Thoughts / 03
Sometimes the answer is a setting you already pay for.
A fair number of the problems described to me do not need anything built. They need a feature switched on in software the business is already paying for every month.
Why this happens so often
Business software is sold on the size of its feature list and bought during a fortnight when somebody had an urgent problem. That person solved the urgent problem, learned the three screens it required, and stopped. Two years later the company is using a fraction of what it pays for, and nobody has been back to look since, because looking is not anybody's job.
Meanwhile the vendor added an integration with the exact other system you use. They announced it in a release note that went to an inbox nobody reads.
What to check before you hire anybody
List every piece of software you pay for monthly. For each one, ask the person who uses it most what it does that they are not using, and then go and read the vendor's integrations page. Not the marketing page. The page that lists what it connects to.
This takes an afternoon and it regularly finds something. A report that already exists. An export that would end a weekly copy-paste ritual. A connection between two tools that someone assumed would need custom work.
What I actually want to build
I take a few projects at a time, so each one uses up something I cannot get back. I would rather spend that on software somebody actually opens every day than on a thing that quietly goes unused because it duplicated a feature you were already paying for.
The good projects are the ones where you can feel the difference afterwards. Somebody stops dreading Tuesday. A job that took an afternoon takes a minute. Nobody has to remember the steps any more. That is the work I am interested in, and pointing you at a setting you already own costs me nothing I wanted.
It also makes the rest of my advice worth something. When I do say a problem is worth building software for, you have some reason to believe I looked at the cheaper answers first.
Recognise your own week in that?
Bring the task. Thirty minutes, no cost, and a straight answer about whether it is worth building anything.
Bernot Software