Dr. Connor Robertson Uses Claude AI for Business Tools

Robertson’s list is longer than most people expect: role-specific onboarding checklists, weekly scorecard templates built from a defined structure, email sequences for specific client types, SOP documents written in a company’s internal voice, training guides built from a brief provided in plain language, and interview frameworks with scoring rubrics. Each of these, he says, would previously have required weeks of manual document creation; Claude can produce a solid first version of any of them in under an hour.

In Robertson’s framing, quality of output depends entirely on quality of input. He recommends treating every interaction like commissioning work from a capable contractor who is new to the business, stating what is being built, who it is for, what format it should take, what constraints apply, and what a successful output looks like. He describes the most common mistake as asking for something with insufficient context, then concluding AI cannot deliver what is needed, when the real issue was the briefing.

Robertson’s example: feed Claude the key metrics, the measurement cadence, the team members who own each metric, and the desired format, then ask it to produce a structured weekly scorecard template with conditional formatting described in plain language. From there, copying the output into a chosen tool and mapping in the data fields produces a custom reporting tool, a process he says a developer would take two days to complete and that AI can complete in an afternoon.

Claude’s first draft is rarely the final version, in Robertson’s assessment, and expecting otherwise is a miscalibration. He frames the real skill as iteration: flagging what needs to change, tone too formal, a missing section, language that does not sound like the

company, with each revision closing the gap between generated and usable output. He suggests budgeting three to five cycles for anything intended for deployment at scale.

Once a Claude-built tool is solid, Robertson recommends layering automation beneath it: connecting it to a trigger in Zapier or Make, letting the automation pull context data and feed it to Claude as part of the prompt, with Claude generating the output and automation delivering it to the right place. That combination, in his description, is a no code internal tool that runs itself.

Robertson observes that most business owners who try this approach stop after the first output, treat it as either a success or a failure, and move on. In his view, that skips the part of the process where most of the value actually gets created. A first draft that is 70 percent right is not a signal to abandon the tool, it is the expected starting point of a process built for iteration rather than one-shot generation. Businesses that get the most out of Claude for internal tooling, in his framing, are the ones that treat every output as a draft to be refined rather than a final product to be judged, which changes both the quality of what gets built and how quickly a team is willing to try the next one.

Dr. Connor Robertson is an entrepreneur, author, and strategic advisor based in Pittsburgh. He is the founder of Elixir Consulting Group, host of The Prospecting Show, publisher of The Pittsburgh Wire, and founder of The Grant Finder. He is also a six-time published author, with titles including Built to Run, available at