Your company has an SOP document. It’s fifty-seven pages long, lives in a shared drive nobody checks, and was last updated by someone who left two years ago. When a new hire asks how to handle a specific situation, someone says “check the SOP” — and then five minutes later says “actually, just let me show you, the SOP is outdated.”
Most standard operating procedures fail because they’re written as aspirational documents rather than captured from actual practice. Someone sits down and describes how things should work. The people who do the work every day never look at the document because it doesn’t reflect how things actually work.
The most accurate procedure documentation comes from recording actual work as it happens. When you or a team member performs a process, capture it in real time with Voice Mode:
Build SOPs from Captured Workflows
“Walking through the client onboarding process. Step one: receive the signed contract via email. Forward it to Mem. Step two: create a new collection for the client. Step three: set up their account in the system — the login page is at this URL, use the admin credentials from the shared note. Step four: send the welcome email, which I pull from the template collection. Important: always cc the account manager. Step five: schedule the kickoff call within five business days.”
This isn’t a formal document. It’s a real person describing what they actually do. And it’s infinitely more useful than a formal SOP because it includes the institutional knowledge that formal documents omit: which URL to use, who to cc, the five-business-day standard that’s tradition rather than policy.
Traditional SOPs are written once and rot. AI notes create living procedures that update naturally. When a process changes — a new step is added, a tool is replaced, an exception is discovered — capture the change:
“Update to the onboarding process: we now need to verify the client’s billing address before setting up the account. Adding this between steps two and three. The billing team requested this after the invoicing issues last quarter.”
Ask Mem Chat to compile the current procedure at any time:
Living Documents, Not Static Files
“What’s the latest version of our client onboarding process based on all my notes?”
The AI assembles the original procedure plus every update and modification into a current, accurate description. No version control headaches. No outdated documents misleading new hires.
The best SOPs are documented by practitioners, not managers. The person who handles returns knows the actual steps, the common exceptions, and the workarounds. The person who manages the monthly close knows which spreadsheets actually matter and which are legacy artifacts.
Make it easy for practitioners to document: a two-minute voice capture of their process is enough. They don’t need to write a formal document. They talk through what they do, Mem transcribes and cleans it up, and the procedure is captured.
Collect these practitioner captures into process-specific collections. When a new hire needs to learn a process, they can ask Chat:
Let the People Who Do the Work Write the SOPs
“How does our team handle the monthly financial close?”
The answer comes from the people who actually do it, in their own words, with the practical details that formal documentation misses.
The hardest part of any SOP is edge cases. The standard process covers eighty percent of situations. The other twenty percent — the exceptions, the unusual requests, the “what do I do when…” scenarios — is where new hires struggle and where institutional knowledge matters most.
Capture exceptions as they arise:
“Exception to the return process: when a client requests a return after the thirty-day window, we can still process it if the account manager approves. This happened three times this quarter. I escalated to the director each time and they approved two of three.”
Exception Handling
Over time, your exception notes become the FAQ that no SOP document ever includes. Ask Chat:
“What are the most common exceptions to our return process, and how have we handled them?”
New hires get answers to the questions that Google and the formal SOP can’t answer — because the answers come from your team’s actual experience.
The act of documenting processes reveals inefficiencies. When someone talks through the twelve steps of a process and realizes that steps four through six could be a single step, the documentation itself becomes a process improvement exercise.
After capturing several processes, ask Chat:
Process Improvement from Documentation
“Based on our documented workflows, are there steps that seem redundant or could be streamlined?”
“Which processes have the most exception notes, suggesting they might need redesign?”
The processes with the most exceptions are the ones that need attention. The steps that multiple people describe differently are the ones that lack clarity. The notes reveal what needs fixing, not just what exists.
For operations teams building systematic documentation, our guide on documenting institutional knowledge covers the broader knowledge capture strategy. And for the team collaboration dimension, AI notes for project management shows how documented processes integrate with daily operations.
The best SOP in the world is useless if nobody can find it. Traditional SOPs require you to know the document exists and where it lives. AI notes make procedures discoverable through natural language:
Making SOPs Discoverable
Instead of searching for “SOP-2024-ONBOARDING-V3.pdf,” someone asks Chat:
“How do I onboard a new client?”
The answer comes from your captured procedures, compiled and current. No document naming conventions to learn. No shared drive to navigate. Just a question and an answer.
Pick one process your team does regularly and have the person who does it voice-record the steps
When an exception occurs, capture how it was handled
Get Started
Ask a new hire to try following the captured procedure and note where it’s unclear
Update the notes based on their feedback and ask Chat to compile the current version
The best SOPs aren’t written by management. They’re captured from the people who do the work.
Cluster