At Amazon Devices, we had a winning product launch strategy. Conversion was strong, and there was little reason to question it. Until I started reading customer reviews. People were buying products they didn't realize they'd chosen, then asking about features their products didn't have. I came to call this Accidental Conversion: when a product successfully drives an action before the user has fully made the decision behind it. For decades, we've designed products to remove friction. But making actions easier doesn't always mean helping people make better decisions. I'll share two Amazon stories: one where we discovered this problem after launch, and another where we caught it before launch. Now AI is taking this further, increasingly making decisions and acting on our behalf. How do we ensure we're optimizing for what people actually want, not just what our systems can get them to do?
Group Design Site Manager, Intuit