One the main buzzwords in the retail world at the moment is “personalisation”; however, as it seems to me with a lot of these terms, I can’t help but feel that there’s a lot of confusion about what this actually means in real life.

Part of the problem is that a lot of companies say they do personalisation, but what they actually offer is something quite limited.

For example, whenever an email provider says that they offer personalisation, but really only let you address your emails to “Joe” or “Mary” based on some name field in a database - I’m talking about {$FIRST_NAME} tags - in my opinion they’re being a little misleading.

In contrast, right at the other end of this very wide spectrum of personalisation sit some far more advanced technologies. But since the term “personalisation” has sadly been soiled by so many lesser pretenders, in this post we’re going to talk about the development of a whole new paradigm in retail user experience: behaviouralisation.

Track behaviours, not attributes

Traditionally, adapting to users meant storing snippets of attributes of customers, such as their first name or email address, or perhaps they ticked a couple of boxes on some retailer website to say that they were interested in “racing cars” or “flower arranging”.

Then, whenever that retailer was running a special promotion on racing cars or flower arranging, Joe would get an email that said “Dear Joe, we’d like to tell you about our new special deals on flower arranging”.

However, there are clearly several key pitfalls to this approach:

  • The data isn’t terribly specific
  • The data is static
  • The data is self-reported

For instance, if in a year from now Joe no longer likes flower arranging, and actually he was only ever really interested in arranging roses anyway, and to be honest he was kind of lying about the racing cars bit, then our “personalised” data about Joe is probably not very useful.

So, a lot of retailers have taken another step and figured that maybe they should adapt to customer behaviours rather than attributes. How about sales?

Track behaviours, not just sales

The next step would be to track whenever customers buy anything, and as a result of this to divide all of your customers into groups that can be marketed to separately.

So for example, because Joe bought a clock last year, he is now a “Clocks” customer, so every now and again he receives an email that says “Dear Joe, we’d like to tell you about our new special deals on clocks”.

And that’s almost good! Actually this represents a big step in the right direction, because now we’re basing our strategies on what customers actually do, as opposed to what they say they do (different things).

But there’s still a long way to go - because the data still isn’t particularly rich:

Customers browse a lot more than they buy So what about other customer behaviours - surely we should adapt to them as well? For instance, let’s say that Joe wants to buy his mother a vase, so one day he looks at 10 different vases for sale on our retailer’s website. One of the vases in particular has caught Joe’s eye, so he clicks on the product page and spends an entire five minutes before leaving. He zooms in on a photo of the product for a closer look, and heck, he even copy-and-pastes the product name to search for reviews on Google.

The following day, Joe receives an email: “Dear Joe, we’d like to tell you about our new special deals on clocks”…


Rather, the way forward is “behaviouralisation”. Think of it as personalisation on steroids. Now we’re going to track, not just our sales or abandoned baskets, but all sorts of different data points about our customers, so that we can adapt our campaigns to our customers on a very granular level.

This way you can build up an extremely detailed picture of customer actions throughout your site - and Joe finally gets his email about vases.

At Persomi, we make behaviouralisation easy.

Here’s a list with some examples of the customer interaction data that we can help you to store:

Any page view
Product views
Product category views
Add/remove to basket
Add/remove to wish list
Zoom product picture
Select text
Use discount code
Ask for live help
Ask for catalog
Contact us
Refer to friends
Search something
Banner impressions
Banner clicks
Email views
Email clicks

This is an extraordinarily rich set of data, that opens up a whole new realm of opportunity for an exceedingly highly customised retail web experience.

Of course, storing all of this information is only the first of the pieces in the jigsaw puzzle, and in future blog posts we will talk about some of the different ways in which you can use all of this wonderful data.

But for now, the main takeaway from this post is: don’t send emails about clocks when your customer wants a vase. Behaviouralise today…