March 2013, March 18-19, Hotel Nikko, San Francisco
User experience really does matter: optimising the mobile purchase funnel
The development process for mobile apps has several phases and the travel industry has gained a fair understanding of how processes work. But while the app developed may look slick, and might deliver on various counts (such as speed, efficiency and simplicity) during the testing phase, it may still not translate to hard sales.
It all boils down to levels of engagement. If an app user is comfortable with features, swiftly moves across various pages and finds the layout easy to navigate, chances are they will return. The likelihood of them completing a transaction will also be higher. So the mobile purchase funnel should be seen as part of the overall user experience.
Current buying patterns validate this. In fact research shows that those users who check out their app multiple times before completing their first purchase are proving more valuable. According to a study by Localytics, users who end up completing a transaction during their first session made an average of 2.8 purchases.
EyeforTravel’s Ritesh Gupta talks to Daniel Ruby, director of online marketing, Localytics, about working on the mobile purchase funnel.
EFT: Firstly, what are the main considerations for choosing particular device and mobile operating system in a mobile strategy?
DR: The target market and budget are two key criteria. If possible, launch on iOS and Android (maybe even Windows Phone - there are fewer devices, but also lower competition) simultaneously. Otherwise, choose the devices or operating systems that align most closely with the app’s target market. One should focus on regular audience measurement, market penetration and user engagement research while also detailing key metrics of different mobile operating systems. If starting an m-commerce app from scratch, such reports help app makers make more educated decisions.
EFT: What should your focus on in the final stages of the purchase funnel?
DR: One should focus on testing, segmentation on the basis of usage and look beyond the download-figure for deeper analysis.
• Do: Test, test, test your purchase funnel. Find the pain points where users are dropping off. Too often, there’s a spot in the funnel that may seem great to a UX designer, but users find it confusing or off-putting in practice. Always tracking funnel performance will ensure that if your app has one of those spots, you can quickly diagnose it and act on it.
• Do: Segment your users based on purchase history. Put your new installs into one bucket, your ‘have never purchased’ but returning users into a second, your small purchasers into a third and your heavy buyers into a fourth. Track each segment individually, determine the comparative value of an individual user in each, and make sure the experience each one is receiving is correct.
• Do not: Focus only on download figures. Downloads are only one measure of success, and only the starting point (ie. the very top of the funnel). There is still some way to go between the download and more sales. Make sure you’re marketing to the ‘right’ users, paying attention to active users, and driving engagement through the rest of the purchase funnel.
EFT: How can one build actionable m-commerce funnels?
DR: Don’t try to build a different funnel for every product or category.
Depending on your app design, your basic funnel should look like App Launched -> Home Screen -> Category View -> Product View -> Add to Cart -> Checkout Start -> Checkout Completed.
EFT: Can you expand further on this?
DR: One needs to consider both screens and Events. They form different elements of an app’s usage.
An event is an occurrence or action within the app - when a user shares something, or when they play a video. A screen is sort of like a webpage – it’s the part of the app on which an event occurs. So a screen may be the featured video of the day, and several events could happen on that screen eg. Play video, pause video, share video, etc.
For custom attributes to improve purchase funnels, these will give you perspective into what users like and what they don’t.
Delving deeply into the funnel means going beyond step 1 - step 2 - step 3 to goal.
Users rarely if ever take the direct approach of going through each step of your funnel in the order you expect them – they’ll go off and look at reviews, or they’ll add something to the cart, remove it, and then add something similar before they purchase it.
Knowing all of these elements - or at least the ones most commonly performed by your customers - can give you the best clues as to what will trigger the final push to the goal.
A good funnel can be made up of both screens and events. Events, like the ‘product view’, have custom attributes attached to them. These attributes can range from product category and product ID to ‘times viewed’ and ‘time on Item. Properly instrumented attributes allow for deeper delves into the funnel while keeping the funnel itself simple and actionable.
In the end, though, users will probably take a limited number of paths from launch to purchase, so make sure your funnels are few but inclusive.