Stop Carrying Rocks: The 3-Step Lazy Audit
- Keith Mortier
- Feb 14
- 5 min read
Updated: Aug 5
The Lazy Audit is a 15-minute weekly review with three steps: write down the three tasks you dreaded most this week, count every click one of them takes, then ask whether stopping it entirely would cost you money. If no, delete the task. If yes, automate it. Tasks over 20 clicks are broken, not just slow.
The hardest-working person in the room is usually the most inefficient.
Early in my career I wore "busy" like a badge. Twelve-hour days, hundreds of emails, constant pings. I looked productive. I was drowning in manual friction.
Then I got lazy, because I was overwhelmed. And laziness turned out to be useful — it forced me to build a system.
The goal isn't to be the person who carries the most weight. It's to be the person who builds the cart.
What is the Lazy Audit?
A 15-minute exercise I run every Friday. Three steps, no tools, no spreadsheet. It consistently finds work that either shouldn't exist or shouldn't be done by a person.
It's called the Lazy Audit because the instinct it relies on — not wanting to do something tedious again — is the most reliable diagnostic signal you have. People don't dread work that flows.
Step 1: What are the three tasks you dreaded most this week?
Write them down. Don't overthink it. What made you groan? What did you procrastinate on? What felt like a waste of your brain?
For most people the list is some combination of invoicing, data entry, scheduling, report generation, list management, and manual follow-up.
These aren't just annoyances. Your frustration is pointing directly at the processes that need to be automated or eliminated.
Step 2: How many clicks does the task actually take?
Pick one task from the list. Do it one more time, but count every click, every tab switch, every copy-paste.
If it takes more than 20 clicks, it's broken. Not "could be improved." Broken. Twenty clicks means at least three or four handoffs between systems that don't talk to each other.
I did this with my own invoicing. Thirty-four clicks to generate and send one invoice. It's now two: review and approve. The system does the rest.
The click count matters because it's objective. "This takes forever" is an opinion. "This takes 34 clicks" is a number you can put next to a build estimate.
Step 3: What happens if you just stop?
For each task, ask one question: if I stopped doing this entirely, would we lose money?
If no — delete it. Stop doing it. You'd be surprised how much work exists purely out of habit, producing nothing. This is the step people skip, and it's the highest-return one, because deleting a task costs nothing to implement.
If yes — automate it. If it drives revenue but you hate it, that's the clearest possible signal that a machine should handle the repetitive part while you handle the judgment.
Why does being busy feel productive when it isn't?
Because your brain is compensating, and it's expensive.
Researchers at UC Irvine ran an experiment with 48 subjects comparing interrupted work against uninterrupted work. The interrupted group finished faster — 20.3 minutes versus 22.8 for the control.
But they reported significantly higher stress (9.46 versus 6.92 on a 20-point scale), along with more frustration, more time pressure, and more effort. They kept up by working harder and paid for it in strain.
That's the whole trap. Fragmented, interrupted, tab-switching work doesn't feel slow. It feels intense — and intensity reads as productivity right up until you burn out or something gets dropped.
The measured cost of that fragmentation is real: Harvard Business Review found workers toggle between applications about 1,200 times a day, losing just under four hours a week — around 9% of working time — purely to reorienting.
You can't feel 9%. You can only measure it.
What does this look like when it works?
A three-person team was spending 12 hours a week compiling compliance reports from multiple sources. They hated it. The click count was astronomical. But they couldn't kill it — compliance isn't optional.
So the Kill Rule pointed at automate. We automated the data aggregation, the formatting, and the first-pass quality checks. The team now spends about two hours reviewing and approving instead of twelve building.
Ten hours a week back. They didn't need a transformation program. They needed someone to look at the Hate List, count the clicks, and build the cart.
What's the most common mistake in process improvement?
Optimizing something that shouldn't exist.
The biggest trap in operations isn't doing things slowly. It's doing unnecessary things efficiently. Before you speed anything up, ask whether it should exist at all — which is exactly why the Kill Rule comes before the automation question, not after.
If you need a broader diagnostic before choosing a task, start with the five signs your business needs an AI audit.
Run the Lazy Audit this Friday. Write the Hate List. Count the clicks. Apply the Kill Rule.
Build the cart. Stop carrying the rocks.
Frequently Asked Questions
How long does the Lazy Audit take?
About 15 minutes a week. Step one takes two minutes, step two takes as long as the task you're counting, and step three takes about a minute per item. It's designed to fit in the gap at the end of a Friday, because an audit you have to schedule is an audit you'll skip.
How many clicks is too many for one task?
More than 20 clicks means the task crosses at least three or four systems that aren't connected. That's the threshold where the problem stops being the task and starts being the gaps between your tools. Under 20 and you're usually looking at a task that's simply long, not broken.
Should I automate a task or eliminate it?
Ask whether stopping entirely would cost you money. If not, eliminate it — that's a zero-cost win and it's more common than people expect. If it does drive revenue but the work is repetitive, automate the repetitive portion and keep the judgment calls with a person. Automating something that should have been deleted just makes the wrong thing happen faster.
What if I hate a task but it needs human judgment?
Split it. Most tasks that "need judgment" only need it at one specific point — the approval, the exception, the call on an edge case. The gathering, formatting, and routing around that moment usually doesn't. The compliance reporting example above went from 12 hours to 2 precisely this way: the review stayed human, everything feeding it didn't.
Can I run this on my team's work instead of my own?
Yes, and it works better, but ask them to write their own Hate List rather than guessing it for them. The tasks a manager thinks are painful and the tasks that actually are painful are rarely the same list — and the gap between those two lists is usually the most useful thing the exercise produces.
Found something on your Hate List you can't kill?
A Find the Leaks call is 20–30 minutes and free. Bring your worst task and we'll tell you what it would take to automate the repetitive part of it. You keep the answer whether we work together or not.
Sources
- Gloria Mark, Daniela Gudith, Ulrich Klocke. "The Cost of Interrupted Work: More Speed and Stress." CHI 2008.
- Rohan Narayana Murty, Sandeep Dadlani, Rajath B. Das. "How Much Time and Energy Do We Waste Toggling Between Applications?" Harvard Business Review, August 2022.
