What should I fix first, and what can wait?
Fix what could hurt you first, then what returns cash, then what returns your time, then what helps you grow. In that order. Not the loudest problem, and not the one that would look best. A broken website is annoying. Customer data in a shared spreadsheet with a shared password is a risk that can end a small business.
Then be strict. Three fixes at a time, each with one owner and one finish date. Write a second list of things you will leave alone on purpose, with a date to look at them again. Deciding what to ignore is the part most owners skip, and it is what makes the rest possible.
The one hour method
Pass 1: write it all down (10 minutes)
Every problem, one list, no sorting. Getting it out of your head is half of it.
Pass 2: the harm filter (10 minutes)
Move to the top anything that could cause legal, financial or data harm if you leave it. Unpaid tax, VAT or PAYE. No backups. Shared passwords. Customer details in loose spreadsheets. No contracts with key customers. No reliable way to be paid. These go first because one incident can cost more than every other item combined.
The scale is real. The government’s Cyber Security Breaches Survey 2025/2026, published on 30 April 2026, found that 43% of UK businesses reported a cyber security breach or attack in the previous twelve months. By size it was 42% of micro, 46% of small and 65% of medium businesses, and phishing was the most common type at 38%.
Source: Cyber security breaches survey 2025/2026, Department for Science, Innovation and Technology and the Home Office, GOV.UK. The National Cyber Security Centre has a free set of first steps for small organisations at ncsc.gov.uk.
Pass 3: score the rest (20 minutes)
Give each remaining problem a score from 1 to 5 on four questions.
- Cash: how much money does fixing it bring in or save in the next ninety days?
- Time: how many hours a week does it give back to you?
- Effort: how hard is it to fix? (A low score is good here.)
- Dependency: does something else need this done first? If so, it drops behind that.
Add cash and time, then divide by effort. The highest score goes first, unless it depends on something that has not been done yet.
Pass 4: three lists (20 minutes)
- Fix now. Three at most.
- Next quarter. Real, but not yet.
- Leave alone on purpose. Each with a review date.
Good candidates to leave alone: anything annoying but cheap, anything that fixes itself when a bigger fix lands, and anything where the cost of fixing is more than a year of the pain it causes.
Two rules that keep it honest
- Three live fixes at once. A fourth means one of the three has stalled. Finish or drop one.
- A fix that needs you every week is not finished. Finished means it works when you are not looking at it.
A worked example (invented, for illustration only)
Take an imaginary twelve person joinery firm. Its list: customer quotes kept in a personal inbox and passwords shared on a sticky note; invoices raised weeks after the job; the owner re-typing every order into two places; a new website; a second estimator; branded vans.
- Fix now: 1. shared passwords and backups (harm). 2. Invoice on job completion (cash). 3. Enter each order once (time).
- Next quarter: the second estimator, once quotes flow through one place.
- Leave alone on purpose: the website and the vans, until the cash cycle improves. Look again in three months.
None of that is a real business. It is there to show the shape of the answer.
Why the website is so rarely the first fix
It is the most visible problem and the least urgent. People polish what they can see. Cash, risk and time sit underneath, and fixing those usually makes everything else cheaper.