CAMPAIGN REPORT ON NEW MEDIA: Interactive research - Analysing and measuring internet traffic is proving a tough nut to crack, but established and start-up research agencies are competing to develop a valid methodology. By Rachel Oliver

When the US auction site ebay (www.ebay. com) proudly announced recently that it had received a whopping 1.5 billion page views on its site, the industry responded with a mixture of disbelief and hysterical laughter. While no-one would dare accuse ebay of out-and-out lying, the methodology that had been employed to reach this figure means the end result is incredibly misleading. To illustrate the point, one bright soul even managed to calculate that it would take every member of the web population worldwide to view 4.92 (or thereabouts) of ebay’s pages for this figure to be even near accurate. Problem was, no-one could prove ebay wrong.

When the US auction site ebay (www.ebay. com) proudly announced

recently that it had received a whopping 1.5 billion page views on its

site, the industry responded with a mixture of disbelief and hysterical

laughter. While no-one would dare accuse ebay of out-and-out lying, the

methodology that had been employed to reach this figure means the end

result is incredibly misleading. To illustrate the point, one bright

soul even managed to calculate that it would take every member of the

web population worldwide to view 4.92 (or thereabouts) of ebay’s pages

for this figure to be even near accurate. Problem was, no-one could

prove ebay wrong.



One of the problems with website traffic measurement is that it can be

extremely unreliable. In the absence of compulsory ABC audits in the UK,

for instance, (only around 130 sites in the UK have actually been

through the ABC treatment), sites can pretty much quote whatever

audience figures they like, based on whatever methodology suits them

best. Many sites have been known to quote the pages viewed on their

affiliate programmes, for example, to bump their own figures up.



As a result, page impressions are no longer a universal means of

measuring a website’s success. It is far more valuable to understand the

behaviour of users on the sites they visit - how long they spend there,

where they come from, where they go to and so on. A site simply

measuring its success by the number of people that visit it is like a

shop measuring its success by the number of people who walk through its

doors.



There is a number of research operations working on furnishing the

industry with more information on how popular the websites are, as well

as how effective online ad campaigns are being. Fletcher Research, for

instance, provides the UK’s only web advertising monitoring service in

the form of Adwatch.



Adwatch tracks more than 1,500 advertisers on around 265 UK sites and

combines the data with traffic statistics and ratecard information to

analyse commercial adspend.



Research companies tend to fall into one of three categories -

datahouses, syndicated research houses or consumer research companies.

MMXI, NetValue and Engage fall into the first category; Jupiter

Communications, Gartner Group and Forrester Research fall into the

second, and NOP and eMORI, or smaller qualitative operations such as

Netpoll, into the third. ’Old’ research operations such as Research

International, Millward Brown and BMRB have carved their place in the

market, as have the vertical players that concentrate on particular

industries such as pharmaceuticals. Established dedicated new-media

research companies such as Jupiter Communications do not tend to view

them as infringing on their space, however. ’We regard them less as

competitors and more as acquisition targets,’ Jupiter’s UK managing

director, Phil Dwyer, says.



One of the latest ’old’ houses to announce it will break into the market

is the US TV ratings agency and established audience measurement company

AC Nielsen. AC Nielsen, which has just bought the UK player Netratings,

employs the same methodology as MMXI and NetValue of selecting panels of

consumers, then giving them software to download on to their computers,

which tracks their behaviour online. The recorded activities are then

used to represent that of the entire web population. MMXI says it tracks

up to 4,000 home users in the UK, though some in the industry believe

the size of the control groups being tested is far too small. MMXI has

stated that it will increase its groups to 10,000 in the near future,

however.



These types of web panels are selected in one of two ways - either

through random selection (random digital dialling in MMXI’s case), or

specific targeting (as in NetValue’s case), where groups of people are

selected based on age, sex, socio-economic status and so on. Currently

these types of panels in the UK only tend to target home users although

workplace panelling is expected to be introduced into the market later

this year.



David Jaffa, the founder of the online education tool SAM Learning,

believes there is much to be said for being able to monitor the

behaviour of site users through the kind of tracking software that

companies such as Engage provide. ’There is a big application service

provider opportunity in this field,’ he says. ’As you can buy an

off-the-shelf e-commerce product, there is also an opportunity for the

likes of Forrester to offer an ASP-type product.’



It is pretty widely accepted in the industry that tracking actual

activity as opposed to telephone/e-mail/pop-up window interviews is a

far more reliable method of collating research on web users’ behaviour.

The perceived problem with consumer research-style interviews is that

people in general do not answer the questions truthfully, either because

they do not understand the questions they are being asked or they simply

cannot remember what sites they have visited and when. Furthermore, for

research on non-web users, it is very difficult to gauge what people

want or expect from a site if they have not been there.



In terms of its quality, the range of research so far produced has been

described by one house as ’patchy at best’, particularly with regard to

any serious efforts made to predict wireless application protocol or

interactive TV usage, which, Dwyer believes, is ’pretty impossible to

predict’. Research houses though will admit that mobile commerce is the

next big thing for them in the UK and software is being developed in the

US that will be able to measure internet usage through mobile phones.



Become a member of Campaign from just £46 a quarter

Get the very latest news and insight from Campaign with unrestricted access to campaignlive.co.uk plus get exclusive discounts to Campaign events

Looking for a new job?

Get the latest creative jobs in advertising, media, marketing and digital delivered directly to your inbox each day.

Create an Alert Now

Partner content

Share

1 Job description: Digital marketing executive

Digital marketing executives oversee the online marketing strategy for their organisation. They plan and execute digital (including email) marketing campaigns and design, maintain and supply content for the organisation's website(s).