You can count on me

When I was at school, there were two types of mathematics: pure mathematics was about learning abstract number work by rote; and applied mathematics was about modelling the real world.

Too many companies are locked into ‘pure analytics’, where they gather numbers, and slice and dice them, but don’t put them to work. Companies that average activity across the whole site, or use measures like visit numbers and hits, are probably locked into this mindset.

For some time, we’ve offered visit scoring. This awards points to different activities on the site, so that companies can measure how engaged someone is. Each visit is scored according to the pages visited and activities that take place. A customer that just paws through the bargain bins is probably worth less to you than a customer who searches for an expensive TV and looks at a few results. Visit scoring enables companies to focus on potentially profitable and unprofitable visits and to adapt the user experience to increase customer participation. The next step is visitor scoring, which scores individuals over time, and considers how often they shop, how recently they shopped and their life time value to the business.

Using visitor scoring as a starting point, companies can introduce customers to others with similar tastes. This can drive social networking, so that site members are automatically introduced to others with similar tastes in products or hobbies based on pages viewed and items bought. This could be a good way to increase participation from those members who only pop in when they receive an invitation from a friend. People can be matched according to how they want to engage with the site too, so that you don’t pair up someone who’s online every day with someone who only logs on weekly.

Customers can also be matched on review sites. When I read bad book review posted by another customer on Amazon, I don’t know whether the reviewer has similar tastes to mine, so it’s hard to trust the review. Visitor scoring offers a simple way to benchmark how similar customers are to each other, so that they know how closely their tastes match and how much they can rely on each other’s reviews.

There will be privacy concerns, but Facebook’s plug-ins have shown that if there are benefits there, people are prepared to share their private data with anyone who can enable them to interact with their friends and the internet in new ways.