Research first, tool second. A Clay table that scores AI startups against Maritime's ideal customer and narrows outreach to the few that fit.

Understand the customer → define the ICP → enrich the signals → score → decide
Most GTM engineers jump straight to the tool. I did the opposite.
I picked Maritime, a YC F26 startup building infrastructure for AI agents. I wanted to show how I approach a GTM problem, not just how I use Clay.
Find which AI startups are a real fit for Maritime, and rank them so outreach goes to the few that matter.
Of the first 20 companies, 2 came out WARM and 18 NO-GO.
That is the point. A good filter narrows outreach to the few companies with real fit.