Day 1 of the asi virtual Conference, November 2, examined “Future-proofing audio measurement” under Paul Kennedy, RAJAR, UK thanks to an array of industry sponsors. Most radio measurement services around the world were seriously handcuffed by the COVID-19 crisis and innovation via various integrated approaches is evolving as the new approach to measuring radio audience hearing/listening. Beyond traditional or digital radio station listening, the increase in podcast listening has further muddied the measurement waters of achieving a comprehensive single audio (all formats anywhere any time) currency.
What have those services that managed to keep going learnt about the impact of the crisis on radio and audio consumption? Has the balance permanently shifted towards passive and online forms of data collection for radio and audio? Meanwhile, are streaming solutions growing in importance and is there anyone out there who has not set up a podcast by now?
During his session, “What’s going on in podcasting – and why you should care”, James Cridland, podnews.net curator, identified the lack of comparative audience data to radio for podcasts worldwide as a critical issue. Podcasting has grown significantly over the last 5 years. Currently data is limited to downloads (dissemination) albeit with a large share of the podcast market, Spotify has audience (delivered to a device) profile data in its walled garden.
Other measurement is limited to weekly podcast “listeners” (in the US via Edison) and some non-comprehensive podcast rankers based only on ‘downloads’, i.e., not listening. James offered that podcasts ‘on-demand’ would generate initial listening and then be programmed by broadcasters for wider appeal. He noted that 85% of podcasts do not make money!
Richard Marks, asi addressed, “COVID-19: what happened?”. A worldwide asi survey of the various radio currencies underlined the variety of methodologies and I suggest their long standing, underlying issues which have been emphasized by a pandemic. Importantly, two countries will likely switch to PPM; several will abandon diaries (there are still diaries!!); and some will move to access panels and/or on-line measurement.
With radio audience measurement being driven by economics not research quality, Richard concluded that overall radio audience measurement needs to be better, less expensive, and more resilient to crises. Yes? And I would like to buy an elegant mansion overlooking the Med in Cannes for less than €1 million.
Giorgio Licastro, GfK shared their, “Innovation in emergency during the Italian pandemic”. No surprise really that GfK’s integrated solution was based on PPM melded with ad-hoc on-line surveys. The new approach did reveal that radio was alive, well, and resilient albeit with some quite different listening profiles by time of day.
During “Tracking the changing patterns of listening in Norway”, Ragnhild Herlofsen, Nielsen was able to show Summer listening reaching the highest ever due to the travel restrictions. Daily reach was up approximately 15%. As in Italy the profile of those hearing changed by time period. The key to Nielsen’s offering based on PPM, which includes location data – an indicator of habits, was to provide programmers daily tracking data and location data (Norwegians go to their cabins in the Summer!) to understand changing “hearing” habits, adjust programming, and retain advertisers.
Paul Kennedy reviewed RAJAR’s “Investigation of new Hybrid approaches” that would be resilient and sustainable in any circumstances. Its use of Face-to-Face interviews came to a scorching halt March 23rd. It did introduce a “Tracking Panel” during the Summer to understand the changes in listening patterns recognizing that this was not a currency. While radio listening generally returned to normal levels in September new lockdowns in the UK will add to RAJAR’s difficulties. Its ultimate integrated solution will likely include Face-to-Face, plus a panel as well as supplementary sources.
RAJAR, backed by ISBA and IPA, apparently does not have Passive Meter Measurement or PPM, in the mix. This surely raises fundamental research questions.
Peter Larsson of Swedish Radio and Jakob Bjur of Kantar offered, “The Swedish RAM – One conclusive radio currency!” Economics played a role in Kantar’s hybrid solution due to the importance of providing radio “hearing/listening” in detail in 25 key market of which only 4 are measured by PPM with a sufficient sample (out of a total of 1500). Their solution to providing the industry a single currency was to continue CATI for the local 21 smaller markets – 30,000 interviews per year – and water mark over 350 stations across Sweden. The two vastly different measurement techniques are then integrated via a model which uses the overall listening patterns in all markets via the PPM sample.
The fundamental question arises as to whether Kantar is doing very precise things (modelling PPM data) with very imprecise data (CATI)?
In perhaps the most forthright and frightening research presentation of possibly the week, Simona Nemes, Triton Digital, discussed, “Invalid Traffic (IVT) Filtration – filtered vs unfiltered streaming audio traffic.” Invalid digital audio traffic can be general or specific and as noted can be much more than simply bots and data center disparities. In terms of specific IVT, Simona stated that, 22%- 45% of session logs were IVT! She also admitted that even with its highly sophisticated software Triton can only filter around 75% of IVT TLH. As stated, “Accurate filtration of different IVT requires very complex sophisticated detection procedures.”
The asi virtual Conference continues Tuesday Day 2, November 3 with, “The Future of Content Delivery: How will linear and on-demand content co-exist?”
The second part of this article will be published on 5 November 2020 (ED.)
