Positions we hold,
written down.

Not a content calendar. These are the arguments behind how we work — each one stated in full up front, so you can disagree with it before you hire us.

• 01

  Stop Automating Broken Processes

  Published note

  Automation is amplification. If a process contains a workaround, a duplicate entry, or an approval that exists because someone got burned in 2019, automating it makes that permanent. Ask what the step is for before you make it faster.

• 02

  AI Is Leverage, Not Strategy

  Published note

  A model is a capability with no opinion about your business. Strategy decides which work matters; AI decides how much of that work a person still has to do by hand. Confusing the two produces pilots that never reach production.

• 03

  How to Find the Right First AI Project

  Published note

  Look for work that runs often, follows articulable rules, consumes real hours, and has an obvious definition of correct. The right first project is boring, and that is exactly why it succeeds.

• 04

  Your Website Is Part of an Operating System

  Published note

  A website is not a brochure with a contact form. It is where demand enters your operation. If what it captures does not land somewhere a person or system acts on, the site is a dead end with good typography.

• 05

  What Should Become an Agent — and What Shouldn't

  In progress

  Agents suit work with branching sequences, tolerant failure modes, and a reviewable output. They are wrong for irreversible actions, ambiguous authority, and anything where being confidently wrong is expensive.

• 06

  Tribal Knowledge Is a Business-System Problem

  In progress

  When the answer lives in one person's head, the constraint is not memory — it is that no system ever asked for the knowledge. Capture is a workflow design problem before it is a software problem.

• 07

  Why Most Dashboards Don't Improve Decisions

  In progress

  A dashboard built from available data answers the questions the data happens to support. A dashboard built from decisions starts with what someone must choose each week and works backward to the number.

• 08

  Map the Process Before Selecting the Software

  In progress

  Software selection made before a process map is a bet that the vendor's assumptions match yours. Map first and the shortlist usually shrinks to two — sometimes to none, which is also an answer.

• 09

  When to Build vs Buy

  In progress

  Buy where your process should look like everyone else's. Build where the process is the reason customers choose you. Most bad outcomes come from building the commodity and buying the differentiator.

• 10

  Where AI Actually Creates Leverage in a Small or Mid-Sized Business

  In progress

  Rarely in the product. Usually in the middle of the operation: intake, quoting, documentation, follow-up, knowledge retrieval, and the administrative work quietly consuming the owner's week.