Last week I blogged about how radio in the UK is measured. If you didn’t read it- you can do here right here.
As I said in that blog post, the pandemic has impacted how radio is measured in a pretty major way due to the inability to perform large scale face to face respondent interviews and recruitment. I think I said that on the face of it, that method seems somewhat “analogue” in 2020 and not wanting to go over all the reasons why that is the case again, this week I thought it would be interesting to look at some thoughts on alternatives and weigh them up. Walk with me a moment…
Phone- don’t you realise people have phones? Whenever I have talked about ratings and how it all works to newbies to radio, one of the responses I get most is about phones. Surely replacing the face to face interview can be done by using the phone right? Seems like a pretty straight forward thought doesn’t it especially when the world is working from home and holding zoom meetings and remote working like crazy right now!
So lets think it through in terms of practicality. As you will know from the blog last week, each radio station has its own TSA defined by postcode- so that is what you are measuring- radio listening defined within a postcode area for each station. The current measurement methodology knocks on doors in a geographic area to recruit respondents. What about replicating that by phone? Lets thing about that for a second… The person living on the street next to yours, that you don’t know, could at the minute get a knock on the door by the current system. So what about by phone? Unless they are by chance on a marketing database- and privacy and limiting random contact has been a major thing for some time now- how are you going to ring them or get their number? Genuine market research calls are allowed when scanning against Telephone Preference Service lists. If you limit the call list to those that have somehow given their number to a database (probably by accident or without thinking), you immediately limit the scope and impact of any research into listening. How successful might those calls actually be? There is a little thing called GDPR too of course… might have been in the news a bit since it came into force and the information (like phone numbers and name and address etc) that companies etc have on you is kind of impacted just a tad! [NB I have literally minutes of fun winding up those “Did you have an accident that wasn’t your fault?” type callers]
Maybe all that is the price to pay for moving away from the door to door approach? Could be, but before making that step its worth understanding and recognising the significant change it would have on the results. That said, the media landscape is abound with people conducting call out research and it produces results which are both helpful and insightful for companies and brands and so shouldn’t be dismissed out of hand. Any change would be a wholesale replacement of the kind of research radio has now, and so the benefit analysis would be that with the hope that one day a vaccine would return to a more free lifestyle, would this change be a worthwhile future-proofing enhancement of the ongoing measurement or the best of the choices available right now to get things up and running again short term solution?
Hey- you know the internet exists right? This is the other common response. Who needs phones and talking- we can all do this online and its easy! True- the practicalities of the above phone scenario exist here though. Think of that person that you don’t know in the street next to yours. How do you know their email address? OK, so what if you didn’t contact them by email (which would probably end up in a spam filter anyway) and instead conducted research via station websites online. Again, this is a common response I’ve had when talking about how radio is measured. Many stations and organisations conduct research with their online and app users. It’s valid insight and gives some useful knowledge on users experience and behaviour. What about all those listeners who don’t visit the station online or via an app but listen to it nevertheless? I would wager that is a sizeable chunk of the audience. The same is true of recruiting via social media in that there are a lot of radio station listeners who don’t follow the station on social accounts.
Maybe thats the price to pay for moving away from door to door? Maybe it is but again, its important to know again that this would be a similar sized change to the one already mentioned. Again, is that a way of future proofing long term or a best choice available right now short term?
Is there anything we can learn from other media? Well yes. BARB measures TV viewing in the UK and whilst the methodology and results etc are totally different (set top boxes, eyeballs instead of ears etc), BARB uses a panel of regular respondents for regular viewing along with a device centric methodology to capture online/ other streaming. Again, this is a different type of research to that which produces ratings for radio and of course radio and TV are very different. Having said that with most phones essentially being a TV in your pocket as well as a radio and music player etc- there is some symbiosis there perhaps? Maybe a regular recruited panel, recruited by a combination of means, combined with a measurement of streaming stats and a de-duplicated combination might be a new version of radio ratings which moves beyond a shorter term pandemic solution, to something which might be more robust and encompass some new media reporting which makes the shift from the current position more worthwhile?
There are some benefit to panels- not least of which you get to see ebb and flow of individual usage over time rather than a single week long snapshot of a raft of people. There is also the downside to that too. Recruitment, whilst still challenging, once the panel recruited would be at a lower level in order to top up or refresh the panel on an ongoing basis.
The task would be huge. BARB recruits 5100 establishments (buildings), equating to 12,000 people it says, compare that to Rajar’s UK wide sample of 24,857 (Wave 1 2020 unweighted sample, UK). Like I said last week- Rajar is a very complex survey with a lot riding on it and there are no simple, easy and problem free solutions!