Most automation projects that disappoint were not built badly. They automated the wrong thing.
The wrong thing is usually whatever was most annoying that week, which is not the same as whatever costs the most.
Four questions, in this order
Take any process you are considering and put it through these. It takes about ten minutes each.
How often does it happen?
Something that happens forty times a week is worth far more attention than something that happens twice a month, even when the twice-a-month task is more irritating. Frequency beats annoyance nearly every time, and annoyance is what usually gets picked.
Does it happen the same way each time?
A process with one path is straightforward. A process with a dozen exceptions is not, and it is where budgets disappear. If your answer contains the phrase it depends more than twice, that is a warning.
What does it cost when it is not done?
Some tasks are irritating but harmless if they slip a day. Others cost you the job. An unanswered enquiry and an unfiled receipt take similar effort and have nothing in common in consequence.
Who has to be involved?
If one person does it, you can automate it and check it against their judgement. If four people do it four ways, the first job is agreeing which way is right. That is often worth doing on its own, even if you never automate anything.
The scoring, honestly
The best first project is high frequency, consistent, expensive when missed, and owned by one person.
That combination is rarer than you would hope. What you are usually choosing between is high frequency but messy, or clean but rare. Take the messy frequent one and spend the time tidying it up, because you get the tidying value even if the automation is delayed.
The one most people should start with
For most owner-run businesses in the GTA, it is the first response.
Enquiries and inbound calls score well on all four questions. They happen constantly. They are consistent, because the first exchange is nearly the same every time. Missing one costs real money. And they usually route through one person, which is generally you.
That is not a coincidence, and it is not a sales line. It is what the four questions produce for a business where the owner is also the bottleneck.
What not to start with
Anything involving money moving without a human, on day one.
Not because it cannot be done, but because you should build trust in a system on work where a mistake is recoverable. Start where being wrong means an awkward email, not a wrong payment.
Also avoid anything you have not written down. If the process only exists in someone’s head, writing it down is the project, and it is worth doing before anyone builds anything.
Do this in an afternoon
List every recurring task in the business. Do not filter, just list.
Score each on the four questions from one to five. Add them up. Look at the top three.
Then ignore the scores for a moment and ask which of those three you would actually trust a system to do. That instinct is usually right, and it is the difference between a project that lands and one that gets switched off after a month.
If you want that done with someone who has scored a lot of these, the free CEO audit is exactly this exercise, run properly, with a written blueprint at the end.
