To pick your first automation, score each repetitive task on three numbers: how often it happens, how much it annoys you, and how long it would take to set up. The task with the best score wins, and you build that one first. No guessing, no analysis spiral, just a little fifth grade math over one weekend.
- The reason nothing gets automated is usually not a lack of time but a lack of a way to choose which task to fix first.
- A simple score built from frequency, annoyance, and setup time turns a vague backlog into a ranked list.
- High frequency plus high annoyance plus low setup time equals your best first automation.
- You do not need software to run this. A sheet of paper and a calculator finish the job in an afternoon.
- Building the wrong automation first is a common reason people quit before they see any payoff.
- The weekend you spend scoring is the cheapest investment you will make all quarter.
You keep saying the same sentence. “I don’t have time to figure this out.” So the repetitive work stays manual, and you stay busy doing it, and the thing that would do the work for you never gets built. The problem is rarely time. The problem is that you cannot pick your first automation because every task looks equally worth fixing, so none of them get fixed.
To pick your first automation, you score each repetitive task on three numbers, frequency, annoyance, and setup time, then build the task with the best combined score first. That is the whole method. A first automation is the single repetitive task you choose to hand off to a system before any other, and choosing it well is what keeps you from quitting halfway through. This is grade school arithmetic applied to your own week.
The reason this matters is that most people stall at the choosing step. They open a blank page, feel the weight of twenty annoying tasks, and close the page. Scoring removes the weight. You stop debating and start counting.
Why does knowing which task to fix first feel impossible?
Knowing which task to fix first feels impossible because your brain treats every annoying task as urgent, which flattens them into one undifferentiated pile of work that all seems equally worth your weekend. When everything feels urgent, nothing gets ranked. And when nothing gets ranked, you default to the task that screamed loudest this morning, which is almost never the smartest first build.
There is a quieter force at work too. Work expands to fill the time you give it, a pattern named Parkinson’s Law. If your days are full of manual tasks, those tasks will happily swallow every hour you have, leaving no room to build the thing that would shrink them. You can read more about how Parkinson’s Law crowds out deeper work and why busy is not the same as productive.
The reason nothing gets automated is almost never a lack of time. It is the absence of a simple way to decide which task deserves the fix first.
What is the scoring method for picking a first automation?
The scoring method is a quick ranking system where you rate each repetitive task on frequency, annoyance, and setup time, then combine those ratings into one number so the highest-scoring task becomes your obvious first build. Frequency is how often the task happens. Annoyance is how much you dread it. Setup time is how long building the automation would take.
Here is how I score each one. Frequency and annoyance get a 1 to 5 rating where higher is worse, meaning more painful. Setup time also gets a 1 to 5 rating, but here lower is better, so a quick build scores 1 and a monster build scores 5. The formula is frequency plus annoyance minus setup time. The higher the result, the better the candidate.
| Factor | What it measures | Scale |
|---|---|---|
| Frequency | How often the task repeats | 1 rare to 5 daily |
| Annoyance | How much you dread doing it | 1 fine to 5 soul-draining |
| Setup time | How long the build would take | 1 one hour to 5 many days |
Let me show the math on a real-looking example. Say you manually send a welcome email to every new lead. That happens daily, so frequency is 5. You hate it, so annoyance is 4. It would take an afternoon to build, so setup time is 2. Your score is 5 plus 4 minus 2, which equals 7. Now compare that to reformatting a monthly report. Frequency 1, annoyance 3, setup time 4. That score is 1 plus 3 minus 4, which equals 0. The welcome email wins by a mile.
How to find the tasks worth scoring
You cannot score tasks you have not named. Before the math, run a short time audit, which simply means tracking where your hours go for a few days. Without tracking, the biggest time wasters stay invisible because they hide inside your routine. A guide on how to conduct a time audit walks through the tracking part if you have never done one.
Keep a running list every time you catch yourself doing something repetitive. Copying data between two places. Sending the same reply. Chasing a status update. By Saturday you will have ten to fifteen tasks on paper, which is plenty to score.
High frequency plus high annoyance minus low setup time equals the automation you should build first. Everything else can wait its turn.
