Spanish newspaper El Pais is suing Nielsen for an ‘unjustifiable’ downgrade in the number of unique users it attracts to its website. Ouch.

El Pais, owned by Prisa, is annoyed at the downgrade because media buyers rely on panel data provided by firms like Nielsen and comScore to determine where to place ad campaigns. The BBC has the lowdown here.

It seems to me that this reflects a bigger issue relating to the blind faith in panel data among advertisers.

Media buyers tend not to apply a pinch of salt to the numbers cited by the panels, which vary wildly, and which can be inaccurate to the point of amusement. This is problematic and needs to change. Yet there is no single source for accurate data, so until that gap is filled then the panels will remain popular.

Prisa itself cites Nielsen data in its own press releases, such as this one (which claims attracted more than five million users to its site in the month of July).

You have to wonder why a publisher would use Nielsen’s numbers in a press release. Presumably, this is for one of two reasons:

  1. The panel data is actually higher than their own analytics data (which is almost certainly more accurate).

  2. The release is primarily aimed at media buyers, who take rather too much notice of this sort of panel data to make their decisions.

The problem with the first reason is obvious – if positively skewed, the inflated usage numbers would provide a false sense of reality to media buyers. I’d love to know how Nielsen’s data compares with’s own in-house Google Analytics or Hitbox data.

The problem with the second reason is that it sustains the flawed practice of buying media on the back of panel-based data. Prisa alone can’t change this industry-wide issue - and there is certainly some value in speaking the language of the lazy media buyer - but ultimately this sort of thing is supporting and strengthening the status quo.

We’ve been making these noises for years (about the need for a joined-up and accurate media buying platform) but so far nothing much has happened.

Hope may be on the horizon in the form of Quantcast, a great site that has improved upon the Hitwise model (Hitwise doesn’t use panels, but instead analyses anonymous data provided by ISPs).

The big difference with Quantcast, compared with Hitwise, is that publishers can install a little bit of code to provide Quantcast with some basic analytics data (eg unique users, page impressions). In doing this a publisher becomes ‘Quantified’ and based on what I’ve seen, the data becomes much more accurate once you do that. It provides a fairer representation of your actual numbers than might be achieved via a panel.

If sufficient numbers of publishers sign up then maybe Quantcast will emerge as the go-to place for buying media, especially if they add buying and campaign measurement tools. It remains a big ask. The real change needs to happen in the minds of the media buyers.

While I will stop short of predicting the absolute decline of panels, I do think that we need greater transparency between advertisers and publishers. And with that in mind, I'm not at all sure that panels help in any respect. They may actually hinder the decision-making process. 

If you're an advertiser you should always question the value of panel-based data if you’re going to spend a small fortune on an online ad campaign. Where does the data come from? What kind of person signs up to a panel? How big is the size of the panel? Where do these panellists live? What age are they? How does the panel data measure up against other data sources? And so on...

And if you are a publisher - such as El Pais - then you need not throw rocks (much less lawsuits) at Nielsen on the grounds of accuracy. If you want accuracy just release your actual numbers, as measured by some proper web analytics software. Or, go and get Quantified.

Chris Lake

Published 28 September, 2007 by Chris Lake

Chris Lake is CEO at EmpiricalProof, and former Director of Content at Econsultancy. Follow him on Twitter, Google+ or connect via Linkedin.

582 more posts from this author

You might be interested in

Comments (2)



You're kidding right? What do you mean when you say 'release your actual numbers'? Everyone has a different definition of a pageview, session, unique user, etc. It's easy to manipulate you own statistics.

Independent third parties like Nielson are used because they are, in theory, impartial. I can't comment on Nielsen's methodology but publishers and
and advertisers will always need someone they can trust in the middle of the relationship.

almost 11 years ago


Laurent Nicolas

Merging panel-based and site-centric data is a huge task. Panel-based data can be higher than site-centric ones in a few cases:
- when the panel is biased (too many visitors of a given web site)
- when the panel is weighted (the number of visitors of the web site is correct, but they belong to socio-demographic groups that are too few in the panel, so the weighting will raise the number)
- when the site has a lot of home visitors (if several members of the same computer visit the web site, they are identified as several individuals in the panel, but as a single visitor in the site-centric tool).
I used to work for NetValue and NetRatings, you can find more details (in french, learn french, it's fun!) about the comparison of the tools here:
The problem for advertisers online is:
- they plan their marketing strategy on socio-demographics data (age, gender, etc.), better measured by panel-based tools
- they pay the result of this strategy on actual banner deliveries, better measured by site-centric tools.
So, no system is perfect!

almost 11 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.