Bill Caskey suggests opening a phone voice note and answering ten detailed questions about your product's problem, objections, sales philosophy, discovery questions, and ideal customer, then transcribing the answers into AI. He says to answer honestly and in depth, as if briefing a marketing intern, so the tool learns your actual sales context.
Brian Neale says prompting mattered when they started about four months ago, but now he uses voice mode to talk out loud with AI and think through ideas conversationally. He describes it as talking to a colleague rather than sitting at a computer writing a prompt.
Brian Neale says they dump Salesforce closed-lost reason data into Claude and ask it to analyze the losses. Claude found financial services was 0 for 18 and a paid referral source was 0 for 22, signaling segments the team should stop pursuing. That turns raw CRM fields into disqualification guidance.
Brian Neale recommends asking Claude to name where your product looks like every other competitor and which attributes are me-too, so you can avoid those claims. He says this competitive analysis goes deeper than comparing features and benefits, which rarely reveals the real difference. It helps surface sameness before it becomes a sales problem.
Bill Caskey says to use your existing clients by name, brand, people, and detail when teaching AI your ideal customer. Brian Neale adds that you can also describe customers you dislike, so the tool filters out bad-fit prospects as well as finding more good ones.
Bill Caskey suggests ending the AI training session by asking what other questions should be added. That prompt surfaced a missing question—why you do the work—which he had not thought to include in the original ten. It turns the tool into a collaborator on the training design itself.
