UX & Predictive Lead Researcher

A Brief Overview

Safelite is the leader in auto glass service, continuously raising the bar for experiences that are easy, fast, and done. Digital is central to that ambition and to how we serve consumers, clients, and colleagues.

We are seeking a UX & Predictive Experience Lead Researcher to increase learning velocity across our digital product organization. You will build new ways to understand and anticipate customer behavior, combining first-party behavioral data, predictive modeling, AI, experimentation, and direct customer research to help teams learn earlier, make better decisions, and invest with greater confidence.

Reporting to the Director of Product Design, you will lead this capability as a senior individual contributor, working across research, design, product, analytics, and data science.

This is a role for someone who sees research as a learning system, not a sequence of studies. You will help us determine what we can learn from existing signals, what we can predict or simulate, what we should test, and when direct customer engagement will create new understanding.

• Increase learning velocity
• Build a learning system that helps teams reduce uncertainty earlier in the product development cycle.
• Use behavioral data, predictive methods, simulation, AI, experimentation, and research together, choosing the method based on the decision we need to make.
• Create ways to explore multiple hypotheses and experience directions before committing significant time and investment.
• Help teams distinguish between what we know, what the evidence suggests, and what still needs to be learned
• Build predictive experience capability
• Develop models and simulations that use first‑party behavioral data to anticipate customer response to potential experience changes.
• Create synthetic customer models that allow teams to explore scenarios, challenge assumptions, and identify promising directions earlier.
• Use predictive signals to inform prioritization and identify where additional discovery or experimentation will create the most value.
• Advance the methods as new data, AI capabilities, and approaches become available.
• Establish confidence in the signal
• Validate predictive methods against observed customer behavior and known outcomes.
• Define confidence levels, limitations, and appropriate use so teams understand how to act on the output.
• Investigate divergence between predicted and observed behavior as a source of learning, not simply model error.
• Continuously improve the models and methods as new evidence becomes available.
• Connect signals to human understanding
• Combine behavioral signals with qualitative research to understand both what customers are likely to do and why.
• Design research and experiments around the uncertainty that matters most to the decision.
• Use direct customer research where context, motivation, unmet needs, or emerging behavior cannot be inferred reliably from existing data.
• Feed new customer understanding back into the broader learning system.
• Turn learning into better decisions
• Translate complex evidence into clear implications for product and design teams.
• Create tools and ways of working that allow teams to use predictive insight without needing to become modeling experts.
• Bring a point of view on where we should explore, experiment, invest, or change direction based on the strength of the evidence.
• Raise the organization’s fluency in evidence‑based decision making, experimentation, predictive methods, and AI.
• Performs other duties as assigned
• Complies with all policies and standards
• Bachelor's Degree or equivalent work experience Required

• 7-9 years across UX research, behavioral science, data science, experimentation, or related disciplines, with significant experience shaping digital product decisions.
• 8+ years across UX research, behavioral science, data science, experimentation, or related disciplines, with significant experience shaping digital product decisions.
• Demonstrated experience building predictive, behavioral, simulation, or decision‑support methods using first‑party data.
• Strong applied statistics and experimentation foundation, including causal inference,…