The Revolution Masterclass on web analytics

everything you need to know about your customers will be laid out in
front of you. Thanks to the number-crunching magic of web analytics
technology, you'll know where they came from, what they did on arrival
and whether or not they converted from browsing to buying. Armed with
that knowledge, you can then set about optimising any campaign you
happen to be running and, in the longer term, use the data to improve
the effectiveness of the site itself.

Well, that's the theory and, as analytics systems become more
sophisticated, there is certainly a real opportunity for companies of
all sizes to gather much more actionable information about their online
audience. These days you can acquire a powerful tool in the form of
Google Analytics for no cost whatsoever, and if you want to take things
to the next level, there are plenty of suppliers who will deliver both
the technology and (if required) the expertise to slice and dice
customer data in pursuit of a more profitable online presence. But in
practice? Well, there are real questions as to whether marketers and
website operators are getting the most from the analytics tools at their
disposal.

Ironically, it's probably the power of the systems on offer that creates
the biggest problem - with so much information available, it can be
difficult to focus on the data that really matters. This was an issue
identified by web analytics company Intellitracker when it recently
carried out a research exercise aimed at improving its service to
customers. "We talked to our clients and people in the industry and one
of the things we asked them to do was tell us what they didn't like
about web analytics," says David Hudson, sales and marketing director,
Intellitracker. "Most of them said the tools were too complicated, but
many also told us that they couldn't get the answers they wanted."

In Hudson's view, the findings highlight a very real dilemma. On one
hand, there is a demand for systems that have the necessary
sophistication to provide a comprehensive view of website activity. And,
on the other hand, there is a concern that it is all too easy to get
bogged down in the information that these systems effortlessly produce.
"Users want power and they want also want simplicity," Hudson adds. It's
not an easy circle to square.

The question is then: how do you cut through the noise, find the
information you want and ensure that the time and money expended on
analytics pays dividends in terms of your ability to meet commercial
objectives?

According to David Honan, web administrator at healthcare company
YorkTest, the first step is to tailor analytic reports to the goals of
the business. YorkTest is a small company that sells range of medical
tests to direct to the public. To bolster its web operation, it recently
deployed software provided by US provider Click Tracks to analyse
activity. In Honan's view, to use analytics successfully, you have to be
clear about what you want. "Someone has to have a really strong idea of
what is required from the data," he says. "Otherwise you find yourself
wading through a lot of 'nice to know' information but you don't
necessarily focus on what you need to know."

Know what you want to know

So what does that mean in practice? Garry Lee, client services director
with analytics company Redeye, says it's important to identify the key
metrics that will help you build a better online business. This requires
thought. For instance, if you consider two closely related sets of
figures, one might be much more important than the other. "Let's say
your analytics tells you that the average person is spending three
minutes on your site," Lee says. "That actually isn't very useful. But
if you also know that 20 per cent of your audience is leaving in less
than 30 seconds, then you can see you have a problem that needs to be
solved."

But as Lee stresses, that second piece of information will only be
actionable if you ask why the bounce rate is so high and then set about
finding an answer. There is a number of ways do this, ranging from a
pathway analysis to track the user journey through the site and identify
points of departure, through to a full-scale usability test or customer
surveys

The upshot is that web analysis is not something that can be fully
automated. To make sense of the data, you need someone to interpret the
statistics and then be prepared to ask further questions - a human touch
is vital. "What we've found is that web analytics works best when you
have a vision team that is actively using the information," says
Hudson.

Broadly speaking, analytics data feeds into three main categories of
online business development - campaign optimisation, marketing
optimisation - notably through behavioural targeting and the rather more
amorphous area of analysing the user experience. As Lee admits, it's
still relatively unusual to find companies working across all three of
these disciplines as part of a fully integrated strategy. "We're still
finding that the people who are buying the behavioural tracking and
campaign optimisation tools are different from those who are buying the
user experience tools," he says.

In some respects, that's hardly surprising. Few companies have the time
or the vision to eat the entire analytics elephant in one sitting, and
it's certainly true that mixing and matching a few data sets is probably
the fastest way to secure a quick win. For instance, YorkTest has
improved the focus of its marketing strategy by segmenting visitors to
the site in terms of their point of origin and whether they convert from
browsers to paying customers.

However, this kind of campaign optimisation can quickly overlap with a
more root-and-branch approach to improving the effectiveness of a site
as a whole. For instance, the view of where customers are coming from
can be segmented in a number of ways - by individual sites, banner
location, search engine, region or even by country. Combine this data
with an analysis of what customers are doing and you can begin to
customise your site to specific groups within your larger constituency.
For example, if you have an international audience, you might find that
the landing page works better for UK visitors than it does for their
American customers. Once you know that, you can create a bespoke
microsite for the US customers.

Hudson admits that in general the technology is underused. "There are
lots of high-level reasons to use analytics - such as understanding ROI
on a campaign, but there are many companies still don't realise the
tactical questions it can answer."

The solution: Honan is a firm believer in the power of analytics to
effect change, but thinks that suppliers and providers could do more to
help their clients get to grips with the potential of the technology
through training and education. "I think that's something that the
industry should address," he says.

MASTERCLASS PANEL

David Hudson is sales and marketing director at Intellitracker, a
company that provides web analytics services for a range of
organisations, including Thomson Holidays, Daily Mail and Oxfam. The
service includes software tools and consultancy.

Garry Lee is client services director at Redeye, which combines web
analytics with related services such as segmentation, ad serving, email
marketing and usability. Redeye's clients include Asda, William Hill,
easyJet and the AA.

David Honan is web administrator with YorkTest, a company that sells
blood-test kits online and by mail order. He has been using an analytics
solution provided by Click Tracks to identify potential partners and
focus marketing spend.

THE PRESS ASSOCIATION TRACKS ITS COPY ACROSS THE NET

As a news and information provider, the Press Association's (PA) online
presence is not limited to its own websites. "A lot of our content is
not hosted by us," says head of digital development, Chris McCormack.
"Other publishers take our content and our news feeds and put them on
their own sites."

Until recently, monitoring how PA content was distributed, edited and
ultimately accessed by the public was a difficult, if not impossible,
task. A story produced in the company's London office could be picked up
by dozens of third-party publishers and tailored to fit their particular
house style and audience. And, in addition to legally distributed
stories, the nature of the web meant that material would inevitably be
picked up and used by site operators who were using the material but
paying nothing for it. What the company needed was an analytics solution
that would augment the systems running on its own servers and provide a
view of how content was distributed and used across the whole internet.
The answer came in the shape of US company Attributor, provider of a
system that enabled content producers to monitor their material beyond
their own servers.

Attributor is allowing the PA gain a much fuller understanding of the
value of the stories it is producing. "One thing that the system allows
us to do is understand the digital shelf life of a particular story. We
can also see how it has been edited and how it has been distributed and
assess demand," says McCormack.

And with the PA seeking to open up a new revenue stream through
sponsorship of its copy, the visibility offered by Attributor means the
agency can furnish potential backers with a full and accountable report
on the audience that a particular story or content package is
achieving.

There's a defensive side to this to. The PA lives or dies on the value
of its intellectual property and any violation of its digital rights
amounts to a loss of revenue. "A lot of people use our stories without
paying for them," says McCormack.

"Once we know who they are, we set about getting in contact and
establishing commercial relationship with them."