Yesterday, I was very fortunate to attend a Web Analytics Wednesday at These Days in Anterwerp, Belgium. These Days defines itself as a "brand interaction company" and is a communication and interactive marketing agency, based in Antwerp and Amsterdam. Close to 50 people attended the event, which is a very significant turnout for the young Web Analytics industry.
After an introduction on building a Web Analytics community in Belgium by Michael Notté, country manager for the Web Analytics Association, a case-study on Pioneer Europe was presented by Siegert Dierickx, Web Analytics Manager for These Days.
These Days manages 20 country web sites, 1 Vlog and multiple pan-European micro-sites, as well as E-marketing campaigns for Pioneer Europe.
Siegert explained that they focus on 2 main KPIs for Pioneer's web sites:
- Sales Leads, measured by the number of uses of the shop locator
- Brand Interaction Hour (BIH), measured by the total number of hours spent by the users on a Pioneer Web site.
In order to make the brand interaction metric more tangible to management, the total number of hours were converted into the total number of customer service representatives needed to achieve the same number of contact hours on the telephone. Wow, so smart!
The thinking behind this is that Web sites are conversations with customers.
An operational metric was then created to measure the effectiveness and cost of campaigns: Cost per Brand Interaction Hour (number of brand interaction hours generated by the campaign divided by the cost of the campaign).
To their surprise, the cost per brand interaction hour ranges from 2 to 2,000 Euros depending on the campaign. Thanks to this indicator, they have been able to optimize their campaign strategy and remove campaigns where internauts spend on average 5 seconds on the site.
On my way back to Lille, I was thinking. Hey, why don't we use the same methodology for support section of web sites? It is usually difficult to measure the ROI for a support section, and consequently these sections are under-invested compared to the branding and product sections.
We could then have a Support Interaction Hour metric, calculated with the total time spent on the support section of a web site, as well as a Equivalent Customer Service Headcount metric calculated with the total time spent on the support section divided by the working hours of an average customer service representative.
I can't wait for my next Web Analytics Wednesday event in Belgium!
To participate to the next Web Analytics Wednesday in Belgium, please join the Linkedin and Facebook groups.
More to read: Michael Notté's and Siegert Dierickx's posts on the event.
