1. Start with the work, not the software
Before comparing tools, write down one recurring task from start to finish: who starts it, what information moves, who approves it, and where it stalls. Most problems turn out to be one or two handoffs, not the whole process.
2. Fix, automate, build or buy
Fix the process
If the steps are unclear or duplicated, software will only make a confusing process faster. Simplify first.
Automate
If the steps are clear but repetitive, such as copying a form into a sheet, connect the tools you already have.
Buy
If many organizations share your need, such as email or accounting, an existing product is usually cheaper and safer than custom work.
Build
Build only where your process is genuinely specific, where you would otherwise pay for several tools that still do not fit, or where the system is part of what you offer.
3. Questions to ask before you spend
- How often does this happen, and how long does it take each time?
- What goes wrong when it fails, and who finds out?
- Who will maintain it in a year?
- Where will the data live, and who owns the account?
- What is the smallest version that would help?
4. Where AI helps, and where it does not
AI is useful for drafting, sorting, summarizing and searching text. It can be wrong, so keep a person responsible for anything that reaches a customer, a donor or a record. Avoid it for numbers you need to be exact, such as payments and totals, where ordinary code is more reliable.
5. Measure honestly
Pick one or two numbers before the project starts, such as hours per week or time to answer an inquiry. Measure them again a few weeks after launch. Treat the change as evidence about your situation, not a promise that applies to everyone.
6. Plan the handoff on day one
Accounts in your name, written notes on how it works, and a second person who knows where things are. A system only one person understands is the same problem in a new place.
Want help applying this to your situation? Tell us what is slowing your team down. See also the glossary.