We use our expertise to answer your question.
A conversion rate optimization (CRO) hypothesis is the foundation for any A/B test or experiment you run on your website, and it’s just like any other scientific hypothesis (“if this, then that” statement) that serves as a basis for investigation. For a website experiment, it contains all the information needed to construct, execute, and make an evaluative comparison for said experiment. There are three criteria that a CRO hypothesis should include to be successful: It needs to be understandable, actionable, and measurable.
While the acronym for CRO is conversion rate optimization, when creating a CRO hypothesis, we apply a framework of comprehension, response, and outcome (our take on the CRO acronym) to each of those three criteria.
C is for comprehension. Here, you’re identifying and understanding the problem, and making an observation by watching shopper behavior and/or gathering data from other sources.
Ex: For the last two months, you’ve observed reduced revenue from a lower average order value (AOV) by comparing your order management system’s last six months of purchase data (month-over-month).
R is your response, or the action you will take to solve the problem you’ve identified. The test variation explains what you will accomplish, the location (i.e., on which webpage(s) the experiment gets applied to), and the audience (who the experiment will impact).
Ex: You want to promote paired products through an inline upsell opportunity on the cart page for returning users who have an item already in their cart. The variation is to show this inline upsell promotion on the cart page, and the UI of the promotion displays paired products relevant to items the user already has in their cart.
O is for outcome. What is the result that you want to achieve? The outcome measures information about the actions taken, or the behavior change you’re trying to affect with the test.
Ex: This test should increase AOV by increasing upsell add-to-cart events, backed up by measurement of multi-item purchase count, order conversion rate, and revenue-per-visitor.
What’s great about our CRO framework of Comprehension, Response, and Outcome is that it maps back to the hypothesis being understandable, actionable, and measurable. You can apply this any time you’re looking to create your own CRO hypothesis for a website experiment.
