Key takeaways
- Most “we need an app” problems are really ownership or handoff problems.
- Check the software you already pay for before buying or building anything.
- Build custom only when your process is how you win work and off-the-shelf tools force workarounds.
- Whatever you choose, pilot it with the people who will use it every day.
When a business outgrows its spreadsheets, the first instinct is often to shop for software — or to ask someone to build an app. Sometimes that is the right call. Often it is not. Before you spend money on a tool, it helps to answer one question honestly: is the problem the tool, or the way the work is organized?
At MAD 90 we treat software as one possible answer, not the starting point. Here is the simple framework we use with owners.
Step 1: Describe the work, not the software
Pick the process that frustrates you most — order intake, scheduling, quoting, job tracking — and write down how it actually happens today. Who starts it? What information do they need? Where does it wait? Where is information entered twice? Which decision has no clear owner?
Include the workarounds. The sticky note on the monitor, the side spreadsheet, the “just text me” step. Workarounds are the best evidence of where a process is failing.
Step 2: Try to simplify first
Many problems disappear when you remove a step, name an owner, or standardize a handoff. Ask:
- Can any approval or re-entry be eliminated?
- Does every step have one clear owner and a definition of done?
- Would a shared checklist or template solve most of the confusion?
If a simpler process fixes 80% of the pain, you may not need new software at all — and if you do, you will know exactly what it needs to do.
Step 3: Check what you already own
Accounting systems, Microsoft 365, Google Workspace and CRMs often include forms, automations and shared lists that go unused. Before buying anything, confirm whether a feature you already pay for can handle the job with a little configuration.
Step 4: Compare buying and building
If a new tool is justified, compare the two paths fairly:
- Buy off-the-shelf when your process is standard for your industry. You get proven software quickly, but you adapt to it, and subscription costs usually rise with each user.
- Use no-code or low-code tools for simple forms and lists with a small team. They are fast to set up but can hit limits as needs grow.
- Build a custom tool when your process is how you win work, when off-the-shelf products force your team into workarounds, or when you need several systems to work together. It fits the way you work, but it takes more time and needs a plan for support.
The most expensive mistake is building software around a process that was never fixed. A tool built on a broken workflow just breaks faster.
Step 5: Pilot with real users
Whatever you choose, start with one workflow and the people who will use it every day. Agree on a simple measure — time per order, errors, phone calls asking for status — and compare it against how things work today. Adjust before rolling it out more widely.
Where AI fits
AI features can help inside a tool: drafting quotes, sorting requests or summarizing notes. Treat them the same way: start with one task, protect sensitive information, and keep a person accountable for anything that reaches a customer. Our article Before You Buy an AI Tool covers this in more detail.
How MAD 90 helps: we start with a review of the workflow, recommend whether to simplify, configure or build, and — when a custom tool is the right fix — scope and deliver it as part of your engagement. Learn more about custom tools & apps.
Sources & further reading
This article reflects MAD 90’s practical perspective and editorial guidance. For the improvement method behind piloting and reviewing a change, see ASQ’s guide to the Plan–Do–Check–Act cycle.
