The 2017 asi Conferences in Nice summed up in just six words

Attempting to summarise the range of topics covered at an asi conference can be an intimidating task. As Research Director I do have the advantage of having seen and reviewed the papers beforehand. I am also a judge for the best papers, and manning the asi twitter feed means I am in the room for the entire agenda. However, the sheer range of topics covered across the three days continues to grow and in recent years this has been reflected in the expansion of the names of the conferences themselves to encompass not just television and radio, but the wider worlds of audio and video.

So what follows is an attempt to look at the main themes that emerged across the three days, resisting the temptation to work through all the papers like a shopping list. So inevitably a number of otherwise great papers may not feature here as I attempt to step back, squint and see what broad shapes emerge.

The first word that leaps to mind is achievement. When reviewing the asi conference three years back, I commented that audience measurement, especially TV, resembled the skyline of a modern city – an awful lot happening, but a lot still under construction. A lot of ambitious talk, with the emphasis on hybrid systems, but a lot still to deliver. Encouragingly we are now seeing quite a few of the systems that blazed a trail moving from implementation into delivery. Total video systems in Netherlands, Sweden and Denmark are now ‘live’ and being used as currency. The much talked about Dovetail fusion in the UK – combining panel with server data – went live in parallel run a few days before the conference and will go public in March. This leads me on to my second key word: visibility.

I have commented in the past that hybrid measurement systems would face challenges in terms of credibility unless the approaches they use are transparent. If not, accusations of black boxes and ‘smoke and mirrors’ would be made.  At the 2014 conference in Madrid I challenged the measurement companies on stage about how transparent they were willing to be in a world of Intellectual Property. They committed to an open and collaborative approach and in Nice we did seem to genuinely be moving beyond broad statements of intent to a reassuring level of detail.

We heard how Auditel is fusing a Set Meter panel to the PeopleMeter panel in Italy and also of a similar project conducted by Ipsos, whilst fusion underpins all of the total video systems listed above. We seem to be moving beyond arguing about the merits of fusion to discussing the specific detail – how to cope with non-domestic viewing, the impact on reach and the dangers of regression to the mean (we even learned what it means!).

‘Regression to the mean’ defined from asi conferences on Vimeo.

My third key word, running across both the video and audio conferences is streaming. At the audio event, US streaming giants Triton highlighted the growth and potential of streaming video and argued it was vital that radio and audio providers take control of their own data assets. Global Radio’s DAX system (Digital Audio Exchange) is significant on two fronts. It blurs the lines between radio and subscription services to promote the targeting opportunities of streaming audio: ‘a rising tide lifts all boats’. It also made a convincing argument that ad revenues from connected audio services can actually be higher than broadcast – by a factor of up to four it was claimed. The better the quality and breadth of the data, the greater the premium: a convincing argument for the power of data to generate revenue.

Voice is my word number four – not just across the audio day but the conferences overall. Radiocentre in the UK highlighted the huge opportunities that voice assistant speaker units like Amazon Echo offer to radio – radio dominates their usage, aided by the efforts Radioplayer has made to embrace the technology.

Meanwhile at the TV and video sessions there was much discussion of the impact that embedded voice assistants will have on how we access content, including an excellent but scary look from asi’s Graham Lovelace into a future of artificial intelligence and frictionless interfaces. My generation grew up in a world of buttons and keyboards, whilst the next expect to be able to touch and swipe. Will the kids of today expect to simply bark their commands?

Our joint session (‘Advertisers, media agencies and digital – how are they doing?’) attracted the biggest audience in asi’s 27-year history. Bob Hoffman (aka the Ad Contrarian) was typically acerbic, launching a scathing attack on both the ethics and effectiveness of online advertising. David Wheldon of the World Federation of Advertisers agreed with Bob’s analysis. The balance of opinion has certainly shifted from the naïve optimism of the early days of digital advertising to a more exacting appraisal of its merits. We do however need to make sure that the pendulum does not swing too far in the opposite direction. Over the three days of the conference as a whole the merits of digital advertising as a part of the media mix were apparent. The challenge is to reassure as to transparency and accountability and to encourage and foster wider industry collaboration.

