about all the times he had shown a marketing team that their new ad was significantly less effective than the campaign they were about to pull. The reaction? General disappointment, cynicism and an illogical determination to still launch the new ad because it was new, because they had spent all this time making it and they weren’t actually very smart.
2. New marketers should resist new campaigns
The single most common move for new senior marketers taking over at a brand is to put the account under review and start thinking about a Big New Campaign. I can count on one hand the number of senior marketers who entered a new role, looked at the existing advertising campaign and agency partner and – with the help of research rather than their lower intestine – opted to keep running things as is. Yet this wear-out data demonstrates very clearly the likely value in doing just that.
To be clear, the extant campaign might be shithouse or driven by a different strategy. In which case a new campaign is appropriate. But, more often than not, the existing campaign is likely still working and the new marketer has bigger, more important boulders to move first as they starts their tenure. Let’s celebrate the brand manager who walks into a new role and does not launch a new campaign, rather than the one who automatically comes up with something new and cool. The odds favour the boring over the ‘brave’, despite what some numpty in a polo neck told you on a stage last Christmas.
3. Dust down old ads
Armed with this data, marketers should also be raiding their corporate libraries for the dusty VHS tapes of their old brand ads. Sure, you can build something new. But that does not mean it’s the right move. Your predecessors from the 70s and 80s were just as smart as you, they just dressed badly. Hell, there is a 50/50 chance they were smarter.
Don’t forget to look backwards before you head forwards. Sitting in your archive might be a wonderful ad that, with a modern tweak, can work again. Three times I have suggested to clients that they re-run old iconic ads and three times my suggestion has been met with blank stares. Why would we run an old, proven, already paid-for, much loved ad when we can create something completely new and unproven?
The message from the wear-out data is that if it wasn’t broke you don’t need to fix it. Just re-run it and celebrate your skill in seeing the value in proven old ads versus unproven new stuff. Just remember that the biggest single deterrent is a sexy soundtrack featuring a famous pop singer. Old ads may not fade away, but licensing arrangements do. The number of times marketers have to pull their great old ads because they don’t have the rights to the soundtrack anymore – you know that blank YouTube link – is a scandal that needs to be addressed.
4. Save next year’s creative budget to spend on media
The average big new campaign will spend 20% to 30% of its annual budget on execution, leaving the remaining 70% for media investment. But what if you knew you could run your ad for longer? That would change your approach, right? For starters, you’d spend more in year one because you know this campaign will be running for many years. You’d spend more on testing too because the stakes are so much higher and longer than they used to be. But then you can look forward to many years where close to 100% of your ad budget becomes working media investment in the years that follow.
I know half a dozen CMOs who believe in ESOV but cannot get more budget to reach their optimum investment level. They are all trying to game the model by driving creativity, codification, attention and the kitchen sink so that their £1m will work like £2m. It won’t happen, of course. But if you accept wear-out does not exist – if you go for better, longer creative – you just earned yourself an extra 25 points of SOV to win the game, versus your new-ad-loving, underspending competitors. This is a HUGE point. Go back and read this paragraph again.
5. Create ads that cross geographical boundaries
There is a clear link between creating an ad that will last the test of time and its geographic and channel expansion capabilities. What I mean by this – and I have zero evidence to back up this point – is that when you look at the great ads that defy wear-out, they also defy the cultural constraints of different markets and different media. The Twinings Tealand campaign is not just a perfect example of one that works across the years. It also works across multiple markets. And multiple media. There is something in the ‘campaignability’ of big ads that transcend time, which means they transcend other barriers too.
6. Advertising agencies are fucked… more so
If you work in an agency and you’ve read this far – which most of you haven’t, because your attention peaks at 200 words and disappears at 400 – you appreciate this is not good news. Add it to the ‘Why We Are More Fucked in 2024 Compared to 1985’ deck that you’ve been building. I’m sorry. But more is not always more. New is not always good. And sometimes your biggest competitor is the quality of your existing work.
Mark Ritson teaches the effectiveness corpus in one of the modules of his Mini MBA in Marketing – the other nine focus on other equally useful stuff. The next programme begins in April 2024.