What you see is (not always) what you get; why return path data is a bad idea
November 2, 2020 11:04 AM
The allure of RPD stems from a notion that there is no reason to keep using panel data of tens of thousand homes, when it is possible to lay hands on actual viewing behaviour of millions
RPD is fraught with privacy issues.
By Paritosh Joshi
Decades ago, I had a conversation with a very senior physician, father to one of my dear friends from college. I didn’t know it then, but it was to be a critical praxis, which continues to inform my understanding of, well, pretty much everything.
“What is the process you follow with your patients?”, I asked him. “Six simple, iterative steps”, he said. Inquiry. Examination. Diagnosis. Prognoses. Prescription. Review. And so back to Inquiry. Those six words warrant expansive examination but we shall not be distracted, and take away only what is pertinent to our subject today.
You never prescribe before you have inquired, examined, diagnosed and laid out the possible prognoses.
Which, as we all know, is alien to our way of doing things. We can no more hear a problem before we offer a prescription, and insist that it is panacea; that problems will “disappear”; just like COVID-19, as understood by Mr. Trump.
Measurement systems, just like all human endeavour, are, at all times, works-in-progress. There is always a new problem to solve, a new vulnerability to patch, a new opportunity to address. BARC is no less. However, problem solving must proceed with the deliberateness of a competent physician. The cautionary proverb, “act in haste; repent at leisure”, should give us pause before we issue prescriptions. Which never gets in the way of people of heft using their bully pulpits to fire them off.
The nostrum du jour? RPD or Return Path Data, from digital cable and satellite service providers.
There being no polite way of saying this, let me say it bluntly.
Big isn’t necessarily better (but, as E. F. Schumacher told us, “Small is Beautiful”)
The allure of RPD stems from a notion that there is no reason to keep using panel data of tens of thousand homes, when it is possible to lay hands on actual viewing behaviour of millions.
Consumers who buy C&S connections aren’t, and don’t have to be, forthcoming about the composition of their households, and will never be asked questions about their consumption behaviours.
Demographic data about households obtained without seeking explicit permissions is illegal. Even when such permissions are granted, privacy regimes limit their usage. GDPR regulations, and their proposed equivalents in India, the proposed Personal Data Protection Bill, stipulate demanding conditions on not merely how such data can be gathered, but who can use, distribute and store it, and for how long.
A set top box with a return path is not much more than a special purpose microcomputer with a fixed or dynamic IP address. It works the same no matter where it is installed. Indeed, you don’t even need a physical set top box to do what an STB does. It’s called STB emulation and you can readily download dozens of variants from the interwebs. Think about this for a moment. All it takes to spoof a functioning STB is a program which you can pick off the digital commons, tweak as per your specs, and voila!
C&S presents a diverse, heterogenous, landscape. All syndicated research shows that Prasar Bharati’s FTA, free-to-air DTH platform, DD Freedish, is by a mile, India’s biggest digital distribution platform. Just because it is free doesn’t mean that it is only used by people at the bottom of the socioeconomic pyramid. Sample this. Entertainment channels are now paying Rs. 15 crores, Film channels, Rs. 12 crores and News channels over Rs. 5 crores per annum as annual carriage fee on the platform. They would scarcely stump up these amounts if all they got, were consumers living on the edge of destitution. Consumers, particularly in the more remote and inaccessible parts of the country, can get a fairly comprehensive bouquet of over 150 channels covering all our major languages and geographies. Freedish doesn’t have a return path and there is no obvious reason why it would seek one.
A measurement system must report the market it monitors with fidelity and accuracy. Panels are designed to represent the population keeping in mind a number of household attributes. The measurement agency, BARC in this case, establishes criteria to include or exclude a panel home from a week’s reporting, the most important one being continuity of reporting. A household will be included if it has checked in at least 5 out of 7 days. In operational terms, about 80% of all panel homes qualify to be included in the reporting sample for the week. This has an interesting implication for the panel. The weight attached to every household in the reporting panel will vary, as the total number of panels reporting from its segment or geography go up or down. There are evolved weighting schemes, such as ‘RIM weighting’ and ‘CELL weighting’ which enable this ongoing precision. Given the anonymousness and enormous scale of data (ibid), it is not clear what an audience researcher would need to do to weight and report off RPD records.
RPD records what was tuned. It cannot tell who, if anyone, watched. Peoplemeters periodically pop reminders which overlay the program then running on the TV asking if there are viewers in the room, and requesting them to poll the remote. It would be extremely intrusive, even offensive, if a distribution platform were to have such pop ups on its general playout.
What do all these infirmities add up to?
RPD is fraught with privacy issues.
If you think panels are open to fraud, wait till you see RPD.
The biggest distribution platform will never be included in the RPD data.
There is no analytical scheme which will allow weighting RPD data back to the underlying population. This precludes fidelity and accuracy, the two most important parameters for assessing an audience measurement system.
RPD is, sort of, okay at tracking what a household tuned to but not what an individual watched.
All those people who have been talking up the joys of RPD will, in almost the same breath, tell you about the widespread fraud in numbers reported by the major digital platforms. About bots and click farms in poor countries, whose only job is to create fake page views and boost their advertising revenues. No major digital platform has allowed independent, agnostic measurement inside its walled garden. You take what the big A, F and G say at face value, and they will not allow you to peer over their hedge to audit and validate their assertions. Since the widespread abuse is hardly a state secret, advertisers have learned to apply a large haircut to claimed numbers to get a more realistic sense of how many people actually saw their communication on a particular digital platform. Marc Pritchard, P&G’s Global CMO, gave voice to what many were thinking when he said this at an Association of National Advertisers conference in April 2019. “Digital media continues to grow exponentially, and with it, a dark side persists, and in some cases, has gotten worse. Waste continues to exist from lack of transparency and fraud… Privacy breaches and consumer data misuse keeps occurring.”
The old proverb says. “Sow a wind. Reap a whirlwind”.
Still wondering why this essay has the title it does?
The author is principal, Provocateur Advisory. He serves on the BARC Disciplinary Council.Views expressed are personal.