September 18, 2024
In my nearly 20 years in the advertising industry, the prevailing narrative has been that it’s television for whom the bell tolls. But in 2024, TV advertising isn’t just a multi-billion-dollar industry; it remains a core component of our digital evolution.
Linear TV (a term I generally dislike, as no one outside the advertising industry even knows what it means) is often labeled a relic of the past. Yet while linear audiences may be in decline, total TV remains strong globally. Indeed, I often wonder whether we’ve missed an opportunity to reflect on everything we’ve learned about TV’s effect and how we once bought advertising on the channel, before consigning it to the margins of advertising history. The lessons from TV’s golden era could well apply to advertising more broadly in the digital world.
What I mean is that there is a lost art in how we plan and buy media—an art that was intrinsically tied to the way we bought linear television. As the industry evolves and, in some ways, outgrows the channel, we’ve replaced much of what made advertising effective with questionable ad-tech, vanity metrics, and hyper-targeting based on flawed and often weak indicators of buying intent. We’ve also seen a weakening of the partnership between clients, agencies, and publishers, a constraint on creativity, diminished attention, and, ultimately, reduced effectiveness.
Earlier in my career, I remember buying TV for some of the world’s most iconic advertisers—Apple, PlayStation, Coca-Cola, to name but a few. And there’s a lot the “digital world,” if such a thing exists, could learn from how we once approached linear TV.
I distinctly remember an era when TV buyers would rib fellow advertising pros with phrases like, “I bet he’s never bought a spot,” spoken in a broad Cockney accent as we swaggered through the office—confident we were the engine room of the agency. Mildly amusing as that was at the time, I’m not sure it holds up now. “I bet they’ve never bought a fraudulent ad impression on a low-rent platform” doesn’t quite have the same clout.
Like many ex-TV buyers, I’m sure plenty will recall the infamous ‘Billets Rack.’ Yes, this torturous device saw us routinely punished by auditors based on the perceived quality and cost of our media buys.
While that method of auditing is vastly outdated today, it did ensure agencies were accountable not just for the efficiency of their buys, but also for the quality of the inventory purchased. This seems almost foreign today, where we rely on platform-based attribution models—wildly inaccurate ones, at that—to measure efficiency (ROAS). We’ve become obsessed with cost-per-whatevers, yet often neglect to look under the hood and ask: What inventory are we actually buying? Are our ads even being seen? And what is the real cost of low-attention, low-quality inventory?
Back in the day, if Apple’s TV buyers had placed a spot on Jeremy Kyle—for my US friends, think a lower-budget version of Jerry Springer that came to represent “Breakdown Britain”—we’d have been dragged over hot coals. Publicly flogged. Consigned to making tea for the entire TV buying floor for a week.
Today’s counter to that is simple: “We let the platforms/data providers/algorithms find the right audiences.” But that’s a fallacy in itself. The identifiers we use to target audiences are often built on weak signals, and advertisers would do well to avoid the long tail of inventory that promises cheap CPMs (assuming the impressions aren’t fraudulent) in cluttered, low-attention, low-quality environments.
There was something to be said for linear TV trading that has been lost in the digital world. Few mandates exist today around maximizing prime-time visibility—the space where premium, high-engagement, high-production value content lives. There’s little focus on securing a prominent position in ad breaks to maximize viewership, or even on the programming, content, or context where ads run.
Instead, we just let the algorithms take over. Or worse still, we let PMAX butcher our ads in the process, and we don’t ask the hard questions.
Why? Because cost has become more important than quality.
I touched earlier on our growing appetite for hyper-targeting, where we layer third-party identifiers on top of one another, building audience profiles so nuanced that it’s a wonder they even exist in the real world.
This was always the digital world’s power play against linear TV. With TV, we were (and still are) constrained by basic demographic audience constructs, with no ability to trade audiences based on affinity, interest, or in-market signals. Yet, there is a growing body of research today that suggests linear TV still outperforms many other channels in terms of return. For fans of econometric models, you’ll often find that TV tends to offer higher payback and a longer payback window than many alternatives. So, how is this possible when targeting on TV is so limited and often perceived as wasteful?
Firstly, good TV buyers understood the importance of context and content. Apple was a prime example of an advertiser that put this at the forefront of their media buying approach. Yes, they traded broad demographic audiences, but they meticulously reviewed programming and understood which audiences had a strong affinity for specific shows, buying media based on both content and context. They carefully curated spot lists that were relevant and seemingly ‘on-brand.’ While this came at a premium, this practice was commonplace across the industry.
Secondly, an increasing amount of research suggests that over-targeting, or hyper-targeting, isn’t the silver bullet we once thought it was. Advertisers who adopt what the Ehrenberg-Bass Institute refers to as ‘sophisticated mass marketing’ perform significantly better than those relying heavily on third-party audience identifiers. Programmatic media buying may bear some responsibility for the decline in advertising effectiveness by perpetuating the myth that more nuanced audience profiles automatically lead to better performance.
