Monday, July 27, 2015

Analysis Paralysis

Numbers nerds (like me!) have been blogging almost daily on Voltair.

Rather than getting smarter with each post, I feel like we all start to send contradicting and conflicting assertions.

We all have been learning less, not more.

Before I crunch anymore random PPM data, it would be nice to know who is using Voltair and who isn’t for one thing. Educated guesses have only the potential to create confusion for all of us.

Yes, A&O&B has a number of clients using the technology right now, but in spite of the perspective that has afforded us, there's a lot we simply do not know.

Like you, I read each new observation, but have ultimately come to the point of reminding myself that “liars figure and figures lie.”

Lets face it.  No one is admitting anything, not because they are hiding anything.  No one knows enough as yet to draw any conclusions.

Things I don’t think anyone knows.
  • Nielsen and Numeris don’t know who has been using Voltair.  Even in Canada, where the ratings firm has asked all their subscribers to unplug their Voltair boxes, who knows if all of them actually did so?
  • Voltaire users don’t know what happened to their sample geography, proportionality, weighting and all the other factors which can move numbers much more than even the uptrends the manufacturer of Voltair has claimed.
Things I think I know:
  • PPM measurement appears to help formats with big cume.  
  • Country, Classic Rock and Pop shares increased when PPM came into our lives.
  • The average age of radio users dropped by about a decade, which may mean that PPM is measuring radio use by younger listeners who didn’t fill out diaries.
  • Talk formats and personalities with long periods of non-music content slipped in the rankers.
  • Sample sizes fell precipitously.
  • Panel/qualitative research replaced random probability sampling.
All of this tells me that we have come to a point that we need less blogging by folks who really don’t have enough information to really know the entire picture.

It think we all can agree that Nielsen's decision to not open their webinar last week to the media was a mistake.  What got reported the next day may or may not have been accurate.  Transparency is the best policy.

It’s time for all Nielsen and Numeris clients to speak loudly to their ratings rep for much more openness and clarity, let alone increasing sample reliability.

It was refreshing recently to learn that Canada's analytics arm has been brainstorming with clients, for example.  We need a lot more of that from Nielsen too.

All of the noise on the technology affecting fewer than 50 markets - albeit the largest ones in North America - is also obscuring an even more pressing issue:  sample quality in the much more numerous diary measurement markets.

Just four years ago we were promised exciting modernizations of the long-outdated approaches being used in radio audience measurement.  After all, Jim Seiler first invented a diary to measure media in 1949! 

Then, Nielsen bought Arbitron and the leapfrog hasn’t croaked once since!

Let’s stop the silly number-crunching without really knowing what is behind the stats we’re all using.  Nielsen and Numeris, please help us all get it right without delay.

It's long past time to focus as an industry on what really will make a huge difference:  increasing sample sizes, better open communication with Nielsen and Numeris clients.

Responsiveness and less secrecy is in order so we all can have more confidence in the data driving our paychecks.