Will the WFA’s cross-media dream come true?
Over the last three decades, asi has been providing an important forum in which all parts of the industry – media owners, advertisers, agencies and research providers – can come together to discuss the future direction of measurement. This was particularly apparent on the Friday 6th November session of our annual conference, taking place this year – as so much of life in 2020 – virtually.
The objective of the session was to establish exactly what the advertisers are asking for with the WFA’s much- publicised guidelines for cross-media video measurement, to understand how Project Origin in the UK is attempting to put those guidelines into practice and to evaluate what pieces of a complex jigsaw need to snap into place for the initiative to succeed.
Will this WFA intervention change media measurement forever or will it be another Project Apollo – the much-hyped but ultimately unsuccessful US project in the Noughties? We polled the audience at the start of the session about how likely they felt the initiative was to succeed. The result was an almost exact 50/50 split between those who felt that the WFA framework was likely to bear fruit in the next three years and those who felt it unlikely. Clearly, we are at a crucial point. There is a definite impetus to the initiative and in Project Origin a concrete project to turn the theory into practice, but still some hearts and minds need to be won over.
What followed was an intriguing and informative session chaired by Richard Asquith who set the scene by highlighting the challenges that have faced cross-media measurement over the last few decades.
“In principle all industry stakeholders – brands, agencies media owners – should be interested in a fully-integrated, de-duplicated cross-media measurement system, but as we know only too well common interest does not necessarily translate into shared endeavour and into making things a reality. Progress has been impeded by a host of barriers including entrenched vested commercial interest, concerns of funding, limitations in the measurement technology, consumer privacy fears and associated legislation such as GDPR.” Richard Asquith
So are we now at a tipping point? Will the WFA initiative be the catalyst for real progress? Brands are now actively intervening in media measurement to reach their ‘North Star’ with the initial goal being cross-platform video measurement.
To understand fully the mission that the WFA has embarked on, Richard interviewed Stephan Loerke, CEO of the WFA. His members are, he says, intervening due to frustration with ‘a status quo that is not tenable longer term’.
Loerke acknowledges that brand owners had been ‘conspicuously absent’ from discussions about measurement issues up to now, but now feel that they have to act due to frustration with the lack of progress. He stresses that consumer frustration with excessive frequency is also a prime motivator for a better understanding of reach and frequency across platforms.
The framework, published in October is a comprehensive vision of what brand owners think measurement needs to look like in the future. Stephan stresses that local adaptation would be needed to take account of services already available in each market, so sees it as a global framework guiding more bespoke local implementation:
“We do accept that measurement is essentially local, it needs to be owned by local stakeholders and it needs to reflect a consensus amongst key players at the local level. We are not seeking to create at a global level a model that is rolled out as a global diktat into countries. That’s not going to work.” Stephan Loerke, WFA
Richard Halton and Joe Lewis of ISBA outlined their progress in implementing the WFA guidelines in the UK as effectively a pilot for the rest of the world. They were able to announce that the Virtual ID approach (VID) has been approved by statistical gurus RSMB. The VID system is the mathematical algorithm that will tie the different datasets together. It has been developed by Google, but is, Stephan Loerke stressed earlier, ‘open source’.
It should be noted that CESP and Médiamétrie in France have also subjected the approach to academic review. Whilst stressing that they endorse the objectives of the WFA framework, they set out some concerns as to how well the VID approach would work across multiple providers and platforms:
“We doubt the ability of the model to provide reach estimates by provider or site. The model can provide these estimates but we believe the accuracy pf the model has been designed to measure total reach and not reach by provider.” Julien Rosanvallon, Médiamétrie
Médiamétrie and CESP claim that individual publishers will not be able to isolate their contribution to a particular campaign in terms of reach and frequency and that the model is driven by individual devices as opposed to single-source panels. ISBA argues that the model to be used in the UK has been adapted to take account of these issues.
Any attempt to model reach and frequency – and therefore direct billions of dollars of ad expenditure – is going to be a hot potato and it is clear that some statistical debate will remain until the Origin pilot starts to produce actual cross-platform video data that can be validated. However, RSMB’s approval of the concept is an important hurdle as it means that the approach is felt to be sufficiently robust in theory to warrant testing in practice.
ISBA plans to have conducted a prototype and pilot by the summer of 2021, whilst agreeing governance and funding to scale up for an ongoing venture beyond that.
“This isn’t just about measures, metrics and pipework, it’s also about underpinning the integrity and auditability of the data with a future-facing independent entity that can guarantee those outcomes over the medium to long term.” Richard Halton, ISBA
Halton stresses the need for participation in that governance from both the buy and the sell side. So what do those on the sell side – the media owners – make of this? How willing are they to play ball?
Robert Schäffner of RTL argues that broadcasters themselves are now cross-platform content providers and so ‘cross-media measurement is not an option or an opportunity for us, it is a must.”
“There is no universal currency that is applicable to all our goods and that carries the value of our total portfolio. We have only a faint idea how the users and the usage of our assets adds up in our customers’ shopping bags. We and advertisers; we share the same dream. We should strengthen local JICs and give them a guiding role to make that dream actually come true and work for us.” Robert Schäffner RTL.
