Better Decisions, Faster: Predicting Human & Agent Behaviors Before Release

Speed is up. Confidence in product decisions isn't. Most teams commit time and resources before they know what will actually work, running the same loop of ship, wait, learn, repeat. That loop is breaking as build cycles collapse and live experimentation gets harder to run reliably.

Why? A/B testing requires live traffic. User research takes weeks. Analytics tells you what happened, not why. By the time you know, you've already shipped the next thing. The traditional model assumes you have time to be wrong. Modern product cycles don't.

This session reframes the core problem: a product decision is a commitment of time and resources made under uncertainty. The question is not how to move faster. It is how to be right more often before you commit.

Tom Charman will show what changes when you move behavioral insight before release. Instead of waiting for post-launch signal, you can simulate how real users move through your product, step by step, and see exactly where they hesitate, misinterpret, or drop off. We'll walk through a concrete onboarding scenario and compare the traditional ship-and-measure approach against pre-launch behavioral simulation side by side.

You'll leave with a lightweight pre-launch simulation framework you can use immediately: define a user, define a task, walk through their behavior step by step. It won't replace your research or analytics stack, but it will give you directional signal before you ship, when changing course is still cheap.

The future of product development is not just building faster. It's being right more often.

Key Takeaways:

• Why ship → learn is breaking in modern product development
• How to identify friction and drop-off points before a single user sees the product
• A new mental model for experimentation that doesn't rely on live traffic

Special Thanks to our Generous Sponsor

Blok: Predict how users and agents behave inside your product before you ship. Replace guesswork with behavioral simulation.

Attendance will be first-come, first-served. We will close doors once we reach capacity; arrive early to secure a spot.