KC May 2023 ratings.

[tall_guy1]

KRBZ is up to a 5.1 (highest share since December 2016), KCKC as well (highest I have seen it.) KZPT only at 2.9 (used to do much better...what happened) and KMJK at 2.5 (same here.) KCHZ and KCJK sit at the bottom with 1.9 each. Thoughts?

 

[Mark Roberts]

OK, so I'm going to ask a naïve question: why aren't these ratings reported in terms of moving averages over a set of time, be it one year or two years? This would seem to be the logical approach in smoothing out the variations from month to month and, I would think, provide a better sense of actual trends.

 

[TheBigA]

OK, so I'm going to ask a naïve question: why aren't these ratings reported in terms of moving averages over a set of time, be it one year or two years?

Actual subscribers receive this information in many different ways. There are dailies, weeklies, yearlies, etc. The monthly 6+ public ratings are basically press releases. Markets that still use the diary are measured in three-month trends.

 

[011nerve]

KCHZ 95.7 had a slight humming sound in the audio when the audio was dead air every time I heard it, bad engineering.

As for 105.1 KCJK, KC has enough Rock stations in the area.

 

[davideduardo]

OK, so I'm going to ask a naïve question: why aren't these ratings reported in terms of moving averages over a set of time, be it one year or two years? This would seem to be the logical approach in smoothing out the variations from month to month and, I would think, provide a better sense of actual trends.

Subscribers get the data each week and month for the PPM surveys. They have software that allows for customization, such as custom dayparts and custom age groups, date ranges, time periods and ethnic combinations. You can get data from just one day or for a period of many months, depending on what you select.

I usually run 14 month reports in PPM markets, with the full 13 annual reports plus "same month one year ago". And I can run those for a wide variety of ages, ethnicities and even individual county zones and gender choices.

I can also look at each report on AQH share, rating, persons as well as daily and weekly cume rating and cume persons and other stratifications. And I can create custom dayparts or even look at hour by hour or quarter-hour by quarter hour if I wanted to.

The one thing I never look at is the 12+ or 6+ AQH share. It is so valueless that they give it away... making it worth every penny of what people pay for it.

There is such a thing as having too much information. I always caution people about digging too deep in ratings as the sample size declines the narrower you get in your report specifications. It is easy to make wrong decisions if you are looking at such tiny samples.

 

[davideduardo]

Markets that still use the diary are measured in three-month trends.

Diary markets no longer have "trends". Every report is a fully weighted three "month" period, with each successive report dropping the oldest "month" and adding a new four week period.

 

[Mark Roberts]

Subscribers get the data each week and month for the PPM surveys. They have software that allows for customization, such as custom dayparts and custom age groups, date ranges, time periods and ethnic combinations. You can get data from just one day or for a period of many months, depending on what you select.

I usually run 14 month reports in PPM markets, with the full 13 annual reports plus "same month one year ago". And I can run those for a wide variety of ages, ethnicities and even individual county zones and gender choices.

I can also look at each report on AQH share, rating, persons as well as daily and weekly cume rating and cume persons and other stratifications. And I can create custom dayparts or even look at hour by hour or quarter-hour by quarter hour if I wanted to.

The one thing I never look at is the 12+ or 6+ AQH share. It is so valueless that they give it away... making it worth every penny of what people pay for it.

There is such a thing as having too much information. I always caution people about digging too deep in ratings as the sample size declines the narrower you get in your report specifications. It is easy to make wrong decisions if you are looking at such tiny samples.

Thanks...but I'm thinking of something different, namely, how to avoid short-termism. It's like the difference between traders and investors. Traders react to the day-to-day movement of stock prices and not much else. Investors look at longer-term trends, which is where moving averages help provide insight by smoothing out day-to-day movements to provide a clearer indication of price trends. What I'm extracting from your response is that the data are available - but what media research folk are looking for are mostly snapshot-in-time averages (or medians). What I would look for in addition to those means/medians are those moving averages, plus standard deviations and possibly more. I agree that there can be too much information and over-analysis, and the point about sample size is well taken. I've always wondered about the statistical validity of audience research, whether radio or TV, and the conclusion I've reached is that it's better than nothing but far from ideal.

 

[TheBigA]

I've always wondered about the statistical validity of audience research, whether radio or TV, and the conclusion I've reached is that it's better than nothing but far from ideal.

This is why you shouldn't rely on one thing, whether it's Nielsen or Edison or any other study. You look at it all, compare it side by side, and that way it's not limited by the specific survey. Nielsen is the way it is because of the cost. The sample is the size it is because smaller is less accurate, and bigger is too expensive. But contrary to what you said, in my view, there never is "too much information."

 

[tall_guy1]

KCHZ 95.7 had a slight humming sound in the audio when the audio was dead air every time I heard it, bad engineering.

As for 105.1 KCJK, KC has enough Rock stations in the area.

The Point and Mix seem to have eaten up whatever ratings they had at one point. I wonder if switching would be wise?

 

[davideduardo]

I've always wondered about the statistical validity of audience research, whether radio or TV, and the conclusion I've reached is that it's better than nothing but far from ideal.

Remember that ratings are much more intended as a sales tool and not as a programming tool.

Ratings are your final grade for the course. Research is your reading and homework that you do to pass the course and get a good final grade.

I like to understand how ratings are done, but when I review them I prefer the big pictures because everything else is subject to sample variations and uncontrollable factors like weather events or breaking news in a market that change listening patterns.

 

[tall_guy1]

Remember that ratings are much more intended as a sales tool and not as a programming tool.

Ratings are your final grade for the course. Research is your reading and homework that you do to pass the course and get a good final grade.

I like to understand how ratings are done, but when I review them I prefer the big pictures because everything else is subject to sample variations and uncontrollable factors like weather events or breaking news in a market that change listening patterns.

Some stations like KCHZ have done poorly for quite a while, but don't notice many changes. Shouldn't the owners be trying to tweak the format more often to make it work? (I am aware they have a lesser signal, but did much better in the past.)