User Experience Research Method to Improve Optimization Testing

1. Define user intent and reasons for possible opposition

It is important to understand why people visit, whether they can successfully visit, and why they leave.

If you mistakenly assume the reason people visit the site, your cause theory and web design assumptions will not be able to reflect the user's perception of the environment. It is dangerous to make assumptions about why users leave without doing a survey.

For example, suppose you assume that the visitor did not do the intended operation (buying) because the price was too high and you then reduced the price and the profit margin was hit. If people don’t really buy it because the price is not the price, but they don’t understand what the services you provide solve the demand, you’re wrong.

Let's take a look at what Netflix does:

Netflix made a very interesting A/B test to understand the factors that affect the conversion rate. Netflix's homepage shows a large number of film and television works, but what different users see is different. As part of its A/B testing, this is a personalized recommendation from the company for users. However, not only that, Netflix also conducted batch tests on users who saw the same video product recommendation. They let visual designers redesign different versions of film and television posters and use this variable for A/B testing. The result is that even with the same work, different posters will still result in different conversion rates. This confirms that the reasons for the user leaving or clicking are diversified.

So don't limit yourself to obvious reasons, but dig deeper into what users may be concerned about. Sometimes users do not necessarily say what they did, so often observations provide more real user propensities than surveys.

2. Discover the defects of the interface

If you ignore major usability testing issues, such as difficult-to-understand interaction processes, or micro-interactions that are easily misunderstood, then you may not get the conversion rate from constant A/B testing because your design There is no source of response to the problem.

For example: If you have provided blanks with several required information that users are inconvenient or unwilling to provide, then merely changing the color of submit button to do A/B testing will not effectively increase the conversion rate. It is a waste of energy. Understanding the real reason for low conversion rates is the key to performing smart, successful testing.

How to discover the flaws of the interface: Usability testing (remote hosting or exemption from hosting, or face-to-face) can be performed quickly and with the help of five or so users can expose nearly 85% of the site's major deficiencies.