CAMPAIGN-I SPOTLIGHT ON: Measuring web activity - How do you count who is doing what, and where, on the web? Alasdair Reid delves into the difficulties of trying to quantify traffic on the web

Don't ever let anyone try to convince you that it's easy to measure

activity on the web. They do, of course. All the time. But as last

week's JICNET announcement showed, these are murky waters. Clear as

soup, in fact - alphabet soup. If acronyms trouble you, look away


JICNET is not to be confused with JICWEB, the joint industry committee

recently in the news for having successfully contributed to the

establishment of definitions in a whole range of metrics - metrics which

have now been officially adopted by the International Federation of

Audit Bureaux of Circulations. The members of JICWEB include the

Internet Advertising Bureau, the IPA, ISBA, the Newspaper Publishers'

Association, the Newspaper Society and the Periodical Publishers


JICNET is slightly simpler in its constitution, being backed by merely

two of the above - the IPA and ISBA - but it has similarly ambitious

goals, albeit in a slightly different neck of the internet woods.

There are basically three ways of looking at research into consumer

activity on the web - site-centric, advertiser-centric or

audience-centric research.

With the first -the ABCe figures - you may well know who does what when

they reach a site but you tend not to know where they've come from and

where they go when they leave. That's what agencies do in

advertiser-centric research when they monitor the behaviour of those who

respond to banner ads and this sort of information is priceless in

building up an understanding of consumer behaviour.

However, both these methods are snapshots - they're absolutely accurate

but ultimately limited because they track behaviour for very brief

slices of time. You can't shadow your customers (or potential customers)

for ever.

For any sort of a glimpse of the bigger picture, you have to take a

small slice of the entire user universe and look at what they do all (or

most) of the time. You have, in short, to set up a panel survey much as

you'd find in old-fashioned media such as television (the Barb audience

panel) or press (the National Readership Survey).

This being the web, however, there are three such panels run

respectively by MMXI Europe, Nielsen Netratings and Netvalue. They each

offer something slightly different. Netratings offers the greatest

detail and ability to analyse the figures.

The issue is that JICWEB wants this area to be more like Barb in TV.

In short: one panel overseen by an industry committee. A good idea? Some

don't think so.

Wayne Arnold, the managing director of Profero, says that there are

potential problems in trying to impose order when the industry may not

be ready for it. And there's a certain irony here, he suggests: "In our

attempts to prove to people that we're a mature medium, there's a danger

that initially you will confuse people. I think it's significant that

the industry is trying to go out there with one single voice - but I

also think you have to be realistic about what you can achieve. We are

really only at the first stage. That doesn't mean that these initiatives

aren't significant - what you have to remember is that this is an

industry that has basically come from nowhere in two-and-a-half years.

We're now making slow and steady steps."

Andrew Walmsley, the chief operating officer of i-Level, agrees that

there is a danger of acronym-induced confusion. He says: "Obviously, the

goal is transparency and reliability in the industry. None of the

available measures in isolation will provide you with everything you

need. There are always gaps. But it's important that people don't see

JICWEB and JICNET as being in competition."

What do the three existing panel studies think of all of this? Surely

the analogies with Barb are potentially alarming - or, at the very

least, represent as much of a threat as an opportunity. With Barb, a

committee draws up a technical specification and then research companies

are invited to pitch for the business. MMXI, Nielsen Netratings and

Netvalue have been making noises about finding ways of co-operating more

closely with each other and with the advertising industry. Privately,

they are happy with the status quo.

Tellingly, media owners appear to be similarly inclined. Caroline Pathy,

the ad sales director of Freeserve, worries about the tendency of the

industry to generate infinite numbers of committees: "It's important

that we get clarity with regards to all the various metrics and look at

this going forward and we are part of the IAB's work on global

standards. Shouldn't the IPA and the IAB talk about this? The fear is

that you create yet another level of committees talking about the same

things. Does that add clarity? And the other point is that imposing one

measure doesn't always add clarity. It's not always as simple as


Pathy implies that this is an issue that will be driven forward by media

owners, which is not only appropriate but inevitable. "A year ago, media

owners, agencies and advertisers didn't have media research people in

this area. Now they do. That is significant. We conduct proprietary

research too. I have to say I am not in favour of a Barb-type approach.

I like the competitive situation where we have three different sources.

We need different measures to understand the various parts of our

business. You can't take a simplistic approach to this," she concludes.