An owner decides to get organised. She starts with invoicing, because it is the thing she likes least and it feels like the obvious candidate.
Invoicing happens eleven times a month and takes about two hours in total. Automating it saves most of those two hours, which is genuinely useful.
Meanwhile she answers forty enquiries a month, each interrupting whatever she was doing, and loses perhaps four of them to slow replies. That is the thing worth several times the invoicing, and it is the thing she did not touch.
Test 1: how often does it happen
Daily beats monthly by a wide margin. Frequency is the strongest single indicator of where automation returns the most.
Test 2: is the answer the same every time
If it varies with judgement, it needs a person. If it is identical, it should have been written down long before anybody considered automating it.
Test 3: does it interrupt paid work
The strongest test of all, because the real cost is not the task but everything it displaces and the time taken to return to it.
Test 4: does delay cost money
Invoicing late costs a few days of cash. Answering late costs the job entirely, and that difference is what ranks the candidates.
Test 5: does it happen outside working hours
Anything arriving in the evening or at weekends is either handled automatically or not handled at all.
Why the answering wins on every test
High frequency, identical answers, constant interruption, immediate cost of delay, and a large share arriving when nobody is available.
Why invoicing feels like the answer
Because owners dislike it, and dislike is not the same as cost. The most irritating task is rarely the most expensive one.
Why scheduling comes second
Removing the back and forth about times is genuinely valuable, and it only matters once the enquiries are being answered at all.
Why follow up comes third
Solo owners almost never chase quotations, and automating a single timed message recovers real work for almost no effort.
Why bookkeeping comes later
Important, infrequent and better solved by a person than a system in most one person businesses.
What not to automate at all
Judgement, difficult conversations and anything that is actually the relationship, which is the distinction drawn in deciding what to delegate first.
The order that works
Answering, then scheduling, then follow up, then everything else, and each step makes the next one easier rather than harder.
The exception worth noting
Some businesses genuinely should not automate the first contact, which is the argument in when a saas should not automate first line support.
The measurement worth taking first
Count the enquiries for two weeks and count how many were answered within an hour. Owners almost always find the second number considerably lower than they assumed, and that gap is the whole business case.
The order nobody expects
Writing the answers down comes before automating anything. An owner who does that one step frequently discovers it solved half the problem before any tool was involved.
Start with the forty, not the eleven
An assistant answering the enquiries removes the interruptions, recovers the ones currently lost to slow replies and covers the evenings, which is worth several times the task an owner would have chosen instinctively. The method is set out in how to automate your top 20 support questions this, and the evening cost in the evening spent replying to messages from the day. See it working on the demo, compare on pricing, and the sector view sits under small business.