Ad Rates 2011: The Big Takeaways

It's that time of year again; one of my favorite articles (Ad Age's upfront ad rates) has hit the Internet. I'll spare you all the stuff about how they are estimates, but they are and you can read about all that on the article. Bottom line for me: for a random on the Internet like myself, this is nothing less than one of the most important windows into the industry that exists for public consumption.

For pretty much the next 12 months (but especially last night and today), this article leads to many a tweet and forum post and blog post where someone links the article and then points out one random comparison as illuminating everything there possibly is to know about TV. I, on the other hand, am much more thorough. This information deserves it.

Just like last year, I'm gonna do four posts on this stuff. Last year, I opened up by revisiting the "Peetooplus" idea (correlating P2+ counts and A18-49 with the ad rates) that I first tackled in depth back in 2009. This year, I'm shortening that because it doesn't really change year-to-year, although a couple of the notes in this post will kinda address that. I recommend examining the peetooplus label if you are somehow reading this blog but still aren't buying into that whole adults 18-49 myth.

In this first post, I'm just gonna highlight a few big takeaways that may not be obvious to the naked eye:

1) The correlation between upfront 2011 ad rates and 2010-11 average adults 18-49 ratings improved for the second straight year. It was 0.90 when I first did this in 2009, 0.92 last year, and this year the correlation coefficient between A18-49 and ad rates is 0.96! And just 10 out of 58 shows were more than 25% away from the average price per demo point per 30-second spot. The correlation being better probably doesn't really "mean" anything except that maybe my organizational skills are better. Anyway, the good correlation itself should drill home that this is the number that is setting ad rates. (Or at least the C3 version of these numbers.) I didn't do the viewership correlation this year, but maybe some day I'll be bored and throw those in with my helpful charts of yesteryear. I am willing to guarantee it will not be nearly 0.96.

2) One of the three posts later this week will be another thorough year-to-year post like the one from last year, but the bottom line: if you'll recall, it was frequently reported around the interwebs circa upfront time that the price of a 30-second spot per demo point (usually called "CPM" or cost per thousand impressions) was on the rise by 10% or generally thereabouts. My scrub of these numbers confirms that's true. I had the average in same-timeslot shows last year at right below $44,000 per average A18-49 demo point per 30 second spot. This year, it's over $48,000, or a difference of +10%! Now the $48,000 itself not a hard and fast number that you should attach to anything, since these aren't the real 18-49 numbers that they're using (those would be the C3 ratings). But the increase is definitely there. VERY important point alert: as long as the networks can continue to get more and more per demo point, the decline of broadcast television is not as precipitous as the ratings make it seem. Things will fall apart when, as many on the Internet like to say, the advertisers "realize" they're paying more for a shrinking pie. (As if they somehow don't now.)

3) OK, let's talk about what's overvalued and undervalued. If you scroll down a ways on last year's intro post, you can see what I identified as the shows that got the most value for their A18-49 demo and the shows that got the least value. Basically, highest-valued were the Fox animated shows, lowest-valued were the Saturday shows.

This year: more of the same. Here are five "highest-valued per 18-49" shows, where "diff" compares the show's $ per A18-49 compared to the average $/A18-49 of all 58 shows in my table.

	A18-49	Price ($)	$/A18-49	diff
The Simpsons	3.21	254260	79119.09	+65%
Family Guy	3.57	264912	74175.36	+54%
American Idol Th	7.10	468100	65971.81	+37%
American Idol Wed	7.82	502900	64311.96	+34%
Sunday Night Football	8.24	512367	62197.23	+29%

As I've said in the past, you really need look no further than The Simpsons and Family Guy to prove that total viewer counts have no real bearing on advertising dollars. These shows have a very high concentration in the 18-49 demo and they're even exorbitantly priced when you line them up using A18-49. When you line them up using all viewers, where they have a fraction the audience of many CBS shows whose ad rates they nearly double, it's even more ridiculous.

But back to A18-49: for the third straight year that I've done this (and most likely going beyond that), The Simpsons and Family Guy are pretty far out ahead of anything else on TV in terms of cost per demo point. (Last year, their "diffs" were very similar at +66% and +53% respectively.) Why? As I said last year, they are sort of a perfect storm, bringing in lots of very young 18-49 viewers (also known as 18-34s) and also lots of male 18-49 viewers. This is a demo that is tough to reach because they watch an infamously small amount of television. And it's a big part of why these shows are still so incredibly valuable and worth the money Fox had to pay the voice actors to secure another two years of Simpsons.

Lowest-valued?

	A18-49	Price ($)	$/A18-49	diff
Dateline Fri	1.42	39610	27881.9	-42%
Cops 8:00	1.42	43555	30728.99	-36%
Cops 8:30	1.58	49735	31425.96	-35%
Supernatural	0.99	32477	32925.99	-32%
48 Hours Mystery	1.20	43400	36069.71	-25%

As last year, it's Saturday shows and a couple Friday shows as well. We'll look at the Supernatural case in a little more depth later in the week, but those un-sexy Friday/Saturday unscripted shows really don't seem to be interesting advertisers in the least. I can only assume it's because those 18-49 audiences who are home on Friday and Saturday are also home the rest of the week (thus easier to reach). I'm kinda surprised Dateline commands such a tiny rate compared to 20/20 ($67,384) considering their Friday 18-49 ratings were nearly identical.

In other words, while I think these numbers support the idea that broadcast primetime as a whole isn't in as bad a shape as the ratings indicate, perhaps they also show that Friday and Saturday nights are in even worse shape than the ratings indicate. And the ratings ain't pretty either. Bad ratings and bad cost per rating equal generally bad times.

So those are the biggies that I can see. Sorry if they're not really too different from what I talked about on last year's post, but I guess it's sort of nice to know that the whole world isn't turning upside down every year!. More to follow later this week!