The second pillar of LevelOps' Operational Excellence framework: building clear, aligned, and measurable processes before automating anything.
AI thrives in structured environments. Without clear, documented, and measurable processes, AI doesn't fix your operations, it accelerates the chaos. That's why Process is the second pillar of the LevelOps Operational Excellence framework for AI success.
Good enough beats perfect. Don't wait for perfection, it never arrives. Start small, learn fast, and improve continuously. The most successful AI implementations evolve through iteration, not endless planning.
Measure performance. If you don't measure, you can't improve. Data-driven visibility helps you see where things are working, and where they're not. Measurement transforms assumptions into actionable insights.
Align with priorities. Focus on what truly matters. Processes should support your top business goals, not create extra work. Alignment ensures that every improvement drives real impact.
Do you have clear objectives and KPIs? How are you measuring and improving performance today? Because what gets measured improves, and strong processes drive results.
Next up in this series: Pillar 3, Tools. Choosing the right technologies for ROI and adoption.
Maxime Lavoie
Founder & CEO, LevelOps
Industrial engineer by training. Former Deloitte supply-chain consultant. Former VP Operations at Goodfood Market. Now writes the ERP, MRP and B2B order playbooks he wishes had existed in 2018.
Operations
LevelOps' Operational Excellence framework for AI success: Team, Process, and Tools. Before bringing in AI, get the fundamentals right.
Operations
The first pillar of LevelOps' Operational Excellence framework: why AI success starts with people. Clear roles, the right skills, accountability.
Operations
Key insights from Maxime Lavoie's Stratégies PME 2025 presentation. How top-performing SMEs use strategic clarity, discipline, and data.
Because a documented process still needs someone accountable for running it, which is what the team pillar settles first. Once ownership is clear, process is what turns it into something repeatable and measurable. In the other order you get a well written procedure nobody owns.
Take the process you run most often and write down how it actually works today, not how it should work. Good enough beats perfect here, because a rough version you improve next month is worth more than a complete one you never finish. The bar to clear is being clear enough to measure.
Only if you skip the second principle. Starting rough is safe when you measure what happens next, because measurement tells you which shortcut actually cost you something. Iteration without measurement is where this goes wrong.
For order handling, two numbers carry most of the weight: how long one order takes end to end, and how often one goes out wrong. Manual entry runs about six minutes per order at an error rate of 3%, and an erroneous order costs about $250 to unwind once you count the pick, the freight, the credit note and the call. That gives you a baseline any change can be checked against.
Aligning with priorities means the shortlist comes from your business goals, not from whichever process is easiest to write down. If improving a process would not move something you are already accountable for, it is not the place to start. Process that supports no goal is just extra work.
No. Everything here is definition and measurement, which you can do with what you already have. Tools are the third pillar and come after, precisely so that you are choosing a tool against a process you can describe and a number you can compare it to.
Book a short demo with a real purchase order and we run it live on your own data.
