The Pre-Flight Check Before Real Research

Before you run a usability study with eight participants, you want to know the prototype actually works. That the checkout button is visible. That the form isn't dead-end. That the navigation makes basic sense.

You could click through it yourself, but you're too close to it. You know where everything is because you built it.

Synthetic user testing is AI agents clicking through your prototype as different personas: the frustrated novice, the impatient power user, the person who always misses the obvious button. They catch mechanical failures before you waste research time on them.

Here's what synthetic testing can't do: feel frustration when a credit card declines three times. Experience joy when something delightful happens. Understand cultural context or accessibility of lived experience.

What it can do: find the checkout button you accidentally made invisible. Flag the form that requires scrolling before you see Submit. Test seventeen layout variations in the time it takes to schedule one human participant.

18-24% of large design teams now use this. Not as replacement research. As the pre-flight check that makes actual human research more productive.

The value is mostly economic. A synthetic test costs around $50 and runs in 10 minutes. A moderated usability session can cost $200-300 per participant and takes days to schedule. If synthetic testing catches three broken flows before you bring in humans, you just saved a week and probably a few thousand dollars.

You still need humans to tell you if it's good. Synthetic testing just makes sure it's functional enough to be worth their time. It's the difference between testing whether your prototype works and testing whether it resonates.