# A four-step order for AI that survives the builder. Source: https://thewiremethod.com/method Site: The WIRE Method Tagline: AI that works inside the business. Founder: Richard Batt Contact: hello@thewiremethod.com This Markdown companion is generated from the prerendered page for AI crawlers. The framework # A four-step order for AI that survives the builder. Most AI projects fail because of operations, not technology. The WIRE Method is the order that fixes that. Below is the high-level shape of the framework. The implementation detail, the worksheets, and the named patterns that go with each step are taught inside the certification programme. ### W, Audit where time is lost 01 A structured operations audit that finds the repeating manual work in the business. This is not a survey or a workshop. It is a line-by-line account of where the hours go. - The Monday Morning Test, the diagnostic that finds the candidate processes - The Time Tax, a calculation method that gives every process a real cost - The Whiteboard Problem, the most common failure pattern at this stage ### I, Identify the first automation 02 A scoring method that picks the right first build out of the candidates. The fastest win, not the biggest opportunity. Documented before a tool is touched. - The Fastest Win selection rule, with the four scoring criteria - The Tool Trap, the trap most teams fall into at this step - A readiness checklist with the three structural gaps every business has ### R, Run the smallest working version 03 A build cadence that ships in days, not quarters. The first version is small enough to observe and is reviewed daily for the first two weeks, with every output checked. - The Minimum Viable Automation, with the deployment cadence - The Four Breaks, the failure modes every running automation eventually hits - The output review protocol that catches the silent drift before it costs ### E, Embed the build into the business 04 A handover protocol that makes the build outlive the builder. Every embed has a named owner, a monthly review cadence, and the documented behaviour written down. Governance is kept light enough for a small business to run. - The Bus Test, the question that decides whether the build is finished - The Governance Layer, scaled for small and mid-sized teams - The Freed Hours Flywheel, the compounding pattern that follows good embeds ## Principles ### Automation, not abdication Automate the work. Keep a named human owner. The point is to free the team for higher-value work, not hand the business over to a tool nobody understands. ### The electrician, not the professor The output is a working automation, not a slide deck. Every step has a testable deliverable. The framework has no place for strategy documents that collect dust. ### Smallest working version first The first build is not the right build. It is the one that proves the outcome cheaply. Real running automations beat polished plans every time. ### Operations is the work The technology is the easy part. The work is finding the right process, documenting it, picking the right owner, and reviewing the outputs. The method is an operations discipline that happens to use AI. ## What the method is, and what it is not ### What it is - A repeatable order that takes AI from idea to a working part of operations - An operations discipline with named patterns and tested failure modes - A credential and a licensing model for practitioners and organisations ### What it is not - A technology guide or tool review (the method is tool-agnostic by design) - A strategy framework for the boardroom (the output is a working build, not a deck) - An enterprise change-management programme (the cadence is small business and mid-market) The problem it fixes ## Three operational gaps cause most AI failures. In every implementation that has come back broken, the cause is one of three gaps. Undocumented processes that quietly assume things nobody wrote down. Unmeasured outputs that drift for months before anyone notices. Unowned automations where everyone is responsible so nobody is. The WIRE Method exists to close all three before they cost. ### Undocumented processes The Wrong Autopsy, the Whiteboard Problem, the readiness gap. Most teams cannot describe their own work in enough detail to automate it. ### Unmeasured outputs The Silent Drift, the Four Breaks. Automations look like they are working until the day they are not, and the cost is already done. ### Unowned builds The Bus Test, the Governance Layer. Without a named human reviewer the build is on borrowed time.