Deep Research Max prompt build — live Q&A capture (complete through Rec 81)

Context

Sam is building a Deep Research Max prompt for Sam Aguiar Injury Lawyers. The deliverable is a packaged folder produced in one Max run. Format is whatever Max can natively deliver in a single output, with the per-URL audit as a CSV (or multiple CSVs) and supporting documents in their native formats, each tagged with its intended file path. A folder map at the top of the output ties everything together.

Claude walked Sam through 81 recommendations one at a time. Sam answered yes, no, or qualified yes. This page captures the full record so the prompt can be drafted from it without ambiguity.

Definition of Winning (illustrative, not exhaustive per Sam at Rec 81)

  1. Top 2 ranking on 80 percent of highest-traffic and highest-converting keywords for personal injury, car accident, and truck accident in Louisville.
  2. Top 5 ranking on the same keyword set in Lexington.
  3. Organic traffic exceeding Hughes & Coleman, Kaufman & Stigger, Karl Truman, Hendy Johnson Vaughn, Gladstein, and Bahe Cook (named in addition to the existing list; Max identifies the full set).
  4. SEO-attributed conversions doubling to 40 per month.
  5. AI-search-attributed cases reaching 15 per month.
  6. Top 3 recommended firm across ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews for personal injury, car accident, and truck accident in Louisville, Lexington, and Kentucky.

These are examples, not the executive scope. Max should treat them as directional, not as the only success criteria.

Other directives captured this session