Moving on from our last post on basic profiling, let's look at increasing the sophistication.
A powerful way of segmenting your database is to look at 'click behaviour' as well as browsing or purchase history.In comparison to profile based segmentation (which relies on customers to actually tell you about themselves), behavioural based segmentation is a valuable way of helping you understand how customers are behaving at a deeper level. Using this information makes it possible to tailor & personalise messages to each individual - you can often learn more about a customer from what they do, rather than from what they tell you. Remember that if you are using cookies, gifs, or any other form of online behavioural advertising, you need the consent of the recipient respecting the local compliance rules for data privacy & similar.
A simple method for segmenting your customers could be on email activity. For example, if you use the date of the last email interaction, you easily build a number of segments which enable you to target and analyse customers based on their engagement level.
- Segment A - engaged within the last month
- Segment B - engaged between 1 and 5 months
- Segment C - engaged between 6 and 9 months
- Segment D - engaged more than 9 months
A similar model can be applied to purchase & browsing data. If used together, you have a potentially very detailed view of your customers - identifying the most active, those who recently purchased & engaged within your program (etc.) and compare them to those who are falling away and may different attention.
Two key strategies that can be implemented using behavioural based activity are:
- Engage with best customers on a regular basis - give them the VIP treatment they deserve & will appreciate.
- Re-engage the customers that have lapsed (.e. segment D) - treat them in a completely different way. They were once interested in you & your products but are not engaging with you - you need to develop a strategy to win them back. A point to note - if people have disengaged due to emails at too high a frequency, then this needs to be assessed and it is important to make sure that they are segmented into a lower frequency emailing group than the main mail-out list.
To make the most of this type of segmentation, it can also be combined with the different types of data that you have collected on your customers. For example, if information on 'product lines purchased' is available, then a targeted campaign that is aimed at recent openers, clickers (segment A) and purchasers of these products could be developed. This type of sophisticated targeting and segmentation will have a direct and positive impact on the performance of your campaigns. The results you should be able to monitor will be significantly higher than a similar campaign sent to an un-segmented database of email addresses.
Behavioural segmentation can also be used to understand customers' likes and interests where you might not otherwise know that data. Email and web analytics systems can group customers into product or interest related categories based on the way in which they interact with emails, web pages or products they purchased previously. Using this information can be helpful in building segments that allow you to target different products to these people. Be careful though, in some cases it can be very dangerous to assume that someone will only be interested in one type of product just because they bought it once before.
Our next post will look at lifetyle analysis
Source: Digital Marketing Association - Whitepaper "A Guide to Data Analysis and Segmentation"