So EGTA is in effect arguing that the priority should be to invest in the current JICs to achieve the WFA goals, as opposed to building something entirely new. Robert also urges the WFA to strengthen its commitment to local adaptation of the guidelines – a lot of progress has already been made locally and the local JICs should be empowered, existing currencies not undermined and existing business models not torn down before new ones are available. That progress is outlined in EGTA’s recently published ‘Advances in Hybrid TV audience measurement.’ This is a comprehensive overview of all the cross-platform video initiatives underway, emphasising the role that TV panels can provide as building blocks for the WFA cross-media framework. So, the message from EGTA’s Katty Roberfroid is: “Don’t reinvent the wheel.”
Katty reports that, whilst broadcasters were pleased to have been involved in the discussions about the project, they have the feeling they were not invited to be fully involved in the design of the solution and that there remain a lot of questions that need answers. Broadcasters are looking for trust, consistency in data quality and transparency. Katty urges against ‘rushing things’.
“Adopting an unreasonable agenda and rushing things could be to the detriment of the final solution proposed, so we really urge that the agenda and deadlines are made realistic to allow for full co-operation.” Katty Roberfroid, EGTA
In order for that Project Origin pilot to go ahead, a number of components need to be confirmed, specifically:
- access to server data from broadcaster players to combine with data from Facebook, Google and other online video;
- access to BARB data to combine with broadcast viewing (still by far the biggest element of all video viewing by volume);
- a panel as a ‘source of truth’ for cross-platform duplication (to check the results make sense) and as a source of all important demographics.
ISBA reports positive progress in negotiations with broadcasters – ‘building confidence’. That is good: without access to data from UK broadcaster VOD the test surely cannot go ahead in a meaningful way.
Access to BARB data may prove more challenging. BARB’s Justin Sampson makes it clear that:
“Duration-based metrics are not a nice-to-have, they are a must-have. We also question whether advertisers really believe that all impressions should be in scope. Why should BARB be comfortable for our audience currency to be integrated with any other data source that uses different audience building blocks? If we do, we help to undermine the principles set out to answer calls for a level playing field made by Mark Pritchard, Keith Weed and indeed the WFA.” Justin Sampson, BARB
Meanwhile, on the Monday following the conference, BARB has gone one step further, tightening its policy on data access for fusion purposes: “…for the purposes of reporting cross-media campaign performance, …licensees will need to demonstrate how the other data sources use industry-agreed definitions which are, where appropriate, objectively equivalent to BARB data.”
Clearly, we need to watch negotiations between BARB and Project Origin with interest, as it is hard to see how the initiative can succeed in the UK – or indeed in other countries – without access to established TV trading data.
With regard to having access to a panel as a source of truth, various options exist, but could a promising possibility lie in Iris, the new UK online survey endorsed by UKOM, and run by Ipsos?
The new service, which replaces the current Comscore panel, goes live in January with a 10,000-person panel. Significantly, that panel is being built out from the BBC’s Compass cross-media panel. That means a proportion of the panel will report not only internet measurement but also broadcast measurement via MediaCell passive metering.
The development of the BBC Compass panel has been tracked at asi conferences over the last few years. Could this be a piece of the jigsaw – the panel as a ‘source of truth’ and demographics – or could Google’s own Screenwise panel or IPA’s Touchpoints play a role? With the Origin pilot happening in the first half of 2021, all should soon become clear.
A fascinating morning session finished with a panel that saw a long virtual queue of delegates quizzing panellists, reflecting just how much interest there is in this topic at the moment. Google was invited to participate in the session but declined.
A burning question, which predates the WFA initiative, is: who actually pays for cross-media measurement? Advertisers are funding the Origin initiative, but what will the funding model be in the long term?
At EGTA’s event in Paris in January (back in the ‘before times’!) advertisers seemed split on this. Some acknowledged that, as the primary beneficiaries, they would inevitably need to contribute, whilst others argued variously that they already pay (by buying advertising), that all media would benefit (unlikely given the objective of efficiency) or that they would only pay if it delivered what they want (so who pays to build it?).
Thankfully, Unilever’s Sarah Mansfield was crystal clear that she believes advertisers will play a direct part on the funding of the initiative.
“We are very much willing to lead from the front in terms of governance and funding and be actively involved going forward, yes.” Sarah Mansfield, Unilever
There has also been concern expressed as to whether the solution will report content as well as advertising. Richard Halton confirmed that it was being set up to have that capability.
“We have the BBC as part of our group as well, so obviously content being involved is a requirement from their perspective. Obviously, the definition of content starts to change as you move across platforms, but the solution has to be built to report content, however that’s defined.” Richard Halton, ISBA
I suspect this conversation may continue however as it is hard to envisage a situation where some platforms publicly report content and some don’t. That may be a deal-breaker for broadcasters.
Overall our session was particularly useful in emphasising the challenges that the WFA initiative needs to overcome. We repeated our poll at the end of the session and the audience remained split, with perhaps a slight shift towards the majority doubting that all of this was achievable within three years.
It is clear that the will is there and it would be hard for any of the various stakeholders to repudiate a system that was demonstrably fair to all parties.
However, it is also apparent that an immense amount of work, effort and co-operation will be required to guide advertisers towards this particular North Star.
If you were unable to register for the event and would like to stream the session, please contact us at firstname.lastname@example.org. In the meantime, you can view the panel discussion that concluded this session below:
Originallly posted by Richard Marks at asi
12th November 2020