The Provocateur – Sonderserie „OUTSIDE THE BOX“: Wavemaker Experten analysieren zum Nachdenken anregende Themen und teilen mit uns offen ihre Meinung und Gedanken.
The inherent frustration of 21st century complexity is that the more opportunity opens up, the more paralysed we become. Nowhere is this more true than in the massively expanded world of television.
More opportunity means more data required to make decisions. Getting more data is rarely the problem – even if its real meaning and consistency is often debatable – the challenge is how to evaluate and apply it. What we all want is a set of comparable measures, distilled to a workable number, leading to good decisions without loss of context.
This is a seeming impossibility in the new world of TV, where the availability of video content to ver-larger domestic TVs and a multiplicity of smaller hand-held devices is creating a hugely uneven playing field.
That’s a big enough challenge, but there’s also the strategically potent force of addressability to consider (and add complexity). If I can use TV with 1st party data to deliver digital levels of precision – including customisation of creative – even as only part of the TV mix, the notion of reach as the strongest single indicator of performance now needs much more qualification: reach of whom, and how?
Homogenising the data to a single source – the World Federation of Advertisers (WFA), working with both the US Association of National Advertisers (ANA) and the Incorporated Society of British Advertisers (ISBA) are just some of the major industry bodies looking at this, alongside ‘TV plus VOD’ initiatives in Australia, Canada, Italy, Netherlands, Nordics, UAE among others – feels like a step in the right direction.
But TV isn’t homogenous – and while every self-respecting TV buyer would argue long into the night on the merits of viewer and programme context, these look like trivial subtleties in the face of new platform, device and addressability dimensions. As a result, the issue these measurement initiatives must grapple with is the risk of poorly reflecting a heterogenous ecosystem, in which ‘reach’ prompts more questions than it answers.
By contrast, channel-by-channel performance metrics, even those built with 3rd party research providers, lack ‘tradeability’. It’s all very well saying that my brand did better than other advertisers on your platform and that I’d get a better result next time with a tweak and likely a splodge more cash. But I’m not (or I shouldn’t be) in the business of making your platform richer – I want to know with reasonable certainty that every dollar spent on each platform couldn’t be better spent somewhere else.
We need to get a better answer to the question before the question: What are the best diagnostic metrics that indicate success? For decades, econometricians have used distributed TV weights – reach and frequency tied to a period – to accurately predict the impact of TV advertising on key business outcomes. TV planners have used those definitions to optimise laydowns that drove growth.
But the assumption that new segments of AV opportunity – contextual, addressable, snackable – will either work to the same rules or contribute on a like-for-like basis to reach and frequency-based rules appears at the very least naïve.
Perhaps the solution lies in a parallel approach to that now being adopted in the world of digital attribution, also being disrupted by the demise of 3rd party cookies. Just as privacy issues are bringing a welcome (if overdue) end to the bluntness of much digital retargeting, so the removal of 3rd party cookies will largely restrict practitioners from looking at the data in such a way as to believe that they can see every consumer interaction and therefore manipulate consumer behaviour in an almost deterministic fashion. The digital piece has only ever been part of the puzzle.
The digitally-based solutions that emerge to answer this need for attribution might be helpful for TV too. We are learning that digital attribution is still achievable using econometric–like statistical methods. This means working with a meaningful level of aggregation in the input data – perhaps daily by activity by platform – to model an output KPI of engagement or awareness or other goal.
This is still fearsomely complex and only five years ago might have caused the lights to dim and PCs to crash, but now we have the math of AI and the computational power to get useful, actionable answers. Answers that will be in some ways superior to what went before – linking as they will to a more specific business KPI, through to tactic by vendor level.
The tradeable nature of these results gives marketers better, more beneficial control. A similar solution in AV looks achievable as well as vital.
David Fletcher, UK Chief Data Officer
