You're Probably About to Automate the Wrong Thing
The CEO reads about automation, gets excited, and picks the most complex cross-department process as the first target. Six months and $200K later, the project stalls. Everyone concludes automation doesn't work.
It works fine. The selection was wrong.
The companies that succeed start boring. They pick a small, repetitive, measurable process and nail it. Then they expand. Every time.
Step 1: Watch What Actually Happens
Process documentation lies. It's always 18 months out of date. People build workarounds, skip steps, and add informal checks that never got written down.
Sit with the person who does the work. Watch them. Document what they actually do, not what the SOP says. Those workarounds? They reveal where the process is genuinely broken.
One client had a "15-minute" data entry process. When we watched, it was 45 minutes -- they were manually cross-referencing three spreadsheets nobody knew about.
Step 2: Measure Before You Touch Anything
Get hard numbers:
- Volume: How many times per day/week/month?
- Duration: End-to-end cycle time?
- Labor hours: Person-hours per cycle?
- Error rate: How often do mistakes happen? What do they cost?
- Wait time: Active work vs. sitting in someone's inbox?
Wait time is the killer insight. A "3-day process" often contains 2 hours of actual work and 2.5 days of waiting for approvals. Automation won't fix a queue problem -- you need to fix the workflow first.
Calculate annual cost of the manual process. That's your ROI baseline.
Process mining tools can accelerate this step dramatically. They analyze event logs from your existing systems to map actual workflows, reveal bottlenecks, and quantify waste -- turning weeks of observation into a data-driven KPI dashboard in days.
Step 3: Score and Pick
Rate each candidate on four axes:
Complexity (low = good): Simple, linear processes automate faster and fail less.
Volume (high = good): A process running 500 times daily delivers way more value than one running twice a week.
Standardization (high = good): Same steps, same formats every time? Easy to automate. Every instance is a special case? Fragile automation ahead.
Impact (high = good): Time saved, errors prevented, compliance improved, people freed up.
Your first target: low complexity, high everything else.
Mistakes That Kill Projects
Automating broken processes. If it's inefficient manually, automation just makes it inefficiently fast. Fix the workflow, then automate the fixed version.
Skipping the pilot. One team, one location, one subset. Watch it closely for two weeks. Fix issues. Then expand.
Ignoring the humans. Automation changes jobs. If people feel threatened instead of supported, they'll sabotage it -- consciously or not. Involve them early.
Fantasy timelines. Simple processes: 4-8 weeks. Complex cross-system flows: 3-6 months. Plan accordingly.
Ignoring the integration layer. Most automation connects systems through webhooks, API integrations, and data pipelines. No-code and low-code platforms like n8n, Make, and Zapier handle straightforward ETL flows well. But when you need custom data transformations or error handling across multiple APIs, plan for engineering time -- not just drag-and-drop configuration.
When to Leave It Manual
Some things shouldn't be automated:
- Changes constantly -- you'll spend more time updating the bot than it saves
- Needs empathy or negotiation -- keep humans here
- Runs once a month for an hour -- ROI will never justify the investment
- One mistake = catastrophic damage -- keep human oversight
Knowing when not to automate is half the skill.
The Playbook
Pick one boring, high-volume process. Map it honestly. Measure it. Automate it. Prove the value. Use that win to fund the next one. Repeat. That's how automation programs actually succeed.