In the early years of the Internet, media professionals were very optimistic about online research. On the Internet, everything would be measurable. The possibility of passive measurements would make a proper audience measurement a piece of cake.

By now we know better. And we think back with sympathy to the times media behaviour could be measured through questionnaires alone.

First generation

When setting up the first edition of online audience measurement, this euphoria was quickly gone. The measurement of unique devices (browsers) was not difficult. But how do you create a good reporting of the number of people visiting a site or page?

Setting up a representative panel was hard enough. How do you find people who want to have their entire online behaviour tracked? And are these people representative of the whole population?

By trial and error and inspiration from abroad, a valid research method finally succeeded. The challenge was to gain trust for the output. The media researcher had a tough job explaining the huge differences between the audience measured (lower!) and the analytics reported by their ICT colleagues.

That’s a big challenge in an industry where marketers like to provide the highest numbers possible.

Disruption of mobile devices

We are looking at the period of roughly 2005 to 2011. Those were the days of the fixed PC and laptops. With the fast penetration of smartphones and tablets, the research design of the first generation became inadequate.

I remember an INMA World Congress back then (!) where one of the leading news sites told us that, at that moment, more Web traffic came from mobile devices than from computers and laptops. It was obvious that the new mobile devices had to be included – simply because half of the Internet traffic was out of scope.

Relatively quickly, new research for a new online media currency was designed. This time the funding appeared to be the challenge. In bad economic times, publishers had to allocate more budget for new research. That was a hard decision because advertising revenues for mobile devices was still very modest.

In early 2014, a new valid method was in place. The initiative was called Dutch Digital Media Measurement (DDMM). The first cross-device audience measurement was ready for use.

The results were startling. Many media brands showed they had doubled their audiences, thanks to the additional performance of mobile devices. What was striking was that not all publishers brought in profit from an audience boost in the same proportion. It was clear that some media brands had done a better job on mobile than others.

Sudden collapse

DDMM couldn’t profit from its state-of-the-art set up for very long. With the introduction of Apple’s iOS8 in fall 2014, a considerable part of the measurement disappeared. The new operating system didn’t show any how iPhones and iPads exactly were used anymore.

This issue was not solved quickly, and the second generation of online audience measurement had its best moments.

New opportunities

The implosion of the measurement also brought new opportunities. Many publishers have become multi-media players that wish to integrate the online measurements into audiences of other media channels, including radio, television, and print. As a consequence, in the first half of 2015, a new future-proof audience measurement was prepared.

The largest media companies in the Netherlands took the lead, amongst which were many leading news media. This research was labeled NOBO, Dutch Online Audience Measurement.

The demanded specs are very ambitious. The wishlist is demanding. The new measurement must not only report online surfing, it has to include online radio listening, online video watching, and online reading.

In short, the new measurement should not serve only pure players, but also news media, broadcasters, and radio stations. And, if this challenge was not ambitious enough, global search engines and social media also had to be reported in a valid manner.

Research design

For this assignment, research firm Kantar was selected out of three global research companies. Clients are VINEX for Internet media, SKO for video (television) audience measurement, and the JIC, a joint effort of media owners, advertisers, and media agencies. Kantar has started building a panel that consists of entire households in which all online devices are registered.

The research design is based on a tagged solution, but it also includes 2,000 untagged respondents to report audience data for international media owners that do not tolerate tagging.

The existing media currencies will take full advantage of the research set up. Broadcasters will use NOBO to complement the data from their television panel. News media will be able to supplement print audience with the reach of e-newspapers, news apps, and sites. And, online radio behaviour will be used to top up the broadcast radio audience.


So, NOBO will be a research project that connects old and new media — with and without tags, audience and census data. It’s all very ambitious, but not impossible.

Kantar combines research components that are already being used in various countries. Despite this, the new set up has the most advanced and bespoke design, fully customised for the Dutch market.

The research company and Dutch media currencies are aware that the whole world is watching this project. That’s why the best possible team is selected, and the best resources are available.

But the work must not only be carried out by Kantar alone. Media professionals with different backgrounds will have to cooperate closely. Media researchers and data experts will all have to use all of their knowledge of data fusion, calibration, and tagging. Offline and online experts, large and small publishers, and print and broadcast marketers are all striving to make this holistic audience measurement work.

It is an exciting project. And it has to happen this year. It is a cliff hanger, and I will keep you posted.