Where your SOPs actually live
Ask most firms where their document intake process is documented, and the honest answer is: it isn't, not really. There's a shared drive folder with an outdated PDF, and then there's what the person currently doing the job actually does — the exceptions they've learned to handle, the shortcuts they've picked up, the judgment calls that never made it into any document. The real SOP lives in a person's head.
The photocopy-of-a-photocopy problem
When that person trains their replacement, some of that knowledge transfers — but not all of it, and not perfectly. The replacement learns a slightly simplified, slightly different version of the process. When that person eventually trains the next one, it happens again. Over several cycles, the process has drifted meaningfully from what it originally was, and nobody decided that on purpose.
A process that only exists in someone's memory isn't documented. It's a single point of failure with a job title.
Why this is invisible until it isn't
This cost doesn't show up until the person leaves — and then it shows up all at once, as errors, missed steps, and "wait, how did we used to handle this?" questions during a period when the firm is already dealing with the disruption of a departure. It's a cost that's easy to ignore precisely because it's dormant most of the time.
What automation forces you to do anyway
Building a system to handle intake, routing, or reporting automatically requires defining the process precisely enough for a machine to execute it — every exception, every routing rule, every edge case has to actually be specified, not just known by feel. That's not a side effect of automating the work; it's the mechanism by which the SOP finally gets written down accurately, once, in a form that doesn't degrade with staff turnover.
Related: Signs your firm has outgrown manual back-office processes →