Ryan Thomas is Head of Direct Channels at Comet, a role which covers the retailer's non-store operations, which includes the website, call centre, its B2B insurance business, and mobile.
I've been asking Ryan about Comet's approach to mobile commerce, as well as testing and optimisation.
How important is mobile for Comet?
Mobile is a big part of our business going forward. Three years ago we were looking at the trends in this area, and there was a lot of discussion around mobile.
We thought it was the way forward then, and decided to really push for mobile, and two years ago we launched the first Comet mobile site, at a time when few other retailers were doing anything around mobile.
This was for click and collect only, but provided a good user experience, and was more of a multichannel play.
We recently made this into a fully transactional site, and incorporated PayPal payments. We didn’t want to stop there though, and wanted to produce an app as well.
Mobile now accounts for more than 10% of Comet’s total traffic, so this was clearly an area worth investing in, and we expect this to double.
Those who pass the opportunity of mobile commerce by will miss the boat. I think there's a big advantage to be gained by getting in early.
How did you approach the design of the app?
We wanted the app to do something different to the mobile site, not just a transactional app, but to actually enhance the multichannel and in-store experience for users.
The aim was to bring the richness of data we have on the website to the app, and by extension, into stores.
The website has lots of ratings and reviews, and PluggedIn data, that can help customers when they are in our stores.
We also have the barcode scanner, which is something we could only do via an app, and which allows customers to scan items in store and see reviews and other content.
Is the barcode scanner intended primarily for this, or for use in competitor’s stores?
It’s predominantly to improve the in store Comet experience for app users, but if a customer want to scan a product elsewhere, a friend’s house for instance, then it will work for that.
A lot of electrical retailers have products behind glass, and don’t necessarily have boxes with barcodes on display. It would be difficult to scan a large item like a fridge for example.
What else can you do with mobile for in store customers?
We’re quite interested in the potential of mobile vouchers, and using mobile and apps to increase loyalty and reward regular customers.
We have lots of interesting ideas in this area, so it’s a question of choosing the best and prioritising them. An app can be what you want it to be. We could use push notifications to promote in store offers for example, though we are careful not to be too intrusive.
Do you see apps as a tool primarily for customer retention, and mobile sites for acquisition?
To a certain extent. With the mobile site, the primary purpose is to make it as easy as possible for customers to make a purchase, but the app aims to allow customers to do things they can’t do on the site.
It’s not just about the purchase, but aims to help with the research process.
Are you looking at designing apps or a version of the site for tablets?
We will be looking into tablets over the next six to twelve months, and testing how the desktop version of the site works on iPad.
We’re looking at how people use iPads and Android tablets, and rather than using a version of our mobile site for tablets, we’d be looking at serving a version of our website that has been optimised for tablets.
You introduced a user generated Q&A section to the website, how has this worked?
This has worked really well, and has grown into a useful resource for customers. We have found that customers are willing to help each other, and our own staff can answer questions and correct any inaccuracies.
We like the community to get involved and answer questions, and this is something we’d like to add to the app in future.
What is Comet's approach to testing and optimisation?
The user experience on the site is vital, and we use multivariate testing (MVT) regularly to optimise the site for customers.
We’re constantly trying out different versions of our product pages, serving different versions to test customer reactions.
For example, we’ve been looking at things like the colour of call to action buttons, trying out red, green, orange and blue.
From what I can see on the product pages, it seems that green won.
Yes, and by some distance, so that’s what we’re using now.
We’re continually playing around with different layouts, one will win, and then we’ll use that version and move on to testing something else.
The key is to continually look at ways of improving, rather than settling on one version. Little changes can make a big difference.
Do you have a team member, or a number of staff that are responsible for testing and optimisation?
We do have one person that is responsible for this, though that isn’t his only role. We also have a team of people who are reviewing the results of tests and potential improvements.
It’s a big part of what our design team does now.
In days gone by, designing product pages and other features of the site would have been decided by a few people in the company, now we can ‘ask’ millions of site users, we had 1.2m visitors last week, so we can find out very quickly what they prefer.
It takes opinions out of the equation, and turns optimisation into a science, as the facts trump anyone’s personal view on what works.
The results are sometimes counter-intuitive, but if it improves conversions and user experience, than you can’t argue with the facts.
It seems a detailed product page is working for you...
We add as much information as possible to help customers, though we have to be careful to avoid crowding the page.
We've integrated Plugged In, and reviews from Tech Radar, Stuff and others. This is so that people don’t need to leave the site for product research.