Software isn't the answer to every problem. There are tons of human challenges where software won't help, or might even make the problem worse.

When it's difficult to enumerate the rules for how a system should work. Or, when there are a ton of special cases or exceptions to the rules, then writing software might be more trouble than it's worth.

I work as a software consultant. There are many times when I saw clients trying to write code to solve messy, human problems. What they needed was better organization, operating procedures, and consensus among stakeholders. None of those are problems software can solve.

How do you decide when a problem is a good candidate for software?

• Are there repetitive steps that have predictable, repeatable outcomes?
• Does having a single source of truth matter?
• Would the parties involved actually use the software & would it make their lives easier?

The answers to the above are not always "yes." Before you build, make sure the problem is a good one for software to solve.