Moving swiftly on to my next key word: Context. One question remains unresolved to my mind, an issue that was touched on by a number of papers and played a big role in our final debate, ‘Who really needs a Total Video currency?’ Is online video, and especially video in social feeds, a similar beast to broadcast video or is it entirely different? Facebook and Kantar Media demonstrated that social video advertising worked in a different way and requires different rules for it to be created and placed effectively. Isn’t this an argument against Total Video measurement? Do we believe that, particularly when it comes to advertising, social video and broadcast are effectively as different as, say, radio and TV? As a result, if we are moving towards Total Video measurement systems, is it realistic for them to encompass all video?

In our Friday afternoon debate it was argued – convincingly I feel – that Total Video measurement systems should maintain a clear focus on measuring exposure – gathering credible and authoritative data on second-by-second viewing. It is then up to the users to argue about the relative merits of a 30-second broadcast versus a 4-second skippable pre-roll.  At least those arguments can be made using the same building blocks, the same agreed data. However if there is pressure for the ‘JICs’ to move beyond that, to attempt to measure relative impact and effectiveness, then troubled times may lie ahead.

There seemed to be agreement in the room that the role of a currency is to provide a single agreed measure of exposure (to be the referee) as opposed to making value judgements about relative impact (to be the commentator).  Our Thursday afternoon session focused on advertising and a number of papers looked at ways in which we can provide a context for audience measurement – Publicis looked at the importance of creative being tailored to platform, whilst MESH looked at the role of context by examining consumer experiences.

My final word: Excellence. We have given an award for Best Paper at the Television & Video Conference for a number of years now, generously sponsored by RSMB in honour of the late great Tony Twyman. This year we not only added an award for Radio & Audio (again sponsored by RSMB), but also commissioned impressive glass award trophies, which were well received, even if their weight may have challenged baggage allowances on the way home! Although the judges’ decision was final, we did also sneak a look at the scores given in the app by delegates, so the winners were themselves the result of a hybrid fusion of qualitative opinion and ‘big data’.

The Tony Twyman Award for best paper at the Radio & Audio Conference went to Lena Brun of Finnpanel in Finland and  Ricardo Gomez-Insausti of Numeris in Canada. Both teamed up to look at their shared successes – and challenges – in introducing online radio diaries. They also drew on data from RAJAR in the UK. The paper was not only insightful but a great example of collaboration across the asi community.

The best paper for Television & Video went to Brian Fuhrer of Nielsen, who achieved the triple crown of excellent content, engaging presentation style and timeliness. There has been much speculation about the likely size of audiences to SVOD services like Netflix and this paper finally lifted the lid. This was achieved not by claimed recall or an ad hoc survey but by using the US PeopleMeter panel itself, allowing a direct comparison using the exact same metrics. A comparison which in the case of Stranger Things (16 million viewers to one episode and rising) may have had the US broadcast networks wincing. At least they now know the size of the challenge. Returning to an earlier word – achievement – this paper felt like a real game changer.

Looking across the event as a whole, at least one new delegate commented that they were surprised and delighted at the willingness of speakers to openly discuss challenges as opposed to ‘aren’t we great’ corporate promotion. In outlining the challenge of measuring TV audiences across eleven time zones in Russia, the system adopted was described as the intersection between the purely scientific and the practical. I am tempted to argue that could also be the perfect description of an asi conference itself, a platform in which new ideas can be tested and debated.

Our grateful thanks to Triton Digital for sponsoring the Radio & Audio Conference this year; to Nielsen for again sponsoring both the Television & Video Conference and the asiConferences App; and to Ipsos Connect and Médiamétrie for sponsoring this year’s two asi Network Social drinks receptions.

See you in Singapore next May and in Athens next November!

This analysis of the conferences was helped by three special ‘aftershow’ videos, shot at the end of each day with delegates giving their immediate impressions of what they had heard. The comments expressed in these have proved hugely useful.

Afterthought Day 1 – asi International 2017 from asi conferences on Vimeo.

Afterthought Day 2 – asi International 2017 from asi conferences on Vimeo.

Afterthought Day 3 – asi International 2017 from asi conferences on Vimeo.

Originally posted by Richard Marks at asi
20th November 2017

Comments

  • NigelJacklin

    I would describe regression to the mean (for fusion) as ‘can’t find FRED.’ You may be able to find a ‘partner’ respondents with and F and an R…and perhaps an E…but as there are around half a million combinations of four letters it’s unlikely that you find a ‘matching’ FRED. Try this…think of a few colleagues in the same age and gender group as you…pick one who watches a couple of programmes (or reads the same newspapers) as you do (your F and R)…what are the chances they’ll have other activity which matches yours (their E and D).