Oliver Bonas is an independent lifestyle retailer with around 80 stores across the UK. Camilla Tress is the connected commerce lead at Oliver Bonas; a role which focuses on creating a seamless online experience for customers across bricks and mortar, and online.
During our #PO event, Ian Jindal spoke to Tress about her experience, tested approaches and top tips around personalisation...
Could you tell us a bit about your experience with personalisation?
“As a business, the customer has always been really important to us, right from the start, 26 years ago. Early on in our online development, we introduced personalisation on our site so I’d say we were pretty early adopters back in 2015. We decided to build it ourselves, so we took the machine learning tool called prediction IO and we built it on site for personalised recommendations, personalised product listing pages and developed it from there.
“It’s taken a little while to develop it properly and we never got to the one-to-one personalisation; we used an aggregated customer personalisation and that’s been quite successful but it’s gotten to the end of its road.”
On machine learning
“You could do it yourself but we have a really good web agency who did it for us - but it is an open source tool. So yes, anyone could actually do it to build their own algorithm and create their own machine learning tool.”
What stopped you getting to one-to-one, hyper-personalisation?
“Another area where we thought personalisation was important was with on-site search where you search for something and the most relevant products appear for you. So we set up with an AI based search company and again, it was an aggregated personalisation for search results and we decided to test one-to-one personalisation versus the overall general personalisation.
“There was almost nothing in it, so I would say to people don’t rush out and think you have to do one-to-one personalisation, it’s going to be the silver bullet for your business. The reason why, for instance with search, if you search for a red dress on our site - we’ve only got probably one or two red dresses. So whether it’s one-to-one personalised or general, you’re going to get the same results.
“It’s trying not to get caught up in the hype around a lot of these things. It’s to really understand what is going to work for your business and where, is what I would say.”
How are you looking forward to build off of your experiences?
“We’re definitely moving forward. We’re definitely trying to improve our personalisation through our email communications and online. As I say, our prediction IO experience has come slightly to the end of the road so we’re looking towards our next steps.
One of the most important things for personalisation, especially for the online world, is to have a singular customer view - a 360 view of your customer. So we have invested in a Customer Data and Experience Platform (CDXP) which helps with marketing automation and really builds up the data around your customer so you can provide a good personalised experience for them.
Tress also highlights the problematic opposition faced by many companies, where data shows that customers want a personalised experience, yet those same customers say they are wary about sharing their data.
She continues: “It’s a constant challenge between providing the personalised experience they want, without being creepy or taking too much data or just not being compliant. It’s about really building trust with your consumers while proving that personalised experience.”
How does your path on personalisations change the way people work together to both deliver and exploit the capabilities?
“I think you have to don’t you, because personalisation covers every aspect of the business and it’s hard to escape it now. Whether you’re personalising what emails you receive; whether you’re personalising the experience as people land on your website; whether you’re personalising your retargeting...personalisation impacts every part of the journey.
“So you have to break down these silos and we find now we’re working much more closely across our marketing, our SEO and our customer services team. It’s not always easy and that’s where something like a Customer Data Platform (CDP) does come in handy because it helps bring the data together.”
Final advice for
“Don’t feel you have to jump right in because if you get personalisation wrong you have the danger of creating a bad experience. So just learn as you go and take it step by step is what I’d say. It is a complex area and it is hard to get right and better to start slowly with the data you have, rather than try and jump through hoops to be top dog.”