Let’s Calculate Television Program Ratings a New Way

It’s something that just about anyone who watches television has had to deal with: the cancellation of a show (or shows) that you love. And it happens every.single.season. Everyone you know watches a certain show, and just when it’s getting good, bam it’s gone. Gone because “no one” or “not enough” were watching. It’s aggravating as all get out because you don’t even get to register a vote in the matter.

Basics

Ratings exist for one purpose and one purpose only: to provide statistical information about which shows that are airing are going to be the most profitable with advertisers. It’s not about art or craft or taste. It’s pretty much just about money.

To this end, any number of ways of collecting the data have been employed. The current version of the leader of TV ratings, Nielsen Media Research, involves hooking boxes to the TVs, VCRs, Cable Boxes, Satellite Receivers, etc. that will transmit data about when the device was turned on and what channel it was turned to. They also have “people meters” which help give demographic information about who is actually watching the program.

In a world with three (or four) major networks plus PBS and an independent station (typical about 30 years ago), this system worked just fine. However, as channels have multiplied and methods to watch have diversified, this system is likely no longer doing the job it’s intended to do. In fact, once cable channels started filling up the dial, the actual usefulness/accuracy of the Nielsens became increasingly suspect.

At the heart of the data gathering is the statistical sample of homes Nielsen uses to extrapolate its numbers. For the nearly 100 million television households there are just 5,000 households whose choice of programs to view matters. Each one of these “Nielsen Households” represent about 20,000 total viewing households. Personally given the breadth of options currently available, I find this coarse level of granularity appalling.

In our current media environment there are hundreds of entertainment channels plus hundreds of Pay-Per-View channels. These shows can be watched live, on a same-day rebroadcast, or time-shifted. Shows can be accessed via television, VCR, DVR, web sites, cell phone subscriptions, etc. While ratings numbers for all of these variables are made available via various services, clearly the shear breadth of choice within an increasingly mobile society informs us that the centrally convenient Nielsens are ill-equipped for the current reality.

A New Path

The time has come to break away from the old ideas of how to gather viewer data. Instead of measuring households in a semi-passive/semi-active way, why not have the breadth of viewers who have a stake in the ratings outcomes take matters into their own hands?

My suggestion uses technology that is currently available, can be scaled, doesn’t require viewers to watch in only one location, and will gather data for shows from a variety of broadcast choice (both live and time-shifted), and will also provide fine-grained demographics. It also puts a little bit of power for those niche (or “cult”) shows whose viewers might be more active and thus ultimately more attractive to advertisers.

The core of my suggestion is a portable random password generator (PRPG). These devices have been around for about a decade. They allow remote computer operators to gain access to secure systems without having to rely on static passwords that can be broken with enough brute-force effort. These passwords changed every minute (or so) on some receiving device and are keyed in only when needed.

Using such a tool, a viewer can send data to a ratings gatherer securely, over Wi-Fi or cellular networks, with no more effort than touching a button. But that alone will not generate the data that is needed, all that does is inform the data gatherer of the identity of the sender. No, for program data there needs to be some keying of the source.

Since the purpose of ratings gathering is advertising revenue, I propose that at random times during a program and commercial breaks (no more than 50% in commercials) a code is shown that can either be entered via a keypad on the PRPG or can be scanned (as like with a barcode). This code would be sufficiently short that it not be onerous, but provide enough information about program, source (broadcast, web, with/without ads, etc), and possibly time. At least three codes per hour, or two codes per half-hour program. These codes do not have to be entered every time they appear (that would be extremely tedious), but if you want to record that you actively watched a certain program, then you can participate.

Relatively simple mathematical pattern match can be employed to ensure that the system isn’t “played”. Also, repeat viewing can be noted as supplemental data, but aren’t included in the main report (this helps to prevent corrupting the primary data).

When the code is entered, you press your PRPG send button and the code plus your password are sent to be counted.

Again, I have to emphasize that for this system to work well, the ability to enter the program code has to be as easy as possible, which is why I favor scanning as an option (with the always allowed option of manual entering of the code). From an advertiser’s standpoint, I’d be most interested in those people who enter commercial codes as well as people you enter the codes in a reasonably time-separated sequence to demonstrate that the entire program was viewed. To do that, the entry of the data has to be as easy as possible.

Because of the ubiquity of cell phones, it wouldn’t be long before the PRPG functionality is incorporated into them, making the process just that much easier.

The Populist Ratings

Who gets these devices? Everyone willing to fill out the demographic information form and send in a small fee to cover insurance on the device as well as the brief transmission access fee… say $20/year for discrete devices and free for cell-phone (or similar tech) devices.

From all the data that comes in, the mining can reveal to advertisers what the true broadcast landscape is. While it’s likely that data-gathering for time-shifted shows will likely become invalid after some period of time (week/month/etc) so that the codes can be re-used, it still provides people who simply don’t have the occasion to watch a show live to have their voice heard.

This system allows the viewers to actively participate in keeping their shows on the air. Sure, there will be some SNAFUs at first as people of “popular” shows deem their show safe and don’t enter data while every single fan of a “cult” show votes their mind. The numbers would indicate that the “cult” show has more viewers. Well, it certainly has more active and passionate viewers. It will only take the cancellation of one or two “popular” shows for everyone to catch on and the ratings will stabilize as everyone starts to participate. How great will it be when you are stuck in some hotel room on some business trip five states away from home and still be able to let your vote count for your favorite show?

The time has come to stop depending on the statistical few to have the same tastes as the variety of viewers across the country. When they are picking from half-a-dozen choices, that might have had some merit, but when picking from about 1,000 choices, those 5,000 Nielsen families seem woefully under-representative. With just a new twist with available technology, there can be upwards of 200,000,000 votes as to what is worthy of being on the air and what isn’t. Whether watched live on DTV, or a week later off your laptop’s hard-drive, you can tell the powers that be that you are watching.

The time to change is now. If my favorite show is cancelled, I don’t want it to be because four people in Iowa prefer to go to bed early on that night. I want to be able to fight for it. I want my friends to be able to fight for it. Advertisers, broadcasters, let us fight for it.