A lazy FIWK discussion on Nielsen ratings or something like that

Royce: Hey Scott,

I haven't had a chance to put this on FIWK yet, but I found this article about Nielsen trying to capture TV ratings vs. online ratings, etc very interesting:

http://www.economist.com/news/business/21571471-measuring-tv-audiences-keeps-getting-harder-counting-couch-potatoes

While I'm gone it's possible you or Aaron could put it on FIWK and comment.  If not I'll probably do a short write up when I get back.... my main point of commentary was going to be that I cannot BELIEVE how outdated (and small of sample size, 22k houses) the old-school Nielsen 'set-top box' rating methodology is.

 

Scott: uh....i probably should avoid posting on this topic...

I haven't read the article, but the quick answer is 22K US households is plenty to track national tv ratings. We have another 300K we use to track local ratings.

 

Scott: OK - just read; some thoughts:

1. Nielsen is slow; as is the whole TV industry
2. It's incredibly expensive to build out an objective measurement system - especially when publishers & advertisers don't want the same thing
3. Fragmentation will only make it more expensive to build an objective measurement
4. The market demands one measurement system for two reasons:
1. Two different numbers (of viewers, sales, etc.) will cause confusion
2. Any measurement company not in the lead in a given market will not be profitable long-term

Yup - somewhat of a biased response in defense of Nielsen.

Aaron: How much information do digital cable providers have?  I really am ok with them seeing what I watch and record.  Just like I figure that Netflix has some pretty good data on what I stream.  I figure that would be one of the easiest ways to get a lot of data without too much additional infrastructure.

 

Scott: One would think. 3 issues historically though:

1. Until very recently, set-top boxes were one-way streets; no information was sent back to the cable/satellite/telco provider. This required a fairly significant CAPEX effort to change, and was implemented in the last decade. However, they didn't design with ratings in mind as they wouldn't make that much $ off of that.
   
2. Therefore, the returned data in today's set-top boxes is still dirty, so the information being received isn't always digestible and well-understood. Nor can it be presented in a manner that works with ad-buys (e.g. GRPs)
   
3. With a perfect feedback system, you still need an independent third-party to aggregate the information, as each service provider would collect data independently.

On top of all of this is also a whole slew of privacy issues/concerns that may make such a system inherently illegal.


Aaron: Are there privacy issues with digital cable?  How is it different than Netflix?  Or did I sign away any privacy rights without realizing it?  I'm sure digital cable companies could do the same.

I'm guessing there is no incentive for the digital cable companies to collect viewership data.  Unless a company like Neilsen was willing to pay them just for the viewing data of their subscribers.

It just feels like in this day and age of digital cable, we should have a pretty good idea what shows are actually being watched or recorded (which counts for a lot, but not 100% with the ability to fast forward through commercials.  My guess serial shows will begin collecting less ad revenue in the next couple decades, but live viewing, such as sports, will start to collect more).  And flying cars.  Why don't we have flying cars yet?
http://www.youtube.com/watch?v=pp89tTDxXuI

Royce: Ugh I knew this conversation would be interesting. Too bad we couldn't post it on the blog.

Scott (9:01pm): Copy paste!

Aaron (9:02pm): Go ahead and just copy the text into a blog post with names.