Software products are starting to all look the same. Seemingly EVERYONE is putting the prompt box front and center (see: Canva, Apollo, Clay, …).

My two cents: It’ll be harder & harder to differentiate on the user experience. We're going to need to differentiate on the agent experience instead.

What I'm seeing from the best products:

1. Product education starts before even hitting the website.

With a prompt box UX, the user needs to come to the product already armed with an idea about how to use it. Folks are educating users through hackathons, communities, slick launch videos & asking influencers to show off what they've built.

Template libraries still work, too. (Half of Lovable’s homepage is now devoted to UGC templates.)

2. The product helps you improve your prompt.

Most of us (myself included) aren't prompt engineers. And a garbage prompt can often mean a garbage result -- and not much patience to keep going.The best products help us clarify our ideas, turning "meh" prompts into "omg" prompts. I particularly love the one-click "enhance prompt" feature in products like Bolt(dot)new and Replit.

3. The agentic UX isn't one-and-done.

The UX shouldn't just complete the task and move on. The best agentic experiences recommend next steps & help you do them.

They can also become the best sales rep for the product: proactively encouraging you to upgrade & then suggesting the plan that's best for you. After all, helping was always the best way of selling.

4. The agent isn't a black box.

The best tools include a preview mode & give users the ability to edit what AI produces. It's not just a "do it for me" experience, it's a "do it with me" experience that leads to an outcome people feel great about.

---

The innovation isn't putting a prompt box in the app. It's creating an agentic UX that (finally) feels personal & maybe a bit magical.