Notes on the invisible job.
Where the money goesThe change order you approved out loud
A client says "sure, go ahead" in the driveway. The crew does the work. Six weeks later you're arguing about $1,840 you already spent. It has one cause and one fix.
July 2026 The software questionWhat's your monthly software bill, and how much of it did you set up?
Do the math on that second number before you buy anything else. Most of what I build lives inside tools my clients were already paying for.
July 2026 How the build worksWhat actually happens in a 7-day build
Day by day: the conversation, the build, the four failure tests, and the moment it goes live. No forms. Your crew learns nothing new.
July 2026 Asking for moneyThe $5,250 invoice I refused to discount
The client pushed back hard. I held the number, kept the relationship, and got paid in full. Every collection system I build comes from that day.
July 2026 Where the hours goHow AI actually saves time at work: 5 workflow examples that matter
AI saves time when it is attached to real workflows. Here are five practical ways small businesses can use AI to reduce admin, follow-ups, reporting, and tool chaos.
From the archive Where to startWhat to automate first in a service business
The best first automation is rarely the flashiest one. Service businesses should start with repeated, rules-based work that saves time without removing human judgment.
From the archive Deciding what to buildThe 10-hour test: how to decide whether an AI automation is worth building
Before building an AI automation, use the 10-hour test: will this workflow recover meaningful time, reduce risk, or improve client experience enough to justify the work?
From the archive Before you buyThe small business AI audit: where to look before you buy another tool
Before buying another AI tool, audit the workflows, handoffs, follow-ups, reports, and underused systems already inside your business.
From the archive The first workflowClient onboarding automation: the first workflow many service businesses should fix
Client onboarding is one of the best first workflows to automate because it is repeated, visible, and full of small admin tasks that slow delivery.
From the archive Before you hireBefore you hire admin help, audit these five workflows
Before hiring admin help, audit the workflows creating the admin load. Some work needs a person. Some work needs a better system.
From the archive Admin workAI for admin work: how to get hours back without hiring
AI can reduce admin work when it helps draft, sort, summarize, route, and remind inside real workflows. Here is where to start before hiring more help.
From the archive Plain EnglishThe plain-English guide to AI workflow automation
AI workflow automation means using AI and automation to reduce repeated work inside real business processes. Here is the plain-English version.
From the archive Shadow AIWhy your team is already using AI, even if you have no AI strategy
Your team may already be using AI without a formal strategy. That creates productivity opportunities, privacy risks, and workflow confusion if you do not set rules.
From the archive Workflow, not promptsAI productivity is a workflow problem
Better prompts help for small tasks. Real AI productivity comes from workflows with clear outcomes and human review.
From the archive The riskAI can make good workers faster. It can also make wrong answers faster.
AI can improve speed and quality on the right tasks, but it can also make bad answers faster. The difference is workflow design, review, and human judgment.
From the archive Skills that matterAI skills for nontechnical workers: what actually matters
Nontechnical workers do not need to become engineers to use AI well. The real skills are judgment, delegation, verification, workflow thinking, and clear communication.
From the archive