Advertisers that have gone all-in on performance media have essentially abandoned the signaling value of media, explains Faris Yakob  those signals are about to get much, much harder to see. Here’s how marketers need to react. 

In 2005, London Business School Professor Tim Ambler introduced signaling theory into advertising from evolutionary biology. Based on earlier research and a bespoke study, he directly challenged the most famous quotation in advertising, regarding the assumed 50% levels of wastage. This ‘wastage’ was historically perceived as a function of the breadth of media available to advertisers which were inherently reaching broader audiences than their actionable market. 

However, Ambler shows that advertising that is perceived to be expensive works better. Ambler explains this is analogous to the costly signaling involved in what is known as the Handicap Principle, whereby animals, most famously the peacock, use wasteful characteristics (the peacock’s tail) to signal biological fitness. Hence, according to Ambler, the wastage is the part that works. 

The initial study used different televisions commercials to understand the impact of higher quality production and found it improves efficacy but indirectly, by increasing perceptions of brand quality, which drive long term growth (Gale, 1999, quoted by Ambler). 

But the biggest impact of the famous quote attributed to John Wanamaker (what a great name for a marketer!) has been on media. As digital media allowed marketers to slice consumers into ever smaller targets, these digital marketers embraced efficiency, using harvested personal data to diminish ‘waste’. 

Marketing thinkers have been building the case for brand investment for decades, but still today growth in media comes primarily from performance. It’s search and the new darling ‘retail media’, which have grown incredibly in a few short years, presumably because it’s possible to track conversions and it’s the closest place to the point of purchase, which has been assumed to be the most impactful. Regardless of whether that’s an illusion caused by the legacy of last click attribution models, the entire point of programmatic and performance digital advertising was that it allowed brands to buy the same audiences more cheaply than in high status and therefore expensive media vehicles, by following them across the internet and harvesting them on long tail sites. Always on, data-driven media engines look for signals of intent in the market – things like searching for a brand or product category most obviously – and use it to focus their spend on people most likely to buy right now. 

The idea of ‘audience buying’ inherently assumes, like media calculations in general, that impressions are fungible and therefore all channels are functionally equal. We’ve always known this isn’t true – the pricing variance across channels indicates differences in ‘quality’ assumed to lead to differences in impact – but this pricing is not robust because quality is measured in different ways, and the relationship between those aspects and business outcomes is complex. 

The draw towards cheaper CPMs is powerful but all experienced media planners understand that this is a compromise. Reach, cost and quality are limiting factors on each other. The better quality media, the fewer impressions you can afford. You can choose to buy more, more cheaply, but what is the true cost in terms of impact? Ignoring, for now, the fact that impressions are increasingly calculated in different ways on different media, it’s obvious that the impact of seeing sixty seconds of film in darkened cinema with your phone on airplane mode is fundamentally different than scrolling past something in the stream. Intuitively, the difference should have an impact, one that isn’t based on issues of viewability (although arguably absolute screen size does matter more than amount of pixels on screen) but rather on the impact of the channel itself on communication – and we now know that it does. 

EssenceMediacom released a new report, authored by CSO Richard Kirk, expanded from earlier work by Thinkbox, called Signaling Success. They make the point that “to date the focus on quantifying media quality has centered on physical measures: dwell time, shared viewing, viewability and now attention.” Their research uses a fictitious brand “to strip out the effect of creative, budget, product quality and distribution, to hone in on the difference made by the media.”

The study demonstrated that the choice of media channel impacted respondents’ perception of the brand and that signal strength is different across channels. The signal strength from both studies was almost identical, suggesting it is an enduring feature of media. It’s built from years of consuming media and shared consumption media (TV, radio, cinema, print) which leads to a stronger signal value. Then they mapped the signal data to their Media Mix Navigator, an outcome based planning tool built from hundreds of econometrics studies. 

This clearly showed that “the signaling data correlates strongly to the long-term ROI numbers we associate with media, but bears little relation to short-term ROI”. Like so many things in advertising, this is a reminder of the importance of brand investment and channel choices that suit that objective, of driving long term growth, rather than simply harvesting customers who are currently in-market.

Advertisers that have gone all-in on performance media have essentially abandoned the signaling value of media, in exchange for better signals they use to target it. Unfortunately, or perhaps not, those signals are about to get much, much harder to see. 

A breakdown between ‘reach’ and impact has been happening for some years due to fragmentation, which is a significant problem for media agencies and clients. Reach is our most powerful weapon and broad reach is considered even more important today in light of How Brands Grow

However, when Zenith did an analysis in 2023 (also masterminded by Richard Kirk in an earlier role), they demonstrated the relationship between cross-channel reach and business outcomes is worsening across the board. Zenith invented a new metric, “right reach”, which factors in duration, attention levels and contextual relevance into reach curves, and showed the impact of this metric on business results is much stronger. “For the same brands mentioned at the start of this piece, across the same weeks, with the same business outcome data, we were able to show that right reach is frequently >100% better at predicting the outcome of a weekly media plan.” Ads in context that people consume more actively for longer work better and working out how much better and whether it’s worth the cost is the job of media planning. 

However, the situation is about to get much worse. Privacy panics about personal data have spread into policy and practices. According to OMG, 78% of all time spent globally on media is now digitally served, across social, digital, streaming video and so on. Inevitably, all media will eventually be digital in some form, either obviously or on the trading back end. 

Following Mozilla, Apple eliminated third-party cookies from Safari five years ago and then implemented App Tracking Transparency in 2021, which forces apps to ask people to opt in to data collection. The majority of people refused, cutting off one of the most important third party signals for performance marketing, especially of digital goods. Google has finally begun the long feared cookie deprecation in Chrome, the world’s largest browser. Further, when performance media is used to target smaller audiences it relies on third-party data that meta-analyses suggest isn’t correct much of the time, and according to even more recent research the increased efficacy required from smaller segments is often not worth the cost:

“Approximately half of those audience segments require the click-through rate to double compared to an untargeted campaign…Our model also shows that narrow segments require a lift that is likely not attainable, specifically when the data quality of these segments is poor” (Overwhelming targeting options: Selecting audience segments for online advertising, International Journal of Research in Marketing).

The signals we use for digital marketing are going up in smoke. 

Inevitably, this will lead to some significant changes in how we understand and implement digital advertising. The price of performance media will go up, as has been clearly demonstrated with prior waves, which will impact how many digital-first businesses operate, because the break-even point will change. There will be a renewed focus on affinity, context, and quality in media. 

No doubt the split will always remain; the division between now and later can never be fully resolved, but it’s likely signal dependent marketers will be forced to embrace the power of signaling. Otherwise, advertisers will simply have to accept the non-deterministic modeling of outcomes offered by the black box of a platform A, even if it isn’t transparent about targeting or even how spend is allocated across the platform, and agencies will render even more power unto the partners we have funded into global dominance.