Set Top Box Data: A Directional Exemplary Vivisection & Translation (Part II)

by Mitch Oscar, Oct 7, 2008, 11:30 AM

• Comment
• Recommend (1)
• Tweet
• 

After I published last week's TV Board, I was admonished by a friend -- a charming, very tall, Olympic-swimming Brit: "You of all people should understand that set top boxes are tuning meters. Therefore the use of the words 'audience' and 'viewers' is specious. Tuning is a surrogate for viewers - and, depending on the channel and demo, can be quite good or quite poor."

Another member of the Commonwealth chimed in with similar feedback and illustrative examples for the uninitiated, like myself: "In your example (and using broad thumb-nails sketches of tuning paradigms), the tuning during daytime could have been done by 'stay-at-home mum,' during the fringes could have been done by 'just-home-from-school kids,' while during the evening could have been 'just-home-from-work-and-just-had-dinner working parents.' That is, there could be (but not likely) zero or little duplication in the data as each of these dayparts could have been tuning by different individuals in the homes. Of course, the likelihood that someone who tuned during daytime or early fringe but didn't tune during the evening is low. However, the inverse is not true. There is a high likelihood that there are chunks of the evening tuning that was tuning for the first time that day and was unduplicated with earlier dayparts."

And concluded with an apt Einstein-ian quote: "Not everything that should be measured can be and not everything that is measured should be."

Certainly I concur. We, "medians," must move the industry closer to understanding actual TV viewership rather than projections based upon non-transparent panels and heliotropic assumptions. I was trying in last week's article to stimulate interest in the topic from the vantage point of someone who is not a researcher -- I know what I don't know -- but appreciates the science as an acolyte and approaches the topic as one would a mysterious artifact -- directionally though a little more adventurous than my golden retriever who approaches something new with his snout and two front paws bending apprehensively towards the object of curiosity, while his two hind legs flee in the opposite direction.

A suggestion: Why don't we get the professionals involved? Form a think tank. Engage inquisitors to have a more profound public dialogue about dwell time, unattended tuning, capping sessions, data hygiene and all that other arcane terminology. Reach out to persons such as Mike Bloxham, Frank Foster, Jim Spaeth from independently minded research; Joseph Abruzzo, Brad Adgate, Shari Anne Brill, Steve Sternberg from the ad agency realm; Tim Brooks, Artie Bulgrin, Theresa Pepe Falcone, Jack Wakshlag, Charlene Weisler from the contential community; and data mongers OTX, Nielsen, Rentrak, TNS as well as a dataminer or two, such as Acxiom's Joshua Herman. And of course, let's not forget the technologists i.e., Invidi, Navic, OpenTV, Visible World and the entities that platform them: cablers, satcasters, telcos and perhaps, in the future digital terrestrialists. A plethora of benefits, to name a few:

Platform/distributor and ad auctioneer commercial inventory sale organizations could prove the efficacy of purchasing multiple dayparts as well as effectiveness in terms of reach, frequency and efficiencies. Thereby increasing the sale of inventory that is presently less in-demand.

Ad agencies could glean the benefits of purchasing multiple dayparts: augment reach and frequency, diminish overall package costs based upon historical CPM baselines, and expand breath of campaign with additive cable networks, traditional pay TV platforms (cable, satellite, telco), or other televisual platforms offering video advertising opportunities such as broadband, wireless and/or out of home.

Astute analysis of set top box cable, satellite and telco platform set top box audience flow would help platform marketing departments plan and implement more effective upsell bundling and feature services messaging i.e., premium, digital, VOD (pay), broadband (varying speeds), telephony (land and mobile) and commerce.

Comprehension of set top box audience viewing ebb and flow would help platforms/distributors and marketers glean greater understanding of usage patterns for interactive TV applications to enhance value proposition to the consumer.

In the interim...

I dug into my bag of relationships and got my hands on one more week of set top box data -- this time, late summer -- from a U.S. market -- different distributor -- that is situated far, far west -- one less far west than the market previously scrutinized. Again I was curious to ascertain whether viewers that watch a particular TV daypart during a given day, such as daytime television, watch any other TV dayparts, (early fringe, prime, late night) during the same 24-hour period. Again I was fortunate enough to be partnering with a responsive pay TV distributor, who had licensed a sophisticated technology that enabled such scrutiny. However, this time I was aware of the limitations of my quest in terms of the differences between viewers, audiences, tuners and inanimate set top boxes.

Vivisection

The following is the un-cleansed, daypart tuner data I gleaned from my vivisection:

Translation

Let's take the stats from Monday as an example. This data indicates that there were 112,955 set top boxes turned on to daytime programming. Of these 112,955 daytime tuners:

87,113 were tuned to Early Fringe TV programming

96,368 were tuned to Prime TV programming

33,736 were tuned to Late Night TV programming

As before, I was skeptical of these numbers. How could there be more STB's tuned to daytime (112,955) than primetime (96,368). So I once again prestidigitated another chart to challenge its predecessor:

For comparison purposes, let's take the stats from Monday as an example. This data (second column) indicates that there were 112,955 set top boxes turned on to daytime programming. Of these 112,955 daytime tuners:

87,113 tuners of daytime programming also were tuned to Early Fringe TV programming, but in the market a total of 187,975 tuners were tuned to this daypart.

96,368 tuners of daytime programming also were tuned to Prime TV programming, but in the market a total of 273,072 were tuned to this daypart.

33,736 tuners of daytime programming also were tuned to Late Night TV programming, but in the market a total of 94,365 were tuned to this daypart/

In closing, last week I was chastised for mis-using the word vivisection thrice - once in the title, once in the body (Frankenstein reference) of the document and a reported slippage from a previous week. It was pointed out to me that "it meant as analogical counterpoint to the use of 'dissection' (STB data is alive, not dead). It is in very poor taste." In my opinion, TV data as it is currently shared with the media community is dead until we breathe life into it or electrically shock the numerologists into publicly showing off the monster's visage, lest the media folks burn down the castle in fear of the monstrous unknown.

• Comment
• Recommend (1)
• Tweet
• 

Be the first to comment on "Set Top Box Data: A Directional Exemplary Vivisection & Translation (Part II)"

Leave a Comment