KyrosWorks · Houston automation guide

What to automate first in a small business: the task that eats your week

A method you can run in one afternoon: list your ten most repeated tasks, time them, rank them, and pick the one task worth automating first.

Every owner has one task that quietly takes more of the week than anyone admits. It is the quote that gets retyped, the follow-up email nobody has time for, the invoice data someone copies into a spreadsheet on Friday afternoon. You know it exists. You probably cannot say how many hours it takes.

This guide gives you a method you can run in an afternoon, with a pencil and a phone timer. At the end you will know which task to fix first and why. You do not need to care about technology to do it. You need to care about hours.

Key takeaways

Why the first choice matters

The first thing you automate shapes how your team sees every project after it. If it works, your team trusts the next one. If it drags on, goes over budget, or never gets used, people decide that this kind of project is a waste of time.

Most first projects fail for the same reason: the owner picks the task that feels most painful, not the task that is easiest to fix and pays back the most hours. Pain is not a good guide. A task that makes you angry once a month may cost you less than a task that nobody notices because it takes twenty minutes a day.

So measure first. The method below takes one afternoon to set up and one week to run.

Step 1: List your ten most repeated tasks

Sit down with a sheet of paper, or ask your office manager to do it with you. Write down the ten things your company does over and over. Do not write the work you are proud of. Write the work that repeats.

Prompts that help:

Use plain names. "Quote a job from a phone call" is a better entry than "sales workflow." "Copy supplier invoice totals into the spreadsheet" is better than "accounts payable."

Ask the people who do the work, not only the person who manages it. The task that eats the most time is often invisible from the owner's desk.

Step 2: Time each task for one week

For five working days, have the person who does each task note two things: how long it took each time, and how many times it happened. A phone timer is enough. A tally on a sticky note is enough.

Do not trust estimates. People guess low on the tasks they dislike and high on the tasks they enjoy. A week of actual numbers is more useful than an hour of discussion.

If a task happens once a month, time it once and write down the frequency. Multiply later.

Step 3: Rank by hours per week

Now do the arithmetic. For each task, multiply the minutes per time by the number of times per week, and divide by sixty. That gives hours per week. Sort the list from the top.

Here is an example of the shape of the sheet. These numbers are made up to show the method. Yours will be different.

TaskMinutes eachTimes a weekHours a week
Quote a job from a phone call25125.0
Follow up on quotes with no reply5302.5
Copy supplier invoice totals into the sheet4402.7
Schedule next week's jobs6011.0

Two things usually surprise people. The biggest line is rarely the one they expected. And several small tasks of a few minutes each can add up to more than the one big task everybody complains about.

Step 4: Pick the one that is boring, rule-based, and frequent

Look at the top of your ranked list and apply three tests. A good first task passes all three.

Boring. Nobody on your team would be sad to lose it. It does not use their skill. They do it because it has to be done, not because they are good at it.

Rule-based. You could write the steps on one page, and two different people following the page would get the same result. If the quote always uses the same price list, the same markup, and the same five questions, that is rule-based. If every quote needs a fresh decision, it is not.

Frequent. It happens several times a week at least. A task that happens four times a year is a poor first choice even if each time is painful, because the payback is slow and the tool gets little use.

Notice that the biggest line on your sheet may fail a test. A task that takes five hours a week but needs your judgment every time is not the first task. The second line on the list, which passes all three tests, is.

When two tasks tie, pick the one where a mistake is easier to catch. Pick the one whose result you can see.

What not to pick first

Some tasks look attractive and are the wrong place to start.

Anything that needs judgment every time. Deciding which supplier to use on a one-off job, or how to handle an unhappy customer, depends on context that changes. A tool can help you prepare, but it should not be the first thing you build.

Anything that touches payroll. If the tool makes a mistake, someone gets paid the wrong amount, and that is a problem for a real person and for your tax filings. Learn on something less costly.

Anything that touches compliance. Safety records, licensing, tax filings, and anything an auditor or an agency may ask to see all carry penalties for errors. Leave these until you have a tool you trust and a person who checks its work.

Anything that depends on records nobody can find. If the data lives in three people's heads and a box of paper, the first job is cleaning that up, and it is not automation yet.

Anything with patient or client records, in a practice. A dental or medical office has privacy rules that add cost and care to any tool. A practice is better off starting with something that carries no patient data, such as reordering supplies, and moving to the sensitive work later with proper advice.

