How Anthropic uses Claude in Product Management

Getting data as a product manager means pinging a data science team and waiting. Lisa Crofoot, Product Manager, shares how Anthropic's PMs use Claude to query product data and build evals in minutes—no SQL required. See how product managers, including the Head of Product for Claude Code at Anthropic, use Claude: www.anthropic.com/news/product-management-on-the-ai-exponential Stay tuned for more stories in the "How Anthropic uses Claude" series.

Lisa Crofootguest

Mar 25, 20262mWatch on YouTube ↗

WHAT IT’S REALLY ABOUT

Anthropic PM accelerates iteration using Claude for data and evals

A product manager describes using Claude to test and iterate on product ideas independently before involving other teams.
Claude Code is connected to BigQuery via an MCP, enabling natural-language access to product data without needing to write SQL.
In a demo, Claude quickly generates visualizations (including rolling averages and overall averages) and iterates to segment results by plan type.
The team also uses Claude to generate evaluation (“eval”) test cases, expanding from a few examples to dozens to assess AI product behavior.
The overall impact is not just automation but capability expansion, freeing PM time for strategy, customer conversations, and decision-making.

IDEAS WORTH REMEMBERING

5 ideas

Claude enables PMs to validate ideas earlier and more independently.

By testing product concepts with Claude before looping in others, the PM reduces coordination overhead and can iterate faster on hypotheses and direction.

Connecting Claude Code to BigQuery turns data access into a conversation.

With an MCP linking product tables to Claude Code, the PM can request analyses in plain language and focus on interpreting results rather than writing queries.

AI-generated charts can exceed “baseline” manual analysis quality and speed.

In the demo, Claude produced a polished graph and added enhancements (7-day rolling average, overall average) without being explicitly asked, then quickly re-plotted by plan type.

Claude reduces the time cost of exploratory segmentation and follow-up questions.

What might take hours manually—changing cuts of the data, generating new plots, checking patterns—can become rapid iterations with permissioned edits.

Claude helps scale eval creation by expanding the test-case space.

Given a product scenario and a couple examples, Claude can generate many more diverse cases (e.g., 50) to systematically evaluate AI system behavior.

WORDS WORTH SAVING

5 quotes

What excites me about being a product manager now, um, is that I can move and iterate much faster. I can use Claude to test my product ideas before I even get anyone else in the loop. I can operate much more independently, which is extremely empowering.

— Lisa Crofoot

You don't need to know how to write SQL, you just need to be a human interpreter of that data and that result.

— Lisa Crofoot

This is, like, pretty impressive. It's done a whole bunch of stuff I didn't ask for, like add a seven-day rolling average, add the overall average.

— Lisa Crofoot

If I had to do this on my own, it would take me possibly hours, um, if not more.

— Lisa Crofoot

I would say it's, like, more than automation for me. Like, it's things that I wouldn't have otherwise necessarily even been capable of doing independently. So it's actually, like, extending what I can accomplish on my own.

— Lisa Crofoot

Faster product iteration with AIClaude Code for PM workflowsBigQuery MCP integrationNatural-language data analysis vs. SQLAutomated charting and segmentationAI eval generation and test-case expansionShifting PM time from coordination to strategy

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.