Jon Bradshaw's recent article (linked here for those interested—and you should be) highlights this fallacy. He references Rikard Wiberg's work, which explores the potential upside of hyper-targeting on Meta and concludes that it’s actually less effective than building broader audience coverage. This is supported by findings from esteemed econometrician Dr Grace Kite, and Barrowman also cites Nico Neumann et al., who provide evidence that second- and third-party identifiers are often significantly less effective than random sampling. Even the use of first-party data is often outperformed by contextual targeting approaches.
So, as we prepare to bury linear TV and consign it to a footnote in advertising history, let’s not forget that context and content may still be the most valuable tools in our targeting toolkit.
For my own broad-stroke approach to targeting, you can find more details here.
I previously touched on another increasingly popular topic: the attribution fallacy. For several years now, we’ve been seduced by a plethora of meaningless digital metrics that bear little relation to advertising effectiveness or outcomes. The ability to track every interaction a user makes led us to believe that every interaction was meaningful. We forgot that some people have larger thumbs than others. As devices got smaller and smaller, we mistook the clicks of these large-thumbed people for genuine interest in our brands, products, and services. In reality, they were just sick of us interrupting their daily digital routines. We started measuring everything and linked it to our digital advertising, irrespective of how meaningful (or meaningless) the data truly was.
Thankfully, as an industry, we’re getting marginally better at filtering out the noise to focus on the metrics that actually matter. But it’s taken a long time.
Meanwhile, linear TV has often been criticized for how intangible its outcomes were to measure. Reach and frequency were the standard metrics—the best available “measure,” if you could even call it that. We could explore lifts in site traffic or brand searches, and for classic direct-response advertisers using unique 0800 numbers, there have always been ways to measure advertising responses on TV to some degree.
However, as a general rule, TV measurement, until the current democratization of econometrics, felt less tangible. But we knew that, and in many respects, it was a strength. The lack of rigorous measurement meant fewer metrics to lead us astray. It allowed advertisers with significant budgets to have more faith that advertising worked, without being hampered by a sea of vanity metrics that bore little relation to actual outcomes.
Digital advertising became accountable, incorrectly, for things like bounce rates, time on site, and average order value because we developed such robust web analytics—not because digital platforms gave us a unique ability to control these outcomes. In reality, advertising can only influence these metrics to a limited degree, and yet we made ourselves responsible for them.
The lesson from linear TV here is that less is actually more. Hindsight is a wonderful thing, but I often wonder if we had been more judicious about what we chose to measure when digital and programmatic were in their infancy, we might have fewer click farms swallowing up advertising dollars today—and perhaps a little more faith in digital advertising (wishful thinking, I imagine).
Another area that has caused some mild frustration with digital media, which I never really felt in my former life as a TV aficionado, is creativity. For years, in digital media—whether display, video, or social—advertisers have been bound by rigid formats. The nature of digital ad platforms, which historically functioned as aggregators of content rather than creators or producers, placed more constraints on creativity than conventional media ever did.
This isn’t necessarily the case today, of course. With dynamic creative capabilities, expansive video offerings, and the rise of creators, it’s become easier for advertisers to be playful in the digital space. But this was one area where TV (and, to be fair, radio, print, and outdoor media) had the upper hand. The relationships you could build with production houses and sales teams allowed for ample playfulness in media execution—something that’s largely been absent as advertising has shifted toward self-serve and standardized ad formats.
Looking back at my own experience, how would we have executed Sony UK’s Little Big Planet 2 campaign back in 2010 without the ability to sponsor Movies on Five? More recently, how would Go.Compare’s excellent work with Family Fortunes and The Voice have succeeded creatively without the linear TV platform? Thames Water were a perennial bridesmaid in several award categories for their exceptional print partnerships. I could go on, but a quick search on WARC, and you’ll find thousands of examples of creativity that, up until recently, were not always viable ‘digitally’—or at least not to the same scale. These partnerships, and the ability to borrow credibility from source content, often breathed life into what can sometimes feel like the transactional function of advertising and media buying.
As I mentioned earlier, the digital world is starting to wake up to the importance of creativity. Tom Roach's presentation to the IPA (Institute of Practitioners in Advertising) (here) suggests that advertisers are, or certainly should be, looking beyond the confines of best practices in their digital platforms. The rise of the creator economy is opening up opportunities for brands to integrate and entertain within popular culture. It’s just taken us a while to get to this point, and it’ll take some time before these behaviors become normalized.
Media is often perceived as a commodity—something merely bought and sold with limited thought. While this isn’t always true, the ‘digital era’ has done little to dispel this misconception. There’s still a lingering belief that one can simply push a button and watch sales materialize. There’s also a fundamental gap in understanding what is causal versus coincidental. In some areas, that gap has widened, as we’ve failed to measure the metrics that truly matter.
In doing so, I believe we lost sight of what makes advertising interesting. We stopped viewing media as a canvas for creativity and storytelling, instead cramming rational proof points into display ads barely larger than a postage stamp. We turned away from borrowing source credibility, shifting our focus from quality environments to cost-efficient spaces. Bound by platform rules, we’ve stifled the exciting, creative potential of compelling media executions. While I sense the tide might be turning, it’s going to be a long road ahead.
And though I may sound nostalgic, I truly believe there’s still a lot we can learn from the way things used to be.
PR
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.