Sample management is the selection and tracking of households, STBs and individuals in the sample—the group of households or STBs capable of returning measurement data or, less commonly, a selected subgroup of these. In some pay-TV systems, such as those using both physical and logical STB or smartcard identifiers, just ensuring continued measurement and identification of the same STBs poses technical challenges.

This area is one in which classical TAM and RPD practices differ greatly. TAM providers can afford only a small sample because of the high cost of recruiting and metering each household. Furthermore, because they quite consciously do not control their samples by income, education or occupation or even geography at a scale granular enough not to be overly diverse, but only by large geographic area, age, and gender (a practice that leaves a Boston corporate lawyer and a Chelsea, Massachusetts bus driver who never finished high school looking exactly identical), they overly rely on the randomness of the sample household selection as a panacea for those failings—which it is not, at those predetermined low sample sizes. The TAM provider will make several efforts to recruit the randomly selected household, and if it fails, will turn their attention to its nearest neighbour. Copious empirical evidence has shown that a different sample recruited in this manner to the same composition by the few tracked attributes, after a year to shake out early participants (which have an increased tendency to leave), can return completely different results, favouring different channels, than the preexisting sample, and that this process can be repeated endlessly with the same outcome.

RPD, on the other hand, is founded on the strategic advantage of much larger samples due to the default availability of measurement collection software on STBs, with no marginal cost beyond connection of an installed STB to a return path, which can be usually justified by business benefits (access to broadband VOD and interactive services for higher-end STBs, and to the latter via phone for lower-end ones). Thus, RPD operators usually accept all the sample they can get. The only cases in which they don’t are when their maximum samples are so large that their computing hardware and software are insufficient to process their data. If the operator understands that a higher sample, especially in the most fragmented viewing landscape ever, is an inherent good from which stem many other benefits, it will not permit such a situation to persist.

RPD operators can balance their samples using a scientifically accepted multivariate algorithm that avoids the need for the sample distribution to be close to that of the subscriber universe as well as the operational difficulty of populating numerous microcells of various attributes (used by some very simple radio ratings services, for example, but not practical for multichannel television). We own an implementation of such an algorithm, which can be run as often as every day. This can include any native information on the household, such as its geographic location, tenure with the operator, subscription level and equipment, and geographically linked external information such as probable socioeconomic status (income/education/occupation) and even psychographics where available.

A bigger problem, which some TAM operators falsely think they avoid by installing people meters, is identifying the individual viewers using specific STBs or means of OTT access. Pay-TV subscribers in even modest-income countries now have large percentages of subscribing households with multiple STBs, aside from OTT devices. People are increasingly viewing on screens that are individual to them or shared with one other person, with occasional exceptions such as major sports events or series season finales. Surveys to describe specific STB users have been attempted by several of our clients but are so costly they can cover only a small fraction of the sample.

Although no systematic solution is readily available as yet, the most promising development in that direction is the division of a pay-TV subscription into independent accounts for each member of the subscribing household, just like the five accounts per subscription offered by Netflix. This offers substantial benefits to each viewer—his own DVR library, prioritisation of VOD selections, the ability to continue viewing from the point it was stopped on any device as on YouTube for logged-in users, individual playlists and recommendations—and in exchange isolates the STB or OTT device to a particular viewer or two and permits viewer, rather than only household or STB, ratings and demographic ratings, matching the TAM offer with individual viewing data that may be even more reliable than those that are supposed to be collected by people meters—but often are not, with a towel shielding the viewers from the annoying bank of red LEDs that flashes when they miss a designated check-in or check-out.