You're automating the wrong thing
Most companies start automation where it looks good — a chatbot, a report generator, a polished dashboard. The real leverage is usually somewhere less visible.
You're automating the wrong thing
Every automation conversation I have starts the same way. Someone tells me what they want to automate: a customer-facing chatbot, an AI report generator, a social media calendar that runs itself.
None of these are bad ideas. But they're rarely the right first step.
Companies pick automation projects based on how impressive they sound, not on where the actual pain is. Those two things rarely point to the same place.
Where time actually goes
Ask most teams where their time goes and you'll get a polished answer. The one they've already thought about.
Have them track their time for a week and you get a different picture.
The real time sinks aren't the big visible processes. They're the micro-frictions: the Excel file that takes two hours every Monday because three systems don't talk to each other. The fifteen follow-up emails after every proposal because the first version always has missing information. The internal ping-pong when a customer inquiry lands in the wrong inbox.
None of these show up in a press release about digital transformation. But that's where the hours actually go.
The glamour trap
I call it the glamour trap. Companies pick automation projects for visibility, not value.
A customer-facing chatbot is visible. Everyone can see it. Stakeholders can point to it.
But if your real bottleneck is a weekly data reconciliation or an internal routing problem, you've spent three months building something that doesn't touch your biggest cost. The bottleneck is still there.
A manufacturing company I know built a customer communication bot while their scheduling team was still doing manual cross-referencing for three hours a day. A consultancy rolled out an AI-enhanced CRM while onboarding new clients via email chains and PDFs.
Not wrong choices. Just not the right first choices.
Finding the right starting point
Good automation strategy starts with an honest audit, not a wish list.
The first question I always ask is: where does time actually go? Not where people think it goes. When teams track their time for a real week rather than estimating, the answer is almost always different from what they expected.
The second question is about failure cost. Most processes have two costs: the time it takes when things work, and the time it takes when they don't. Some processes run smoothly 95% of the time and cause most of the damage in the remaining 5%. That damage is invisible in standard capacity planning.
The third is volume. A process that takes 45 minutes daily for six people has more automation potential than one that takes three hours once a month. You do the math once, look at the result, and wonder why you didn't do it sooner.
And then there's complexity. The most time-consuming process isn't automatically the best starting point. A high-volume process with clear rules and clean input data delivers faster results than a complex one with many exceptions, even if the complex one looks more valuable on paper.
What makes a good first project
Three things that matter in practice.
Time saved should be measurable. Not "feels like less work." Measurable. Fewer hours per week, fewer errors. Without a measurable result, you can't build a case for the next project.
It should go live within six to eight weeks. A fast win builds more organizational trust than a nine-month project with first results late in the year. Speed of feedback matters more than scope in the early stages.
Real people should feel the difference. If someone on the team says their day actually got easier, that's more valuable than any slide deck. It's what gets the next project approved.
Looking impressive externally is not a criterion for the first project. That comes later.
What happens when you get it right
Once the first project delivers real results, something shifts. Teams stop asking "should we automate this?" and start asking what's slowing them down. The second project goes faster because the skepticism is lower. The third goes faster still.
Companies that pick the wrong first project often don't get a second one. The initiative stalls. Automation gets filed away as something they tried once.
The entry point matters more than most people realize.
The uncomfortable part
Starting automation means being willing to look at the unglamorous parts of your business first. The processes nobody includes in presentations. The tasks everyone knows about and nobody talks about.
That's where the value is.
Once the first project works, the results justify the second. By that point, you can go after the visible, externally-facing projects too — but on a foundation that holds.
Want to find out which processes at your company have real automation potential? Our free Automation Check gives you a clear picture in 30 minutes.