Interaction effects occur when the effect of one variable on a dependent variable depends on the level of another variable. In the context of A/B testing, interaction effects can occur when the impact of one experiment is influenced by the presence of another experiment.

For example, consider two experiments: Experiment A tests a new color for a "Buy Now" button, and Experiment B tests a new placement for the same button. If the experiments are run separately, they might both show positive results. However, if they are run simultaneously, the combination of the new color and new placement might confuse users and lead to negative results. This is an example of an interaction effect.

In the context of overlapping A/B tests, interaction effects can increase variance and skew results. However, strong interaction effects are rare. It is more common to find smaller non-linear results.

While these can skew the final numbers, it’s rare to find two experiments collide to produce outright misleading results. Effects are generally additive which leads to clean “ship or no ship” decisions. Even when numbers are skewed, they are in the same ballpark and result in the same decisions; overlapping experiments can be trusted to generate directionally accurate results.

Interaction effects that lead to broken experiences are fairly predictable. Broken experiences or strong interaction effects typically occur at the surface level, eg. a single web page or a feature within an app. At most companies, these are usually controlled by a single team and teams are generally aware of what experiments are running and planned. Interaction effects are easily anticipated and can be managed at the team level.

Here's an example of an interaction effect that could break the user experience:

• Experiment 1: Moves the “Save File” Button to a menu

• Experiment 2: Simplifies the UI by hiding the menu

If these two experiments are run simultaneously, they could combine to confuse users, as the "Save File" button would effectively be hidden. This is an example of an interaction effect that breaks the user experience, and should be avoided.