Profound

Intro

In 2025, I joined Profound as the sole product designer. The New York–based startup helps brands understand how AI agents like ChatGPT, Perplexity, Google AI Overviews, and Copilot talk about them. Processing over 100 million AI queries each month across 18 countries and 6 languages, Profound supports companies including Indeed, MongoDB, Ramp, and Rho.

With more than 60 percent of consumers starting product research with AI assistants, knowing how these conversations happen has become critical for marketing teams.

The Problem

AI answer engines are changing the way people discover products. Instead of typing keywords, people are asking AI assistants conversational questions. Traditional SEO tools have no way of showing what is actually being asked.

This shift created three big challenges for marketers:

No visibility into trending prompts and emerging topics

No clear sentiment analysis or share of voice for AI results

No way to see which competitors dominate AI conversations

Early users told us they did not want raw data dumps. They wanted clarity, trends, and actionable insights.

Role and Team

I worked closely with engineers and data scientists, leading design from start to finish. My responsibilities included research, UX flows, UI design, and building a scalable design system. I also worked with the founders on product strategy, deciding which insights mattered most and how to present them in a way marketers could act on.

Approach

User Research

I spoke with growth teams, content strategists, and AI researchers using Profound's early prototype. We found that AI search patterns often had little overlap with Google, that AI conversations covered the entire marketing funnel, and that marketers wanted the ability to group related prompts, track sentiment, and see which websites influenced AI answers.

Competitive Research

I analysed leading SEO and keyword-tracking tools including Semrush, Ahrefs, Ubersuggest, and Google Trends to see what they offered and where they fell short. While they excel at keyword rankings, backlink analysis, and search volume trends, none could surface AI prompt-level insights or reveal how AI agents respond to real questions. This confirmed that Profound had a clear opportunity to lead in a new category.

Design and Development

Hierarchical Explorer allowed users to move from broad topics to specific prompts, making large datasets easy to navigate

Bulk Analysis let marketers compare up to 200,000 prompts with advanced filters for date, region, language, and AI platform

Answer Engine Insights provided dashboards with visibility scores, sentiment breakdowns, and share of voice trends

Agent Analytics gave human-readable visualisations of AI crawler activity across domains

Design System used modular Figma components to maintain consistency and speed up engineering

Challenges

Translating billions of real-time data points into a clear, usable interface

Designing around privacy restrictions by summarising and grouping prompts instead of showing raw text

Turning technical crawler logs and sentiment models into practical, actionable insights for marketers

Solution

The finished product brought multiple capabilities together in a single platform:

Answer Engine Insights to track mentions, sentiment, and competitor share of voice

Agent Analytics to monitor AI crawlers and link their activity to traffic

Conversation Explorer to reveal trending questions and hidden AI search patterns

A scalable design system that could support both enterprise clients and a $499 per month self-serve plan

Impact

Ramp increased AI visibility in the accounts-payable sector from 3.2 percent to 22.2 percent in one month

1840 and Company went from zero to 11 percent AI visibility in remote staffing, entering the top five in their category

Profound now processes more than 100 million queries per month across 18 countries and 6 languages

Early adopters reported a 25 to 40 percent lift in AI share of voice within 60 days

Reflection

Profound showed me how rapidly shifting industry trends can create entirely new product categories. I learned to design within data privacy and technical constraints while making complex AI insights accessible. The project reinforced the value of aligning user experience with evolving behaviours in how people search and consume information.