Managing Complexity: 2023 Conference summary (part one)

In this first of two articles, our Research Director Richard Marks looks across the three days of our 2023 conferences in Nice to highlight some of the main themes and reflect on the current direction of travel for media and its measurement.

Currency is not staying in its lane

The cross-media imperative is seeing JICs, MOCs and currencies increasingly looking to widen their remit, to the extent of not even necessarily staying in their original lanes:  Vividata in Canada has morphed from a print JIC to a cross-media service; AGF in Germany has widened its coverage to measure publisher video ads and online display; whilst NMO in the Netherlands has brought together video, audio, text and online measurement in a co-ordinated system with cross-media measurement the ultimate goal.

It’s clear that the lines of demarcation between the JIC silos are eroding, illustrated by the fact that our joint session at the conference was hosted by Denise Turner of Route in the UK. An outdoor JIC CEO chairing a session at a Radio and TV conference? It makes sense when you look at the importance of video to the outdoor industry. Also, as we heard both in and around the conference hall, many of the challenges now facing measurement – server data, panel costs, metrics and definitions – are now common to all, not just those currencies that started life as TV, radio or print surveys.

The politics of measurement

In the three decades that asi has been holding its annual industry gatherings, the politics within the measurement business and the pressures from without have never been greater. In our Friday measurement panel we asked what the typical timeline was for building the complex but effective ad measurement systems that had been presented during the conference. There seemed to be some consensus around the need for about 18 months to build, test and launch the measurement itself, but an acknowledgment that industry politics, consensus-building and legal issues – around different interpretations of privacy issues for example – could all add considerably to that timeline.

Measurement matters

Nonetheless, whilst those politics can be deeply frustrating, we should be glad that the business of measurement is so high profile. If measurement is getting people engaged it is because what we do does actually matter: guiding billions of dollars of ad spend, content investment and technological development. It is surely rewarding to work in an industry that clients really do care about. Whilst this can make building consensus difficult, would the measurement community rather have the wider media industry be indifferent to what it is doing?

The real impact of the WFA’s North Star

It has been four years since Project Origin and its US equivalent led by the ANA were announced at the asi conference in Prague, tasked to deliver the WFA North Star of cross-media measurement. Both are currently encountering difficulties, particularly when it comes to getting acceptance from the broadcasters, a critical component for their ultimate success. So, again, industry politics are an issue, which does make one wonder why the WFA selected for its pilot markets two of the most political media markets on earth.

Origin was not in attendance in Nice and is focusing on the launch of pilot data next year. However, regardless of its long-term success, the WFA North Star initiative has acted as a wake-up call to local currencies around the world and significant progress is being made, particularly in cross-platform video measurement.  asi delegates witnessed real progress shared by, amongst others, Médiamétrie in France, Kantar in Brazil and GfK with IMDA in Singapore. Meanwhile, Nielsen in Denmark were winners with TV 2 of the Tony Twyman Award for best paper at the TV & Video Conference with an account of their strategic use of four-screen measurement (see video below).

Measuring ads not just content

A common theme is that content is more straightforward to track across platforms than advertising. Initial progress has been on the content side, but now the focus is on tracking campaign performance across linear and on-demand itself; for example the TTVAM system in Finland and the Replay ads system built by Mediapulse in Switzerland. Ad server data is a critical component and it’s essential to gain a better understanding of how that data can be made to work effectively. As an industry, we create panels and surveys that are lovingly crafted with the end goal in mind. Ad server data offers huge opportunities, but it is crude oil that needs to be refined to deliver reliable measurement.

‘Fit for TV’: help or hindrance?

The last year has seen efforts to more precisely distinguish ‘premium’ TV content from other forms of video content in terms of perceived quality and transparent regulation, as outlined by Justin Sampson with Barb’s ‘Fit-for-TV’ definition of its measurement remit. Some content on platforms like YouTube would meet this definition, but much would not. Google argues that restrictions on what should and should not be measured in a cross-media system should not be determined by siloed industry JICs, but that all video should be measured with the MRC two-second minimum definition of a viewable impression as the universal starting point for calculating an audience. So, the fundamental point of difference is this: should the TV currencies preserve an element of perceived quality, transparency and regulation in what they measure, or should data be allowed to run free? Are concepts like Fit-for-TV and Premium TV making the remit of the currencies clearer or are they adding an additional roadblock to cross-media measurement? Perhaps they are doing both.