Why setup time has to count against the score
A task can be wildly annoying and happen constantly, but if building its automation takes two weeks, it is a terrible first project. First builds need to finish. A finished automation that saves you twenty minutes a day teaches you the whole process and gives you a win you can feel. That win is what gets you to build the second one. Subtracting setup time protects you from starting with the hardest thing in the pile.
How does this compare to other prioritization frameworks?
This scoring method is a stripped-down cousin of bigger prioritization frameworks, trading their broad categories for three concrete numbers you can rate in seconds, which makes it faster to run when the only question is which task to automate first. The famous frameworks are excellent, but most are built to sort an entire business, not to pick one weekend project.
| Method | Best for | Speed to decide |
|---|---|---|
| Frequency and annoyance and setup score | Choosing one automation to build next | Minutes |
| Impact effort matrix | Sorting a project backlog visually | An hour or two |
| Eisenhower matrix | Sorting all tasks by urgency | Ongoing |
If you want to see a plotting-style approach, the impact effort matrix groups tasks into quick wins and major projects by charting them on two axes. It is worth knowing. For picking a single first automation, though, three numbers beat a four-quadrant chart because you finish and move on.
Running the whole thing in one weekend
- Friday night and Saturday, list every repetitive task as you notice it.
- Sunday morning, score each task on frequency, annoyance, and setup time.
- Do the arithmetic. Frequency plus annoyance minus setup time.
- Sort the list by score, highest first.
- Build the top task. Only the top one.
That is the full weekend. By Sunday afternoon you will have either built your first automation or know exactly what you are building next Saturday. Either way, the choosing paralysis is dead. You traded it for a number.
Parkinson’s Law got its name from Cyril Northcote Parkinson, a British naval historian who first described it in a 1955 essay for The Economist. He noticed that work expands to fill the time available for its completion, an observation that explains why a task you could finish in an hour somehow takes all day when you have all day.
The version of this I run into most often is a business owner who has ten automation ideas and has built zero of them. When I ask why, the answer is never time. It is that they keep rebalancing the list in their head, swapping the top spot every week, so no single task ever stays number one long enough to get built. The moment they score the list on paper, the top choice stops moving. One client sat with the three numbers for twenty minutes, found a daily copy-paste task that scored highest, and had it running by the next weekend. The scoring did not make them smarter. It made the list hold still.
How do I pick which task to automate first?
Score each repetitive task on three things and let the number decide. Rate frequency and annoyance from 1 to 5 where higher means worse, rate setup time from 1 to 5 where lower means faster, then calculate frequency plus annoyance minus setup time. The highest score is your first automation. It takes about an hour once your task list exists.
How long does it take to score my tasks?
Scoring a list of ten to fifteen tasks takes under an hour. The slower part is noticing the tasks in the first place, which is why you spend a day or two just writing them down as they happen. Once the list exists, the math itself is a few minutes with a calculator or a sheet of paper.
What if two tasks get the same score?
Break the tie with setup time and build the faster one. A tie means both tasks are worth doing, so pick the one you can finish sooner. Finishing a first automation matters more than optimizing it, because the completed build teaches you the process and gives you momentum for the next one.
Do I need software to run this scoring method?
No. A sheet of paper, a pen, and a calculator finish the job. A spreadsheet makes sorting easier if you have more than fifteen tasks, but the method is deliberately low-tech so nothing stops you from starting. The point is the decision, not the tool you record it in.
Why do I keep failing to build my first automation?
Usually because you start with the hardest or most dramatic task instead of the highest-scoring one. Big first builds stall out before they finish, which kills momentum. Subtracting setup time from your score protects you from this by pushing quick, high-value wins to the top of the list where a first project belongs.
Can this method work for a whole team, not just me?
Yes, with one adjustment. Have each person score the shared tasks on their own annoyance, then average the ratings. Frequency and setup time stay the same for everyone. The averaged score surfaces the task that drains the most people, which is often a stronger first automation than any single person’s pet peeve.
Spend one weekend on the list. Write down every repetitive task Friday and Saturday, score them Sunday, and build the top one. Keep the scoring sheet somewhere you will find it again, because you will run this every time you are ready for the next build.
Save the scoring math for the weekend you finally sit down to build. The method is not going anywhere, and neither is that backlog until you count your way through it.
… and see more of my posts in your Google results.
