Tuesday, February 15, 2011

Why Have Ratings Been So Low?

Since Thursday, people have been tweeting and e-mailing me asking the same question... what happened? Why are ratings so low!? OMG OMG OMG!?!?!?

Yes, well, people are very passionate. And confused!

I believed I had the beginnings of an actual answer that wasn't "well, on Thursday, there was an abnormally highly rated NBA game on TNT during prime time" or "the Justin Bieber movie came out so all teen girls have ceased to exist outside of the cineplex."

Here's what happened: starting Monday, January 31st, Nielsen changed their methodology. I know, right? It seemed suspicious, but I didn't have a handle on exactly what the change was.

Of course, the more I looked into it, I found it only affects overall HUT/PUT levels (and therefore share - a metric I continue to believe even with these changes it entirely and utterly useless), and not actual program ratings.

Here's what the change is...

Before: Nielsen credited DVR viewing as a HUT/PUT (household or person using television) when something was RECORDED, and no matter when you played it or how many times, your viewing was credited once and only once, and credited to the time period of the recording.

For instance, if you recorded Modern Family on Wednesday at 9pm, even if you played it on Sunday at 3pm - and because it was so funny, again at 5pm - you would be credited as a single PUT on Wednesday at 9pm.

Now: Nielsen is crediting DVR viewing as a HUT/PUT when playback happens, and every time playback happens.

Back to the Modern Family example... your Wednesday 9pm recording watched on Sunday at 3pm then again at 5pm? You will be credited to the overall number of persons using television on Sunday at 3pm and at 5pm. MAKES SENSE, DOESN'T IT?

Here's the catch, and here's why it doesn't actually explain the low ratings of the last week or so: program data remains the same. Recorded programs have always been credited with viewing regardless of when it happens. As above, it only impacts the overall HUT and PUT levels.

All of which is a roundabout saying "darn, I thought I had something when in fact I don't" to explain why ratings have been lower in the last week. This is why scientists test hypotheses.

So, the speculation continues.


Bill Gorman said...

Travis, interesting to read that about HUT/PUT calculations, but as you note that's not going to have anything to do with individual show ratings.

And there's no way multiple viewings of a recorded show is going to count more than once.

It was Valentine's Day! (Friday too!) ;)

Travis Yanan said...

Hey Bill, as soon as I published I was like, "No, that multiple viewing thing increasing ratings further in Live+3 or +7 playback makes no sense" and so deleted it... not soon enough, it seems!

I understand the holiday effect, as well as Friday... but can a big NBA game (that doesn't even score a 2.0) really account for the oddly low ratings this Thursday? I want to figure the mystery of Thursday and the series/season lows out... not having any luck, though!

cas127 said...


I have to respectfully disagree that the Nielsen changes are meaningless.

If I understand your explanation correctly, then it seems like HUT/PUT figures (paid attention to by billion dollar advertisers allocating funds across TV, print, internet, etc.) are going to be artificially goosed due to double counting.

In brief, if I am understanding things correctly, *one person* watching something twice is going to be counted as *two people* for the purposes of HUT/PUT calculations.

This seems like a slimy way of trying to sneak a fast one by advertisers who may not track changes to Nielsen's methodology.

And the general public.

Travis Yanan said...


I would agree (and, my kneejerk reaction, which I edited out of the post was exactly the same) except that the change is not affecting PROGRAM ratings. It's only going to affect overall HUT/PUT levels (meaning the sum of all households/people using their televisions at any given time).

As advertisers buy based on program ratings (okay, technically, with C+3 supposedly being the new standard, they buy based on commercial ratings), not overall viewership levels, the change is meaningless because the program ratings won't see that artificial goosing.

cas127 said...


I agree that individual programs won't be affected, but I guess my point is that Nielsen/TV networks are playing definitional games in order to obscure the ongoing erosion of TV usage due to ever increasing internet usage.

In brief, the change wasn't calculated to increase ratings/revenue for any one show, but rather to shift/claw-back advertising dollars that are migrating to online.

With regard to that, both Nielsen TV Measurement and the networks share a common interest.

I don't have any specific HUT/PUT data but I would be shocked if HUT/PUT aggregate (all network) figures haven't fallen markedly pre and post internet.

Those are advertising dollars lost to another ad sector.

I wonder if the MRC has made any comment on this change.

But if the MRC is really dominated by Nielsen/the Networks (rather than the advertisers) then it may be a case of the fox watching the hen house.

Another thought, a la AOL/HuffPo - to the extent that individual online media companies have aggregated "reach" on a scale similar to TV (measured online by monthly unique visitors, measured for TV by HUT/PUT) then these individual online media companies can offer "single-stop" "reach" equivalent to TV networks (sorta).

I wonder if "one stop" online media companies achieving "reach" parity (sorta) may have played into Nielsen's HUT/PUT change.

Note II: Thanks, btw, for all the hard, ongoing work of putting out the ratings...we appreciate it even if actual thanks are few and far between...