A crisis of definitions

Jonathon Wells of Nielsen: re-evaluating standards

Arguably ‘Fit-for-TV’ is at least clearly defined as a concept, unlike many other critical terms that were referenced across the three days. For example, are we too busy arguing about the minimum duration of an impression to take a step back and agree what an impression is? The MRC in the US makes clear that the definition of an impression is a building block for an audience itself, the successful delivery of an ad to a device. We then get into the further definition of ‘viewable impressions’, but even then who (and how many in the home) viewed that impression remains the critical step in calculating an audience.

An impression records a device, not a person. In that sense the word is well-chosen as it provides a (vague) impression of the audience, but not the audience itself. Yet the term ‘impression’, perhaps because of its usage in the world of internet research, is often used instead of audience. We need to start talking the same language or we have a Tower of Babel which makes progress much harder.

To that list of definitional challenges, we can add the tendency to use ‘cross-media measurement’ when we really mean cross-platform video measurement. Yes, cross-platform is an important first step towards cross-media measurement and the conference showed real progress on the former but, to be fair, very little progress in moving beyond video.

As was highlighted at a number of points during the conference, the situation is not helped by the absence of any commonly-agreed standards for media measurement. GGTAM (Global Guidelines for TV Audience Measurement) is now two decades old and restricted to panel operation in effectively a different era. There has been a huge degree of innovation since then, but could common standards be agreed for complex hybrid systems with router meters, server data and data fusion?

Delivering an attentive audience

Nora Schmitz of Ipsos (left) and Catarina Sismeiro of Annalect: attention is not just ‘eyes on screen’

The measurement of attention is a phenomenon that is growing in significance and coverage at our asi conferences in recent years. Again though, precision in definitions is important. As the Attention Council’s Andy Brown summarised, there is a lot of work going on from companies using varying approaches to measuring attention. Ipsos and Annalect looked beyond visual attention to evaluate the relative utility of three forms of attention, specifically eyes, mind and body.  Attention is not just ‘eyes on screen’.

We also need to be clear about the difference between measuring content (the ability of media to deliver an attentive audience) and attention to the ads themselves. It is easier to see that the former could somehow be baked into a currency as an additional metric to aid media planning. Route in the UK for example already evaluates viewability within the currency calculations, but it is viewability of the sites themselves, not of individual ads, which will depend on the creative. It is the media owner’s job to provide an attentive audience and the buy-side’s job to make the most of that attention. The first of these is potentially within the remit of audience measurement; the second is more about ad testing, not currency measurement.

It’s about people not devices 

Across all three days for both audio and video at recent conferences we have been hearing more and more about the opportunities provided by census data to measure audiences. This could be set-top box data, podcast streaming, connected TVs or ad server data, whilst an increasing number of services are adding router meters to track devices within the home. Hybrid measurement has progressed from being a new concept to the accepted norm in many markets now and the availability of this data from devices has only served to amplify the importance of persons-based panels to convert these devices into people. Kantar presented data from a range of their measured markets that clearly demonstrates that the number of viewers per TV screen varies greatly by market, day part, season and genre, showing why persons-based panels are so vital to turn devices into people, to turn impressions into actual audiences.

So much happened at this year’s conference that we decided to split our analysis of the main themes into two parts, so stay tuned for the sequel next week in which we will be discussing content measurement, passive audio measurement, first party data and clean rooms, sustainability, AI, global guidelines and our CEO interviews amongst many, many other themes!

We are very grateful to all our generous sponsors for their support during this event: Nielsen for sponsorship of the TV & Video Conference and asiConferences App; GfK for sponsorship of the Radio & Audio Conference; Ipsos for sponsorship of the asiNetwork Social Drinks Reception; and dataBreeders for sponsorship of the Tony Twyman Awards for best paper at both the Radio and TV Conferences.

The 2024 asi International Radio & Audio and Television & Video Conferences will be held on 6th to 8th November 2024 in Venice, Italy.