Commentary
My TV Cord-Cutting Research Is Better Than Yours. Let's Go Down The Rabbit Hole
- by Wayne Friedman , Staff Writer, January 30, 2019
One recent bit of research from Waterstone Management Group, which surveyed some 5,000 respondents: a whopping 59% have cut the cable cord (really!) and another 29% are thinking about it (whoa!).
Waterstone said its data comes from 5,000 respondents in a December 2018 survey who were 18 to 69 years of age, including current cable subscribers as well as former cable subscribers.
The research did not say, from its website, whether the data was modeled to project an estimate for all U.S. pay TV subscribers. We guess it isn’t. Thankfully.
Then there is this: A recent survey from TiVo says 81.2% of TV watchers plan to stay with their current pay TV provider — with 18.8% planning to cut the cord — we suspect they prefer an alternative, digital pay TV live, linear service. Other research from Kagan says cord-cutting is growing at a 12% annual compounded rate, totaling 40 million by 2023.
So there are lots of numbers — and big differences — to consider. Including this: Waterstone says 70% of respondents interviewed from states such as Idaho, Kentucky, Tennessee, Wisconsin, Nevada, Arizona and South Dakota, have “cut the cord.” (Not revealed here: What happened to cable service repair technicians in those states after losing customers?)
Yes, there is some cord-cutting going on. But we need more — granular information from these disgruntled and disrupted media consumers.
From the Waterstone data, Broadcasting & Cable says only 12% of cable subscribers are “committed” to continuing with their traditional pay TV service. It says this number was derived by adding 59% — those who cut the cord — to 29% — those planning to leave. This means only 12% want to remain with cable.
B&C says Waterstone used Amazon Mechanical Turk for its research efforts.
Maybe this is the key. MTurk is a “crowdsourcing marketplace,” according to Amazon, which outsources “processes and jobs” -- some of which can include conducting simple data validation, data collection and analysis, as well as more subjective tasks, like survey participation. Ah. Hah.
Still scratching your head?
The bottom line is that any seemingly out-of-nowhere rapid changes in TV disruption come at a price. And action.
Are many pay TV companies now filing for bankruptcy? Going out of business? Firing cable pay TV technicians? If not, I leave you to read the media tea leaves.
Wayne Friedman is West Coast Editor of MediaPost. You can reach Wayne at wayne@mediapost.com.
I applaud your skepticism. Even when samples are truly random, respondents sometimes lie.
What's amazing about the Waterstone study is that those in charge at this company allowed the press release to go out with those absurd findings which tells me that they are clueless about what's happening in TV and, especially about cord cutting. It is to laugh.
Agree with Ed - croudsourcing always contains a leval of bias in that the pannels are not representative. I saw one such study that was 85% female, but reported as 'adults'. Scarborough reports the number of traditional pay TV households at just under 3 in 4. Logical numbers. The decline they note over the last 6 years is about 13% - also seems logical.
It's not just crowd-sourcing that is the issue (though it can produce hilariously absured results as per the above).
There is an underlying issue of legitimate research companies conducting online research (and not just into online activity, which would introduce a clear bias), by buying lists.
The gross majority of these lists are opt-in (i.e. people who opt-in to get points - I am on a couple of lists just to see how crap the questionnaires are, but thanks for the points). The lists are rarely kept current or even verified, Further, many of these 'respondents' are on multiple lists and de-duplicating them is extremely complex.
We very carefully use these lists for audience measurement purposes where applicable. But one issue we consistently see is that the response rate (bias alert!) is so low that we can exhaust multiple lists of hundreds of thousands of opt-in members in just one sweep. And to think that there are dozens if not hundreds of surveys concurrently in the market. Give me a carefully planned and managed panel any day.
This is absurd.
This is why research professionals engaging in online quantitiative research use quotas/weights--it helps make a sample representative. It is because online research is not truly random that such methods must be used. Without them, you end up with Waterstone-like data.
Machael, even with sample balancing I doubt that you could have saved this "study".
Studies like this belong in the circular file aka trash. I think we would have known if almost 2/3 of cable/dbs subscribers had cut the cord. Cable sub revenues would have collapsed. Financial filings would have shown the disaster. Why would somebody choose to do a survey like this when data already exists on this issue? An embarrassment.