Three signs the boring bit of your day is fixable

Sign one: you do the same five-step thing twice a day in a slightly different order

Open inbox. Find the new orders. Copy customer details into the spreadsheet. Update stock figures. Email the warehouse. Some days the email comes before the stock update. Some days you skip step three because there were no new orders. The five steps are the same.

This is the most common pattern, and the most fixable. The variation between days is rarely about judgement. It is about whichever bit of input arrived first that morning. The work itself is mechanical.

Test it: write down the five steps. If you can write them down in the order they happen, you have something that can be built in code. If they vary, write down the rule that decides the order. If you can write the rule, you have something that can be built in code.

Sign two: a load-bearing spreadsheet that one person knows

Every business has at least one. It started as a simple tracker. It now has six tabs, ninety formulas, four colour-coded states, and three sheets that pull from it. One person on the team knows how it works. They are slightly stressed when they go on holiday. The business runs on it.

The spreadsheet is doing useful work. The problem is not that it exists. The problem is that the logic only lives in one head and one file, and the cost of an error is unbounded. The fix is not to replace it with a piece of software the team has to learn; the fix is to formalise the logic the spreadsheet already represents and add a layer of validation around it.

Test it: ask the person who maintains it to describe what would break if it disappeared overnight. If they go pale, you have one of these.

Sign three: the email-PDF-paste loop between two tools that already have APIs

This one is the most expensive and the easiest to spot. Tool A produces a thing. Someone exports it as a PDF or CSV. They open tool B. They paste the values in, possibly transforming them in the middle. They send a confirmation. The two tools both have public APIs that they are deliberately not using, often because the loop predates the integration becoming available and nobody went back to revisit it.

Test it: pick any task in your week that involves moving information from one piece of software to another by hand. Look up whether the two tools have integrations with each other or with a workflow layer. The answer is almost always yes.

What is not a sign of fixable

Workflows that need real judgement, where the right next step depends on context only a person can read. A salesperson deciding how aggressive to be with a follow-up. A support agent reading the tone of an angry email and deciding how to respond. A finance lead deciding whether an unusual expense category warrants a conversation with the department head. These are not plumbing. They are the work.

Automating the plumbing around the judgement is the goal. Automating the judgement itself is usually a category error and rarely pays back.

How to test it for the cost of five minutes

We built a free audit tool that asks ten questions about how time gets spent in your business and produces an estimated weekly time-loss figure plus three personalised recommendations. It takes about five minutes and you see the results without needing to give us an email address. Most teams who take it find at least one workflow on the list that they had not consciously thought of as fixable.