Optimising the Customer Experience: Conversion Rate Optimisation Part 2
Optimising your customers’ experiences on your website—and your mobile website—is the first step toward creating frictionless customer journeys, and making sure your customers connect personally with your brand. Every time a user bounces from your site, you’ve wasted money.
In this series of articles, I’m walking through the entire process of optimising user experiences. In the first article of the series, I talked about creative consistency—making sure the content customers see when they arrive on your site follows directly from the advertising creative they clicked to get there.
But creative consistency is just the beginning. You’ve still got a lot of decisions to make in terms of how to organise the sections of the landing page, how to draw attention to the most salesworthy attributes of your products and offers, what graphics to use to present your offer, and how to manage flow throughout the page.
That’s where A/B testing and multivariate testing come in.
Building recipes
A/B testing puts multiple versions of a page—or a piece of content—head to head against each other. Multivariate testing often builds on A/B testing, enabling you to isolate the elements of the page that are actually driving user interaction.
The principle behind multivariate testing is simple: Generate a variety of versions of your page, in terms of content, layout, offer presentation, and even text and button styles, and see which of those versions works best for your chosen goals. In other words, multivariate testing can help you build different “recipes” for user experiences—where on your landing page the “killer questions” should be placed, how many steps the onsite conversion process should take, what kind of validation you should require, and even which colors and calls to action buttons to use.
Multivariate testing often generates a large number of variations that need to be tested—requiring a lot of site traffic in order to gather the necessary results. One way to speed up this process is to run what’s known as a partial-factorial test—that is, test a subset of all these variations of elements.
Comparing recipes
Once you’ve generated your “recipes,” it’s time for some testing. In this phase, you’ll actually show each of version of your page to real customers, and measure the ways they respond to the various layouts, offers, and button and text styles you’re testing.
You’ll be measuring how users respond not only to each “recipe,” but also to each individual element on the page, from buttons and text to images and ad creative. By combining a large number of individual use cases, you can determine, in real time, which recipe generates the most conversions.
Many CMOs still trust their marketing and web teams to handle this kind of optimisation manually, on the fly—but that’s a waste of money, because it means paying for hours of work that’ll never be used. It’s much more efficient to use a testing platform, take the guesswork out of landing page design, and let your audience, and the data, tell you which design will generate maximum conversions.
The effectiveness of testing
At our recent customer event, the Adobe Summit EMEA in London, a number of brands were gracious enough to share some of the wins they’ve seen from their testing programs.
Will Harmer, Senior Manager of Insights and Optimisation at EE, the largest mobile network operator in the United Kingdom, described how he and his team have been optimising their product detail page—the single most important page on their shop site. Customers are poised to make a major buying decision on this page, and an improvement by even a fraction of a percent can deliver big financial returns.
Harmer and his team decided to use multivariate testing to test a theory they’d been working on.
EE customers, Harmer knew from earlier analyses, have many choices between phones and plans—and customers often find those choices overwhelming. Harmon and his team believed that by showing visitors the most-purchased phones and plans on their website, this social proof would boost visitors’ confidence in their plan of choice, leading to conversions.
Harmer and his team tested this theory using multivariate testing, they immediately saw a 9 percent lift in orders per unique visitor.
The team is now exploring how social proof can guide customers through other complex decisions. For more information on this concept, take a look at this article from my colleague Blair Keen on the Econsultancy blog. For many other testing examples, you can access the full recording of the Summit EMEA session here.
Once you’ve optimised your landing pages and the site experience, the next step is to start thinking about the content you’ll need to support the conversion funnel on your site, and all other aspects of the onsite experience. This will help to ensure that as many users as possible will not only enter the conversion funnel but get all the way through it and become customers.
This is one aspect of what we call onsite content velocity—and it’s exactly what we’ll be examining in the third and final article of this series. See you there.