How I Audit a Whole Business Before Adding Any Automation
I walk you through the step‑by‑step audit I run on a 200‑person owner‑operated firm, exposing hidden waste before any SaaS or AI is installed.
When a client asks me to sprinkle bots on their workflow, I first ask them to stop. The temptation to chase the latest AI headline is strong, but the real cost lives in the invisible hand‑offs that never got documented. I learned that lesson the hard way on a 200‑person, owner‑operated firm where we tore out 21 SaaS tools and rebuilt a single operating system that the business now owns.
Map Every Interaction Before You Touch a Screen
The first day of the audit I sit with the heads of each of the 16 departments and ask them to walk me through a single transaction from start to finish. I watch the paper forms, the spreadsheet shortcuts, the email chains, and the occasional phone call. I record every hand‑off, every approval gate, and every manual data entry point.
I refuse to start with a diagram I found on a SaaS vendor’s site. Those templates assume a generic flow that rarely matches the quirks of a family‑run operation. By watching the real people, I discover that the finance team still uses a legacy ledger for inter‑branch transfers, while the field crew logs work in a handwritten logbook that later gets scanned.
The output of this step is a visual map that lives on a whiteboard or a simple Lucidchart file. I label each node with the person responsible, the tool used, and the data format. The map becomes the single source of truth for the next phase; if a stakeholder disputes a step, we point to the map, not to a vague memory.
I also note the latency at each hand‑off. A delay of a few hours may be acceptable, but a day‑long pause often signals a manual bottleneck that automation could relieve—once we know exactly what is waiting for what.
Finally, I flag any step that repeats across departments. The same inventory check appears in both the warehouse and the service crew, but each team uses a different spreadsheet. Consolidating that check later will save duplicated effort and reduce data divergence.