I went to #canvasconf recently (organised by 383) and listened to Liz Crawford, CTO of Birchbox.
Liz had some deceptively simple insights into the concept of personalisation in ecommerce today.
Here are some of the things I took away.
Personalisation is more than a recommendation algorithm
Five years ago, those that were personalising their ecommerce service were ahead of the curve and ‘data scientist’ sounded a bit like a made-up job title.
Fast forward to today and everybody is doing it, creating deeper relationships to enhance customer lifetime value.
Personalisation is now more than a recommendation algorithm (‘customers who bought this, also bought..’).
Though this tactic has raised average order value (AOV) and conversion rate for Birchbox, personalisation has to be taken further – into the core of the experience.
Personalisation that disappears
Birchbox’s service is about trying, learning and buying, so shopping and discovery is designed to be easy, efficient and fun.
This concept of ‘personalisation that disappears’ was defined by Liz as attentiveness and serving the right message. This becomes personalisation that the customer is warmly aware of in the background but does not adversely affect the joy of discovery.
Messages such as ‘Liz, were you looking for this?’ are displayed when a customer (who has never bought) returns to the Birchbox site after previously viewing an item.
‘Popular in Hair’ may be displayed alongside a top product from this category, if a customer has recently browsed this category.
Liz pointed out that it’s easy to A/B test this stuff, incrementally increasing conversion without compromising customer satisfaction.
Thinking about every touch
Whether it’s web, offline (the Birchbox boxes), bricks-and-mortar stores, apps, push messaging or social media, they can all be personalised.
Liz advised to think in terms of customer events and not simply current marketing investment.
So, when emailing customers, do this in response to customer activity or change, rather than part of a pre-planned marketing calendar, however sophisticated your segmentation is.
This includes tactics such as a ‘shop our recommendations for you’ email, including products chosen based on recent browsing history.
Explicit personalisation and the power of customer involvement
One of the interesting questions asked of Birchbox’s data analysis is ‘what samples to send what customers?’
Birchbox does this in a number of ways. Firstly, there are restrictions on what some customer profiles can receive.
So, if a customer states they have curly hair, they’ll obviously never receive a product for straight locks.
The challenge is to produce a valid allocation of users to samples, that will also allow for optimal personalisation. It’s a challenge that’s characterised by having a high chance of mistake, as too many mismatched products will conceivably lead to cancellation.
There’s also explicit customer involvement in personalisation. By inserting an element of choice for subscribers, Birchbox found engagement with its service increased.
Subscribers initially received a message saying that ‘this month, you can choose one sample’. There was such a spike in usage that it was decided this would be a regular monthly feature.
Finding this balance between educated choice and discovery is a powerful mix.
Editorial and ecommerce
On the straight ecommerce side of the Birchbox business, editorial is used to help personalise the shopping experience.
So, aside from products, reviews and original content are surfaced. A/B tests show customers have been returning to this editorially-supplemented experience more frequently.
Don’t be limited by the data you have today
Liz described the omnichannel customer experience as one that should induce a feeling of ‘oh! you know me!?’.
This entails thinking about what the customer wants to hear, for example ‘your points are expiring, remember to spend them’, and delivering these messages in a timely manner.
Liz went on to warn about a marketing plan based on current marketing investment. Marketing, she stressed, is about triggered messaging (blocking and tackling).
The take-home lesson was ‘don’t be limited by the data you have today.’ Use customer involvement and personalisation that disappears to help customers discover and enjoy your product range.
For more brand stories involving data and customer experience, why not join me at the Festival of Marketing, November 11-12th in London.