This is the title of the report Richard Marks made for the IPA, the Institute of Practitioners in Advertising in the UK. It was published on the 20th of June 2013 and it is already a classic. It is compulsory reading for all of us in the industry. So first download it from the IPA site and read it.
The complete title is a summary of the report in itself:
No controversy between Big Data and Audience Research, it is an opportunity to combine the two. The report is littered with quotes from experts in advertising and research from all over the world including the famous Don Draper, which makes the report most readable.
Recently Marks has started this discussion to distinguish “Big Data” from “Deep Data”, claiming the industry currencies are already Big Data as “they strive to represent the big picture”. The report strongly defends the merits of the industry currencies and does not support the revolutionary idea that, in the near future, industry currencies will develop from Big Data. I agree with Marks, but, in my opinion, he could have voiced more criticism of both the current JIC currencies and of the big data believers.
The page 17 with the table of comparison between the two will be a prominent slide in a lot of coming presentations on conference podia.
But in that table the existing (UK) currencies are shone in a better light than they deserve. All UK industry currencies, off course including Touch Points, and including JIC REG get their pat on the back.
Not surprising knowing the background of the author and the political position of the IPA
In the comparison of “passivity of measurement” Big Data get a High and Industry Surveys a Low, that is clear, but why mention between brackets –“BARB excepted” – TAM panel members in the UK do not have to push buttons anymore?
In the comparison of “speed of delivery”, Big Data get a High and Industry Surveys a low with again, a note –“BARB excepted”-, probably because of the overnight reporting, offered by BARB. Marks does appropriately mention the live data concept of some kinds of Return Path Data, and he must be aware of some TAM countries that deliver live ratings instead of BARB.
These small details show that Marks, (or the IPA briefing?) has the tendency to avoid any criticism on the existing currencies. Yet all of us in media research know that we face challenges with response rates and other looming problems, like “the striving for the complete picture” as media consumption on other devices does not lend itself to easy measurement.
Now what about Big Data in the report?
There is a very brief chapter on Privacy issues around Big Data. (but with the most beautiful quote from Danah Boyd, Microsoft: ”Never forget that Big Data is Soylent Green. Big data is made of People”.)
Marks is underestimating the privacy issue of Big Data. It is an easy flammable political issue that travels fast even across borders in our media society. Being a mobile operator, you can argue with your clients and your authorities that the individual data are totally anonymised before they are used for research purposes and that opt in contracts are safe. However one single breach can damage huge investments at the original data owner side and consequently on the easily underestimated investments in “baking a data set into your research business….”at the buyer side.
Another topic that Marks hardly covers is the cost factor of Big Data. There is always an original owner, one step up from the individual who “owns” his personal data and early experiences in investigating big data for media measurement suggest that these owners are very well aware of the value of the data and are asking high prices for it. That puts a serious question mark at the score Low in the cost comparison in the table. ( Marks also makes a brief note on that issue). Recently on the annual EMRO conference experts warned that views are too optimistic that Big Data are
1. Cheap to get, and
2.Easy to merge with classical currencies.
Aside from my critical remarks, the report gives an excellent overview of the opportunities and pitfalls of the future of using Big Data in combining with existing Audience Measurement Currencies.
Return Path Data can give a unique view into the long tale of media consumption of niche channels. Site centric data can give insight in the relative importance of mobile day-to-day usage of news sites, long before we have solved the passive measurement of (all) individual mobile devices. Marks mentions the opportunities possible by overlaying return path viewing data with EPOS (Electronic Point Of Sale) purchase data and our research hearts make a jump.
Marks makes a very good point in mentioning that most often Big Data are “machine data” on a household level more than “individual “data like most of the current currencies. He takes a firm position in that debate against that HH could be a better advertising target than the individual.
Most important is his statement that we have to learn to work with Big Data and that we have hardly experience so far. “For research agencies, there may be a shift to become both creators and curators of data. They have to evolve from data farmers to data chefs, selecting a combination of homegrown and bought-in Big Data to cook the insight meals for their clients”.
So: Don’t wait, giddy up to the IPA site and read it so you can say you’re a trailblazer.
25th of June 2013
Leendert van Meerem