A useful way to hold this: the first build should be something where, if the tool is wrong on a Tuesday, you notice by Wednesday and fix it at no real cost.

Why the first build should be small and fixed-price

Once you have your task, the temptation is to ask for everything: a full system, every task on your list, connected to everything else. Resist it.

A small first build has three advantages.

It gets finished. One workflow, installed on your own accounts and phone, can be built and in use within a few weeks. A large project takes months, and months is long enough for the owner's attention to move on.

It gets used. People adopt a tool that does one thing well. A tool that does twelve things half-well gets ignored.

It teaches you what you actually want. After two weeks of using the first tool, you will know what the second one should be, and you will describe it better than you could have on day one.

Fixed price matters for the same reason. When the price is written down before work starts, the builder has a reason to keep the scope tight and you have a reason to say what you actually need. Hourly billing rewards the opposite. Ask for a named deliverable, a price, and a date.

This is how we work at KyrosWorks. The Plan is $750: we walk your operation, time the repeated tasks with you, and hand you a one-page plan naming the first task, what it costs, and what it gives back. The Build starts at $4,000: one workflow, built and installed, with a one-page guide and thirty days of fixes. Care starts at $1,000 a month: we keep it running, fix what breaks, and make one small change a month. You can cancel Care any month. Nothing starts without a signed page of scope and price.

You can run the afternoon method in this guide yourself and never talk to us. If you want someone else to check the list, that is what The Plan is for.

What this looks like in Houston

The same method gives different answers in different kinds of company. These are examples of the pattern, not case studies.

A trade or fabrication shop. The owner is on a job site and the quote waits until evening. The ranked list shows quoting at the top. It passes the three tests: it is boring, it follows a price list, and it happens daily. The fix is a quote tool where the owner answers five questions on a phone and the customer gets a priced PDF in two minutes. We built this for Heights Mobile Welding, a one-person trade in the Heights, in two days.

An HVAC or plumbing company. The list often shows callbacks and follow-up at the top, because jobs are lost when nobody calls back for a day. The first fix is an automatic reply that confirms the request and gives the owner a clean summary, so no request is missed.

An industrial distributor. Requests for quote arrive by email, in different formats, and someone retypes each one into a spreadsheet before pricing it. That is boring, rule-based, and daily. It is a strong first task. Pricing the quote is not, because that takes judgment.

A freight broker. A small number of emails, sent again and again with different details: rate quotes, pickup confirmations, document requests. The list shows email at the top. The first fix drafts the routine ones and leaves the owner to approve each before it goes.

A professional practice. The front desk spends hours on reminders and paperwork, but most of it involves client or patient records. The first pick is a task that carries none, and the sensitive work waits until the first tool is proven.

In each case, the first task was chosen by the sheet, not by the owner's mood.

What to do this week

  1. Write your list of ten repeated tasks today. Thirty minutes.
  2. Ask each person who does a task to time it for five working days.
  3. Next Monday, do the arithmetic and rank the list.
  4. Apply the three tests to the top three lines and cross off what fails.
  5. Write one page: what the task is, who does it, how long it takes, and what a good result looks like.

That page is the start of a scope. It is also all you need to get a clear price from anyone who builds this kind of tool, including us.

Ready for a second opinion?

If you would rather have someone walk your operation with you, we will show you the price before any work starts.

See what a fixed-price build costs →

Questions owners ask

How do I know which task to automate first?

Pick the task that is boring, follows the same rules every time, and happens often. List your ten most repeated tasks, time each one for a week, and rank them by hours per week. The top of that list, filtered by those three tests, is your first build.

Should I automate payroll or compliance work first?

No. Payroll and compliance errors cost real money and can bring penalties, so they are a poor place to learn. Start with a task where a mistake is cheap to catch and fix, such as quoting, follow-up emails, or copying data between systems.

Why should the first automation be small and fixed-price?

A small first build gets finished and gets used. A fixed price means you know the cost before work starts, and a written scope means you know what you will get. If the first one works, the second one is easy to justify.

Do I need to buy new software to automate a task?

Often you do not. Many repeated tasks can be handled by a small tool built around the accounts and devices you already use. The first step is the list and the timing, not a software purchase.

Will Hays runs KyrosWorks in Houston. Sixteen years running purchasing and operations, now building the fix himself.

← Back to AI automation for Houston companies