Introducing Segment metrics: The end of one-size-fits-all personalisation
The idea of personalising content for people has been around for a while. When I was a kid it was about getting your name on pens, notebooks, anything that you could write on really. There was a lot of practicality to personalisation for children — after all, it’s harder to lose a headband when your name is plastered across it. Did anyone own or want to own one of these headbands in the 90s? That was living the dream.
More recently the idea of personalisation has been centred around trying to get to know a person specifically, rather than just getting their name. This is great in theory, however the practice of this is sometimes a little tricky.
Imagine you’re personalising a homepage banner for men and women. It sounds like a good idea initially, but if we think about how massively behaviour and opinions can vary from one man or woman to the next, suddenly you realise that you’re not getting that close to your audience at all. In this instance if a woman comes to a site, without additional information about her, she’s going to gets the same one-size-fits-all message as her mother.
Let’s be honest, one-size-fits-all only really works with socks, so it’s surprising that this method often ends up being used to create bespoke experiences for visitors.
About 12 months ago at Qubit we introduced the concept of Adaptive Targeting by building our segmentation tool, Segments. We wanted to get away from the one-size-fits-all approach and help businesses to deliver genuinely tailored personalisations for each type of person coming to their site.
It’s sounds obvious but we believe that personalisation should be focused around the customer.
To deliver personalisations that genuinely make a difference to a customer’s journey we need to know who they are and why they’re there.
The latest Segment update, which comes in two parts, is designed to make it easier for users to get this contextual information.
1. Segment builder upgrade
Since the initial release of the segment builder, we’ve been hard at work making it more detailed, more powerful, and easier to use.
The builder has a brand new look, with brand new conditions to match. As well as having access to contextual and behavioural data, Segments is now integrated with our Datasets tool to allow uploaded offline datasets to be used as a basis for personalisation. And if you’ve created a Visitor Pulse experience for your visitors you can actually create segments based on the answers people have provided.
The crucial thing here is the ability to combine multiple types of data.
In the UX world we often talk about the dangers of letting assumptions control the direction of a product. It’s the same for on site personalisation. By creating segments of visitors based on multiple complex attributes, we can avoid making assumptions about the way customers shop.
2. All new metrics!
The segment builder helps with physically splitting your traffic into key groups but to get to know your customers properly, and to know how best to speak to them, you need to be able to see what they do after they’ve joined a segment. This is where you can find out key information like how many sessions it takes for someone to become a VIP, what device new visitors use most, and what products are most popular with lapsed customers.
This has led to us developing the all new segment metrics view.
For each segment of visitors created you can see:
- Top level numbers like Conversion Rate, RPV and AOV.
- The breakdown of how many visitors have joined, left and remained in the segment each day.
- The most popular products and categories for people in the segment
- What experiences are targeting the segment and the impact they are having on segment performance
With this information you can start to build up a picture of who is actually in the segment and create personalised experiences based on interesting behaviour.
In this particular segment you can see that the most popular categories with new visitors are Shoes and Knitwear. From this you could create a recommendations experience on the homepage for first time visitors to showcase your newest items in these categories. You could even show when products are about to go out of stock to make sure they don’t miss out.
Additional context can inform your whole personalisation strategy. Basic personalisation is positive but without a wider understanding of your audience, a truly personal experience can never be achieved.
As always, this is just one step along the road to better personalisation. But we hope it takes people in the right direction — thinking of the customer first, and leaving the one-size-fits-all socks at home.
Sarah Benson. Product Designer at Qubit — wearing ill-fitting socks 😩
Give me a bell on twitter @sarahbenson18 or give the 💚 a quick click if you enjoyed the read.
Thanks to the whole Adaptive Targeting team for all the hard work and brain power. And special thanks to Suganya Sivaskantharajah and Giovanni Luperti for help with editing.
