Most teams know something is broken. Fewer can name the one workflow that burns the most hours every week. Automating the wrong process locks in waste. A tight audit fixes that.
This is the same first pass we use at isomore.tech (Compass Ventures LLP) before any fixed-price build.
Why 20 minutes is enough
You are not writing a process encyclopedia. You are ranking pain by:
- Time — hours per week across the team
- Frequency — how often it runs
- Fragility — how often it fails or needs a hero
If a flow is rare, cheap, and stable, leave it alone.
Step 1 — Scan the last seven days (5 min)
Walk chat, email, and calendar. Mark every moment someone:
- Re-typed data between tools
- Asked “what’s the status?”
- Waited on a spreadsheet owner
- Copy-pasted the same message to customers or vendors
Those are automation candidates.
Step 2 — Score and pick one (5 min)
For each candidate, rough-score 1–5 on time, frequency, and fragility. Multiply. Highest total wins.
Rule: one system first. A single shipped workflow beats three half-built bots.
Step 3 — Map the happy path on one page (7 min)
Write only:
- Trigger — what starts it
- Systems — CRM, inbox, sheet, billing tool
- Handoffs — who touches it
- Done — what “finished” means
Anything that depends on someone remembering a step is a defect waiting to ship.
Step 4 — Choose the tool after the map (3 min)
| Job shape | Default tool |
|---|---|
| Stable rules, structured data | Script / n8n |
| Messy language or judgment | Claude skill + human gate |
| Queue of exceptions | Internal tool + automations |
If you cannot describe success in one sentence, you are not ready to build.
What good looks like
After the audit you should have:
- One named workflow
- A one-page map
- A first-system scope you can price
That is the input we take into a discovery call—and into a fixed-price build that usually ships in 2–6 weeks.
Next step
Want a second pair of eyes on your map? Book a discovery call with isomore.tech.