Case study: Attic and Button

Attic and Button noticed that they were experiencing a high rate of dropoff after visitors added items to their carts. When they reviewed their funnel, they suspected the links in the funnel might be distracting visitors, so they ran an experiment removing the links. Their hypothesis read: “If we remove the extraneous links from our checkout page, we will see a reduction in dropoff in our checkout funnel, because visitors are currently dropping off in the middle of the checkout process and are likely confused or sidetracked by the links.”

When they ran their experiment, they found that their dropoff rate actually did go down! But what if it hadn’t? What if their hypothesis had been incorrect?

The improvement in completed purchases is about more than just lift. It confirms Attic and Button’s hypothesis and understanding of their visitors and site experience. If the hypothesis had failed, they would have still gained insight because their work was based on a specific and measurable hypothesis--and if they followed their micro-conversions, such as number of people creating a store account or navigation to different product pages and saw an impact there, they’d be even more certain that they were on the right track. They can use these insights to help decide what to optimize next, and how they’ll measure results.