How to audit your business processes for automation
Most businesses can reclaim 30-50 hrs/week, but only if they audit before building. This 5-step framework produces a prioritised list of automation candidates with explicit impact × ease scoring.
Tools you'll need
Steps
- 1
Catalogue every recurring task
For 30 days, list every task done >3 times. Note frequency (daily / weekly / monthly), duration (minutes per occurrence), and total hours/month (frequency × duration × 4).
- 2
Apply Pareto cut
Top 5 tasks usually consume 60-80% of recurring manual hours. Don't waste effort optimising the bottom 50%.
- 3
Triage each task into 4 buckets
(a) Automate now — rule-based, no judgement. (b) Augment with AI — judgement needed but pattern-recognisable. (c) Human-only — relationship, strategy, creativity. (d) Eliminate — doing this at all is the bug (often 10-15% of recurring work).
- 4
Score automate / augment tasks on impact × ease
Impact = hours/month × hourly cost × 12 (annual savings). Ease = build hours (1 if <4 hrs, 3 if 4-16 hrs, 9 if >16 hrs). Sort by impact/ease descending — top 5 are this month's roadmap.
- 5
Decide tool per task
Linear pipelines <100 runs/day → Zapier. Branching / iterating → Make. High-volume or state machines → n8n self-hosted. Long-running drip → dedicated SaaS (Customer.io).
- 6
Ship one workflow per week for 12 weeks
Resist parallel building — context-switching kills velocity. Each Monday: pick the highest-impact next task. Each Friday: review what shipped, log baseline metrics, tune next week's pick.