Cost of invisible work
The calculation, with a table you can read your own row from.
Read the articleWorth reading
When does business
automation pay off?
Most conversations about automation focus on what can be done. The more useful question is when it actually pays, and what has to be true for it to work.
At a glance
- The three conditions
- Repetitive, material, and not going away on its own
- The test
- Could you write the steps down and would they hold for the next fifty times
- When to skip it
- Unstable processes, rare tasks, and decisions that are the valuable part
- Where it starts
- The Time Leak Audit, from £950 excluding VAT
Condition one
The task is
genuinely repetitive.
Automation works on patterns. If a task follows the same steps with different data, the same process with a different client name, invoice number and date, it is a candidate. If it needs fresh judgement each time, or varies so much that no template holds, a system adds complexity rather than removing it.
The test is blunt and it works: could you write down the exact steps in order, and would those steps still be right the next fifty times you did it? If yes, it is probably worth looking at.
Condition two
The annual cost
is material.
Building anything has a cost, and that cost has to be justified by the value of the time it frees. This is why the calculation comes before the build rather than after it.
Work out the annual cost of the task at your own rate, then compare it against a real quotation for fixing it. Not an estimate, not a rule of thumb, and not a figure from a website. A ratio you can defend to yourself is the only evidence that matters here.
Condition three
The task is not
going away on its own.
Some admin problems solve themselves. A new hire absorbs the work, or a change in the business model makes the process obsolete. If that is likely within six to twelve months, the window shortens and the case weakens.
Most do not go away. Client onboarding does not get easier as you take on more clients. Invoicing does not simplify as revenue grows. What is frustrating at ten clients is still frustrating at twenty, and by then it takes twice as long.
The honest part
When it does not pay off
Being clear about when this is the wrong answer matters as much as knowing when it is right.
Do not automate
- A task that happens less than once a week and takes under thirty minutes. Building and maintaining it costs more than doing it.
- A process that is not stable. If the steps change often, you will spend longer updating the system than doing the work.
- Work that needs nuanced judgement that shifts with context. If the decision is the valuable part, keep the decision.
- A business that is very early. Processes still being discovered should not be set in concrete yet.
- A genuine capacity problem. Sometimes the right answer is a person.
What good looks like
- It runs reliably without anyone watching it
- It is transparent enough that a failure is obvious
- It handles the common case and routes the exceptions to a person cleanly
- It was built after someone mapped the process, not after someone chose a tool
- Your team can run it without the person who built it
The most common failure is not technical. It is building the wrong thing: automating a process nobody properly understood, at a step that was not the bottleneck, with a tool chosen before the problem was diagnosed. That is why the diagnosis always comes first.
Where to start
Do the calculation
before you spend anything.
Which tasks repeat most often? How long do they actually take, measured rather than remembered? What does that cost in a year at your effective rate? Which of them are stable enough to be worth fixing safely?
That is what the Time Leak Audit produces: a clear picture of where the time goes, the cost of each category, and a ranked list in order of impact. Under two hours of your time, from £950 excluding VAT. If the numbers do not support building anything, the report says so.
Before we talk
What people ask
about payback.
Three questions that decide whether this is worth your money.
What payback period should I be looking for?
That depends on your own cost of capital and how stable the process is, which is why I will not publish a target multiple. What I will do is calculate your annual cost properly and put it next to a real quotation, so the ratio is yours rather than mine.
Can you tell me the cost before the Audit?
I can tell you the Audit price, which is from £950 excluding VAT. I cannot honestly price a build before mapping the process, and any consultant who does is guessing or padding.
What if only one of the three conditions holds?
Then it probably does not pay off yet, and you should hear that before you spend anything. The free fifteen minutes is usually enough to tell.
Keep reading
Related
Your business
Run the numbers
for your business.
Tell me what your week looks like and I will tell you whether removing the work makes financial sense, and what would have to be true for it to pay.
15 minutes · No charge
Prefer to write? [email protected]