1 of 1- RadioActive
- Moderator
Offline 
Radio Veteran Not Discouraged By Elmnt FM's Poor Ratings
Dave Charles has worked for a ton of successful radio stations, both on air and off. But the veteran radio consultant may have taken on his biggest challenge yet as one of the key cogs in the two Elmnt FM outlets. The Indigenous stations in both Toronto and Ottawa have been on the air for three years now but in almost every rating book, they continue to show virtual zeros.
Charles doesn't believe it and is now on the hunt for a new rating service that he says will actually show people are listening and responding.
“Numeris ratings PPM ‘people meter’ system hasn’t yet responded to the Elmnt fm format,” Charles tells [FYImusicnews.ca]. “We’re looking at other ways to measure our true audience on all platforms.”
Dave Charles Is Fighting To Win With Twin Elmnt FMs
- radiokid
- Member
Offline
Re: Radio Veteran Not Discouraged By Elmnt FM's Poor Ratings
Some would argue it almost would be worth it for them not to be in the book and sell based on a target audience/hope/dreams like Evanov does with Proud FM. It also would give them more fixability musically to stand out and not try to be all things to all people as they are now.
- RadioAaron
- Member
Offline
Re: Radio Veteran Not Discouraged By Elmnt FM's Poor Ratings
Bentwater wrote:
The lowered tiered stations couldn't count on national $ through the agencies, consequently, their boots-on-the-ground reps were expected to be pretty good at selling the sizzle and power of radio.
I don't think that can work in 2021 in Toronto.
There just aren't the same businesses that are themselves independent and making their own marketing decisions while at the same time being big enough for a mass marketing medium to be efficient.
I see 4 options to do radio profitably here:
1. Aim to get ratings through Numeris and the agency buys that follow. (This is basically what they're trying. If they make it any more PPM friendly, they'll lose the Indigenous identity.)
2. Focus on a 3rd language group, and do business within it, including hosts that sell their own airtime. (Not an option in this case)
3. Go hyper-local to a part of the city, and focus on the market-within-the-market. Like Sauga 960. (Not an option in this case)
4. Go all-in on Indigenous programming, and finance through grants, underwriting, and sponsorship. This is the best option IMO. There'd be a lineup of companies that would love to have "financially supported independent Indigenous media" on their public record.
- Skywave
- Member
Offline
Re: Radio Veteran Not Discouraged By Elmnt FM's Poor Ratings
The real issue here is that the Numeris sample size for the Toronto market is woefully inadequate. My understanding from a meeting with Numeris a few years ago is that there are only 600 PPMs for the Toronto market survey. They are measuring audience shares in a universe of more than 5 million people.
When Numeris publishes their top line summaries, they never talk about "confidence level" or "confidence interval". Typically you want a confidence level of 95%. The confidence interval is another way of stating statistical significance. A confidence level of 2.5 means that the estimated percentage share of audience could be 2.5 percent higher or lower than the real number.
Go to surveysystem.com/sscal.htm and you can play with a couple of online calculators.
So here's what's interesting. For a sample size of 600 and a confidence level of 95%, the confidence interval is +/-4 %. Numeris lists 24 radio stations in the latest survey for Toronto. Seventeen of them have shares under 4%. Fifteen have shares under 2.5%. So if they calculate that you have a 4 share, it may be 8, or it may be zero. Duh!
If your share is 2.5%, a .5 confidence interval represents a potential error in the share of +/- 20%, whereas for CHFI, with a share of 11.4%, the .5 confidence interval would translate to an error in share of only 4%.
Using the calculator, to have a confidence interval of 1% (which would still represent a significant potential share error for the low audience stations), the sample size would have to be 9604.
An interesting fact regarding sample size versus the size of the universe: once you get above 300000 population, the required sample size for a given confidence interval remains static.
The fundamental issue is that when you are trying to measure very small responses from a large number of stations and want a reasonable confidence interval, you must have a very large sample size. Right now, a paltry six people meters adds up to a 1 share. And this is without getting into the issue of how the meters are distributed relative to the age demographic breakdowns, geographic area, household income, average hours tuned, etc.
- RadioAaron
- Member
Offline
Re: Radio Veteran Not Discouraged By Elmnt FM's Poor Ratings
The real issue here is that the Numeris sample size for the Toronto market is woefully inadequate.
Absolutely, and really well explained. Unless you're doing an ultra-mainstream format with a large CUME, you're either going to have no ratings or be subject to wild bounces as panelists come in and out.
Elmnt need to operate like that's never going to be solved, but I'm not sure exploring other measurement options is the answer. I think the whole business model needs to change.