Sunday, January 20, 2008

A Thousand to One

If you were handicapping a sporting event and put a team’s odds at 1000 to 1, you’d be saying their chances of winning were a real long shot.

I came across my own thousand-to-one long shot last week while analyzing a fall book for a client. It was the PPDV (per person diary value) for a male cell and it came in at 1,052. That’s one diary equaling 1,052 people. 1052 to 1 if you will.

When I calculated the same cell for females, I got a second cell with a PPDV nearly as large: 987. 987 to 1.

Would you consider the reliability of the ratings for that demo a sure thing or a long shot?

PPDVs in and of themselves are not the sole measure of reliability, but they certainly are an indicator. Average 18-34 and 35-49 PPDVs have been noticeably on the rise since 2003 and, as of two years ago, had already approached 700 and 600 respectively.

Now it appears things are getting worse.

I contacted some fellow consultants in several formats including country to see if they’d begun seeing 1000 PPDVs. Unfortunately they had.

One national rep’s researcher even said 1000+ PPDVs are now “very common” and are moving out of the 18-24 cells and into the 25-34 and 35-44 cells, and for women as well as men.

I also checked with Mike Oakes who does a good bit of behind the scenes ratings research for A&O and he recalled a near-1600 PPDV for 18-24 males and an over 1700 PPDV for 18-24 females - and this was in a major market two years ago. He also recalled PPDV instances over 1000 for men and women 25-34 and one for men 35-44.

We’re already living with rankings where stations with the most, raw QHRs can rank 4th, 5th or worse in a market, and where stations with far less are the market leaders.

We’re already living with PD Advantage reports which tell us that, on successive books, our 10-year core audience is 20-30, then 40-50 and then 30-40.

Now it appears that 1000+ PPDVs could be the next big thing to play in your market.

I ‘get’ the enormity of the challenge of measuring an audience, and this isn’t meant to be a rant or a witch hunt. It’s a recommendation – and a warning - to you that if you’re not in the habit of regularly checking your PPDVs when you receive your numbers, you need to do so starting today. And if you don’t like what you see, you need to call for action.

If PPDVs consistently reach into the 1000+ range, the odds are that confidence levels in our ratings will posted as “long shots.”

1 comment:

Dave Lange said...

Mike
Amazing observations and facts here, but will Arbitron every really improve the sample??? Who do you 'call for action.' Having been on a few of these 'calls' you get a 'sorry we're doing the best we can' response and off we go to another book with low sample returns.

The real problem is the limitations of relying on land line phone numbers for the sample database. We are leaving the land line world, much as we left telegraphs in the 50s, at an alarming rate for researchers.

Arbitron has to move quickly on new samples databases. But, how many Advisory Panel meetings, tests, experiments, and a slow acceptance of anything new in ratings will it take?

The industry - buyers and sellers need to speak up quickly.