The Netherlands has just released its first results using a new app-based audience measurement system. We spoke to Johan Smit, director at the Platform Media-Adviesbureaus (Platform Media Consultancies) to find out more
RedTech: Please outline how the new passive electronic audience measurement is different from the previous one.
Johan Smit: The measurement changed from active, recall (filling in a diary on mobile, PC, laptop, tablet or paper) to passive (an app on mobile phones registering what the media consumer is listening to). In essence, it changed from what people thought they listened to, to what they could actually hear.
RedTech: Why were the changes necessary? What motivated this project?
Smit: The former method was based on recall, memory. From other countries and earlier tests with different passive measurement technologies we learned that this creates a substantial overestimation of the listening time and a slight underestimation of the reach. After several tests, we concluded that a passive measurement was a better representation of the real listening behavior than the radiolog system.We chose the app over other technologies, although all have their pros and cons. The app added value as it could be combined with Reality Mine technology, measuring online consumption via data (4G/5G) and could be used for TV/video measurement via audio matching.
After several tests, we concluded that a passive measurement was a better representation of the real listening behavior than the radiolog system.”
New technology also supports daily reporting on a minute-by-minute basis, as in the past, it would be a monthly rolling two-month average. Through watermarking/encoding we now add the platform of listening (linear, via IP, app) directly and continuously, instead of through separate research.
RedTech: Which stakeholder benefits the most from this new measurement system?
Smit: All stakeholders benefit:
- Advertisers. Better reporting of actual listening behavior; chosen technology enables cross-media reporting and planning; better control of audio investment; better comparison to other media.
- Media Agencies. Radio planning has matured in one giant leap forward; more work, but better advice; radio planning becomes interesting again. More added value for the client, as more precise planning is possible, and as added value optimization is now possible, too.
- Radio Stations. Events, be it an hour a day or a week, can now demonstrate increased ratings; better pricing of inventory; more accurate insight into profiles of listeners; better comparison to other media and data to support commercial propositions. More insight into how (parts of) programs resonate with the listener and more data to support build of programs or day parts.
RedTech: What have you learned about radio listening behavior from the first results?
Smit: We see slightly more reach and substantially less listening time. We also see some differences in profiles. For instance, some gender-oriented stations look to be less outspoken, others on the other hand, more so.
Morning listening peak is longer, but with a lower peak. People listen to more stations than before. Events have an impact, varying from substantial to huge.
RedTech: Did anything stand out as suprising?
Smit: Based on previous tests, morning peak behavior was certainly expected, although the extent was not. We are still looking into this, although we have some reasonable explanations, we need to validate those.
We over-estimated the willingness of the 13–17-year-old demographic. We had to return to a form of recall for this group. From focus groups, we know that privacy concerns (from the respondents and their parents) played a role, but also that they lose interest and leave the panel after a few months. We did a qualitive study at the beginning of the project to check our questionnaires (did respondents interpret the question the way we intended and did we interpret their answers the way they intended?). A curious and eye-opening finding was that respondents see listening to the radio as “listening to talk radio.” Listening to a music station was classified as “listening to music.”
A curious and eye-opening finding was that respondents see listening to the radio as “listening to talk radio.” Listening to a music station was classified as “listening to music.”
RedTech: Do you expect to see a re-allocation of advertising budgets from one media channel to another or a rebalancing as a result of the new study?
Smit: Yes, on both counts. As far as reallocation is concerned, stations can now demonstrate added value (in incremental reach or additional contacts) in relation to other media.
Specifically in the day parts, other media are consumed less. They will be able to adapt their rate cards better to reflect that. Agencies will be more interested in using radio in their media plans as it will be in the plannings systems directly next to the other media types. I expect a re-balancing between the stations. Smaller stations will have to rethink their commercial proposition as the new data shows different profiles. Events will attract more budget, so the other stations might want to revise their commercial proposition during such an event.
RedTech: You could do much more with this new method. What are some of the challenges?
Smit: As far as additional opportunities, the major obstacle is the acceptance of change in the world of audience measurement.
So, for now, we are focused on getting everybody to see the opportunities of this new measurement system. Stations have to adapt their commercial policies and then refine them, agencies have to retrain all their forecasting and planning models and advertisers have to get used to a new currency.
From a technology point of view, I don’t see many limitations. Data science and statistical knowledge are stretched, however. As we no longer have just one measurement technology for any media type (except print) fusing the data from different data sources is reaching the limits of our current knowledge. As this is all very new, we are learning and building up our expertise and knowledge.