An excellent, clear and comprehensive piece about digital audience measurement was published in The Media in November. Written by Byron John from Habari – one of the foremost digital media agencies and, for a number of years, the sole local representative of global businesses like Facebook – the article argues that the IAB’s measurement data, supplied via measurement partner Effective Measure, is “flawed” because it excludes many global publishers. The global players have to date declined to be measured by the IAB and EM.
The central point in the piece, explores whether there is a viable solution to the problem. This solution, John argues, is to make the Effective Measure “panel data” available to planners. The panel – based on a browser plugin that tracks several thousand self-selected users as they browse the web – does not require the participation of publishers and thus cuts through the complex politics involved in persuading global players to allow Effective Measure and IAB to measure their sites.
The piece ends with a call to the “digital industry” to question why the IAB have this data but are not providing it to planners.
Let me emphasise upfront that this piece is factually correct in many of its assertions. The panel data is indeed a valuable set of information about Internet behaviour in our market and is , as a policy of the IAB SA, not released as a part of our data set to the market. John contends that this is damaging the perception of the size and impact of Internet spend and is therefore an impediment to the growth of the industry.
Presented like that, this sounds like a bizarre and self-defeating decision. And it would be if that were the whole story.
But there are two critical points that this piece misses.
Firstly, the presentation and discussion of the validity of the data needs to be challenged. Even if the active panel is 10 000 strong, and even if that is four times as large as the TV measurement panel, the sample bias cannot be underestimated. This is an entirely self-selected group, on desktop only. To imagine that it can be compared with the meticulous, painstakingly chosen sample group that form the TAMS panel is misinformed.
While one can re-weight the sample in line with AMPS (as we do with our survey data), I dispute that we have enough of a representative group in a panel chosen this way to base our assertions confidently about our audience on it. At best, it is an interesting approximation of web use in South Africa. The audience measurement statisticians would poke very large holes in the collection methods.
That, however, is the lesser of the responses to be made here because these issues can be fixed – statistically, technically and otherwise.
The larger point that the article fails to explore, is one of funding and the local publishing industry. This is something that I frequently find agencies losing sight of.
The IAB is the third iteration of an organisation setup by the local online publishing industry. Since then it has come to represent digital agencies, brands and educators, and operates a Measurement Council that is a Joint Industry Committee which also incorporates other key players. Critically one of the key roles the IAB plays is to appoint and manage Effective Measure and the official digital currency.
However this role is administrative, not financial. Effective Measure earns almost 100% of its revenue in South Africa from local publishers. The publishers pay for the technology to be here. Agencies and other interested parties currently do not pay toward this research. They are simply consumers of the information.
Some global publishers (say, the BBC) have implemented Effective Measure tags, which fire for South African users on their site. I believe John covers this aspect of his argument too quickly: that it would be infeasible for global players to implement localised tags because it would “slow down the user experience”. This is the same publishers who are adept at delivering not only localised but often personalised ads in milliseconds.
I don’t accept that.
The many global players who will not be measured in the local market refuse simply because they don’t need us. Their own ad networks and tracking technologies have more information about the local audience than we do, and they have no incentive to share that. They command over 50% of local adspend and the IAB and Effective Measure have little to offer them. They simply have no need for our currency and ecosystem.
That is their choice – it’s not personal, they operate the same everywhere in the world. However this is where the interests of agencies – who are happy to advertise with these players – and local publishers, who are trying to attract local adspend, diverge.
To expect our publishers to promote the interests of global competitors who are contributing nothing financially to local audience measurement; will not tag or contribute data to the market; and represent a severe threat to the sustainability of their businesses is a big ask.
The argument about the benefits this would accrue need to be developed further. Local publishers are being asked to believe that by showing a more complete view of local audience browsing habits, and emphasising the fact that global players command more eyeballs, this will improve their lot in a cumulative sense. A ‘rising tide lifts all boats’ concept.
It’s fair to say that thus far local publishers aren’t buying it. Particularly as the PwC Adspend Study, which we conducted this year, and which absolutely could change buying behaviour among advertisers, struggled to get the support of many large digital media agencies (Habari, to be fair, did contribute). By contrast, just about every local publisher reported their numbers.
So, from an IAB perspective, if agencies want the panel data to be released there are three options (not mutually exclusive):
- Come up with a more persuasive argument that shows publishers how releasing this data – data which they pay for from a panel which they effectively recruit – will benefit their businesses.
- Make a fiscal contribution to Effective Measure and audience research. Whilst SAARF has always been funded by media owners, it’s true, and thus made data available for free to planners, this is not without a carefully constructed funding model to which the agencies were a close party. If the IAB can levy digital media spend in some way, or if the agencies were simply prepared to contribute financially, they would be in a stronger negotiating position. Getting the data to be statistically sound requires considerable funds which the IAB currently does not have.
- Persuade global players to tag up or provide their data to us. The amount that SA agencies spend with these companies is not insignificant and if we stand together as an industry – rather than creating a stand-off within the industry – perhaps we can bring them to the party. This isn’t going to make local publishers’ businesses flourish, but at least everyone is playing by the same rules.
The IAB is not a body that needs to be tackled by the industry. With 250 members it is the industry, and we have active and productive councils in which these debates are alive and happening. It’s your organisation and the best way to change the status quo is to get involved.
Jarred Cinman is chairman of the IAB SA.
