GoogleA cloud of scattered orange, black and gray shapes on the left passes through three translucent panels and converges into one orange sphere on the right.

Building research systems for speed

Turning rapid research into a repeatable system for bringing user evidence into fast-moving product decisions.

Google Maps + Google AdsSurveys + Concept testing + Usability testing2-week studies

The problem

Product teams were moving fasterthan traditional research could support.

Across Google Maps and Google Ads, teams were making rapid product and experiment decisions, while research could take weeks to recruit, execute, and synthesize.

The challenge wasn’t simply to conduct studies faster: how could we make research more responsive without compromising the quality of the evidence?

My approach

From doing research quickly to designing for speed

On Google Maps, I led rapid research across global markets, running 15+ studies in six months. On Google Ads, I helped set up a rapid program for advertiser research that ran 8 studies in two months.

The model streamlined scoping, recruitment, and synthesis, and fixed recurring problems like participant quality and unclear scope. Focused studies could go from question to decision in about two weeks.

Can research keep up? → Where can research change the decision?

On Google Ads, one rapid study gave the team the evidence to cancel a planned launch and move the resources elsewhere. On Google Maps, research shaped launches later featured in Google’s Q4 2020 earnings call and covered by TechCrunch and PCMag.

Impact

  • 20+ studiesDelivered rapid research across Google Maps and Google Ads, supporting fast-moving product decisions across consumer and advertiser experiences.
  • 5x daily usersResearch informed Google Maps launches that contributed to a 5x increase in daily active users.
  • Changed decisionsResearch influenced design and experimentation across multiple projects, including evidence that led the team to cancel a planned launch and redirect resources.

Two weeks was fast enough to matter.Teams could wait for evidence instead of guessing.