SOPs Are Code

If you have SOPs, you're closer to deploying agentic workflows than you think.

Most CS leaders who hear "agentic AI" picture something that requires a development team, months of build time, and a six-figure implementation budget. The technical framing — agents, workflows, APIs, orchestration — makes it sound like infrastructure work.

It's not. And the reason it's not is something Lewis Thompson said in our webinar that I want to expand on here.

SOPs are code.

What This Actually Means

A Standard Operating Procedure is a step-by-step description of how a task should be performed. Input comes in. Steps happen. Output goes out.

That's also exactly what an agentic workflow is.

The difference is execution. An SOP is executed by a human who reads it, makes judgment calls, performs the steps, and produces the output. An agentic workflow is executed by an AI that reads it, makes judgment calls, performs the steps, and produces the output.

Same structure. Different executor.

Which means if you have an SOP for something your CS team does repeatedly — call prep, churn risk assessment, handoff documentation, renewal readiness review, expansion signal identification — you already have the blueprint for an agent. You just need to give it to Claude Code and let it build.

The Translation Process

Here's how it works in practice.

Take an existing SOP. It doesn't need to be perfect. It doesn't need to be comprehensive. It just needs to describe what happens: what triggers the task, what information is needed, what steps are performed, what the output looks like.

Feed that SOP to an agent — through Claude Code or a similar environment. Tell it what tools you use, what data sources are available, and what you want the output to look like.

The agent will ask clarifying questions. It will identify gaps in the SOP. It will figure out how to connect to the data sources it needs. And it will build a working prototype of the workflow.

You don't write code. You describe what you want, in plain English, based on a process your team already runs.

That's it. That's the starting point.

Why This Matters for CS Teams Specifically

CS operations are full of SOPs that are partially followed, inconsistently applied, and resource-constrained.

The call prep SOP that takes a good CSM two hours and gets skipped when the calendar is full. The churn risk review process that should happen monthly but actually happens when someone notices something is wrong. The handoff documentation standard that the sales team ignores and CS has to reconstruct from memory.

These aren't bad SOPs. They're good SOPs that nobody has the capacity to execute consistently.

An agent doesn't have capacity constraints. It doesn't skip the call prep because the day got busy. It doesn't do a lighter version of the churn review because there are thirty accounts and only so many hours. It runs the SOP every time, completely, for every account.

That's the value of the translation. Not that you're doing something new. That you're doing something you already know works, consistently, at scale, without adding headcount.

Where to Start

The best SOP to start with is the one your team skips most often because it's valuable but time-consuming.

That's usually call prep. Or churn risk assessment. Or handoff documentation. Or expansion signal review.

Pick the one where inconsistent execution is costing you the most — in missed signals, in unprepared CSMs, in churn that could have been caught earlier. Build that one first.

You're not starting from scratch. You're starting from the SOP you already have. Feed it to the agent. Let it build. Iterate from there.

Your SOPs are already code. You just haven't deployed them yet.