Redesigning AI.Google to differentiate, excite, and build trust.

WEBSITE RE-DESIGN • DISCOVERY • USABILITY

Role

Senior UX Researcher

Timeline

Feb - May 2024

Team

Instrument (Agency) in partnership with Google

PROBLEM

Over time, Google’s AI website had become a text-heavy repository of information.

Despite Google’s history and deep expertise in AI, the site failed to excite key users about products and instill trust and confidence in their approach.

It was thorough, highlighted dense principles & objectives, and dove deep into AI responsibility rather than allowing consumers to understand relevant applications.

OPPORTUNITY

Re-position the site to serve tech influencers and everyday consumers.

Through conversations with stakeholders, the team determined that we needed to design a site that would differentiate Google from its competitors. With a strategic re-design, we wanted to:

Tell Google’s story as a legacy AI leader

Excite consumers about AI products

Build trust around AI safety & responsibility

RESEARCH APPROACH

I led two rounds of research: discovery to support strategic direction and usability to validate and pressure-test designs.

DISCOVERY RESEARCH ARTIFACTS

Interviews covered existing AI perceptions and reactions to 3 competitor websites alongside Google’s website.

I designed a moderator’s guide to address the following objectives:

  1. Understand perceptions of the AI landscape, including current and historical leaders in the industry

  2. Evaluate existing attitudes around AI responsibility and safety

  3. Evaluate how participants respond to existing Google and competitor websites

Discovery research helped lay a strategic foundation for design work, resulting in 5 design principles:

  1. Broaden mental models of AI. Work to expand audience understanding of AI beyond LLM applications.

  2. Lead with concrete details. Embrace the specificity that our audience looks for.

  3. Humanize scientific authority. Fold in Google’s long-standing and rigorous scientific legacy in highly relatable ways.

  4. Demonstrate responsibility through transparency. Position Google as the responsible, safe leader through examples.

  5. Design for human/AI partnership. Show rather than tell the relationship between AI & humanity.

USABILITY TESTING ARTIFACTS

Following design iterations, I led a round of usability testing focused on comprehension and resonance.

After a few weeks of design iterations based on the design principles from discovery research, I conducted 10 additional usability tests. These tests focused on reviewing clickable prototypes of the redesigned site. Pages included:

  • A redesigned homepage

  • A models overview

  • A Gemini ecosystem page

  • Navigation

After usability tests, work-shopping, and synthesis, I outlined findings mapped to specific design recommendations unique to each of the pages and features tested.

At a high level, we heard our audiences key into a deep need for human, inclusive, and safe messaging throughout Google’s AI website.

OUTCOME & IMPACT

The two research sprints informed the project at strategic and tactical UX levels.

Taken together, the two sprints informed the design of ai.google, particularly by understanding key audiences needs and mental models.

Recommendations around simplification and humanizing overly technical parts of the website were key in making the site more useful to both tech influencers and consumers. We shifted the overall tone from corporate to human, and elevated key responsibility & safety initiatives in more engaging ways.

Beyond the improvements, this work built client confidence in the team. They appreciated our persistence in reaching a specific, niche audience in the usability test, and were impressed by how well our initial proxy group represented the attitudes and perceptions held by true target group.

TAKEAWAYS

Calibrate research rigor to the learning moment.

A key challenge was recruiting influencers with large public audiences in the discovery round. Pivoting to speak with “proxy influencers” - people with smaller but meaningful audiences, either within companies or on LinkedIn - allowed us to gather early insights that quickly informed design.

While the team iterated on prototypes, I was able to recruit higher-impact influencers to validate those initial findings and ensure our insights held true for the broader audience.

This reinforced the importance of flexibility and importance of balancing methodological rigot and speed. It helped me to concretely see the value of validating early insights with the most representative participants when possible.