🎙️UX+DEV SUMMIT 2018 - Pedro Marques
summary
Uses machine learning models to detect patterns like repeated date searches (example: a user searching Seattle for next week then again for another day) and surface a one-click alternate-date prompt that jumps to cheaper options. Implements profiling and A/B tested buckets to avoid annoying mismatches, measures booking conversions and engagement drop‑offs, and iterates on hypotheses rather than blind personalization. Requires large event volumes (Booking.com scale) for stable profiles but the concrete takeaway is find pre-failure patterns and proactively fix them.
AI-written summary · about
chapters (7)
- 0:00Welcome and introduction
- 3:17Pedro’s background and career
- 9:52Predictive design explained
- 13:26Balancing helpfulness and annoyance
- 16:33Key takeaways and actionable insights
- 19:44Putting users at the center of design
- 20:46Closing and contact details
related talks
31mWhy data science and UX research should be best friends
30mMobiconf 2018: James Butler, Booking.com
12mMichelle Brownstein 🎙 Front End Design Conference 2019
28mThe Impact of AI on UX Design
28mCrafting Hyper-Relevant User Experiences
26mShapeshifters: The #1 UX Mistake You Don't Know You're Making with Sophia Prater
56mEssential Analytics and Coding for the ‘Complete’ UX Professional
02mFebruary 2018 Highlights