Neil Mason interview – measuring the effectiveness of Web 2.0 and Social media sites

In this E-marketing Essentials interview I ask web analytics expert Neil Mason for his thoughts on how we should assess social media sites and site channels. Many are talking today about Web 2.0 initiatives, but how do you monitor and improve their effectiveness.

I have known Neil as a fellow web consultant for several years from when we have been speaking at conferences from E-consultancy and E-metrics. We haven't had a chance to catch-up FTF for a while, but I was keen to interview him by email get an update on using web analytics to assess customer participation Web 2.0 sites.

Q1. For me, community, participation and associated user-generated content are a key
part of Web 2.0. Which metrics can site owners use to assess these using a
web analytics package like Google Analytics?

[Neil Mason]: As with many things to do with "€œweb
analytics"€ the answer is "€œit depends"€. Underlying activity levels can be
tracked using standard metrics such as sessions, visits and page views. You can
also look at unique visitors if you can trust the metric.

You then might want to measure and track things like the number of postings or the
volume of comments that are being left. Whether that can be picked up in a web
analytics package may depend on the type of data collection your are using and
the structure of the site.

Q2. There are some useful ratios out there around the proportion of community site
visitors who post original content, comment and just read or lurk (1:100).
Which ratios do you trust?

[Neil Mason]: I think one of the main challenges is to work out what the "€œsignal to
noise"€ ratio looks like. Out of all the activity that might be happening on a
site, how much of it is generating value? What"€™s valuable will depend on the
objectives of the site "€“ that"€™s no different to any other website of course.

If I was running a community site, the kind of metrics that I might find useful might
include, the percentage of registered users that are active and the average activity level per user, i.e. how
many people are engaged and how engaged are they?

Q3. Where do you stand on the page views vs unique visitors visitor sessions debate of
how to evaluate media / Web 2.0 sites. A blend of each I guess?

[Neil Mason]: It"€™s always good to have more than one view of the world.

We all know of the potential reliability problems of using unique visitors, so I would tend
to steer clear of that metric, particularly if you"€™re using 3rd party cookies.
Page views is a traditional publishing metric but is also under threat from the
trend towards using rich media applications within the page.

Visitor session is a clean metric in the sense that it"€™s pretty well understood and
doesn"€™t suffer from some of the integrity problems that some of the other
metrics do.

I would also try and get some feel for activity levels on the site by looking at time
based metrics. However, don"€™t use average time per visit as it"€™s not
representative of the true way that people behave on websites.

Most people visiting a website don"€™t stay very long and then there are a few at the other
end of the scale who distort the figure. Use a measure like the number of
people spending more than a minute of the site

Q4. What about loyalty for Web 2.0 sites? It has long bugged me that "returning
visitors" isn't an accurate or actionable metric. Which other measures can
be used short of full RFM?

[Neil Mason]: That"€™s a bit of a tough one. We are mainly dependent on persistent
cookies to help us track repeat visitors to websites. If the cookie get"€™s
deleted then that causes problems. Other approaches such as tracking IP
addresses etc are less reliable.

The most solid approach is to use a login procedure and to track how many times people
are logging in. However, a login procedure is not sensible for a number of
sites, so there isn"€™t an easy answer I think.

Q5 Many Web 2.0 sites use ads as their main revenue model. What is the best way to
assess "eyeball monetisation" - don't ya hate that expression?

[Neil Mason]: I guess by looking at the bank account? The proof of the pudding is in
the eating!

[Editor]:
OK "€“ maybe that wasn"€™t such a great question. I was expecting Neil to refer to
some standard publisher revenue model metrics such as eCPM (effective amount
earned for every thousand page views), EPC (Earnings Per Hundred Clicks) or for
a subscription site, Revenue per user/active user in a period.