Every growing business eventually faces the same fork in the road: a process is painful, and something has to change. The instinct is to jump straight to “we need an app” or “let’s buy that software everyone uses.” Both can be right. Both can be expensive mistakes. This is a simple framework for deciding among the three real options – build, buy, or automate – without guessing.
The three options, briefly
- Buy an off-the-shelf product and use it as-is.
- Automate – connect or script the tools you already have so a manual process runs itself, without a big new system.
- Build a custom tool shaped to your exact workflow.
The questions that decide it
Run the painful process through these, in order:
- Is this process generic, or is it your competitive edge? If it is generic – accounting, payroll, email – buy it. If it is the thing that makes you better than competitors, lean toward building.
- Does a tool exist that fits without painful workarounds? If yes, buy or configure it. If every option requires bending your business out of shape, that is a signal to automate or build.
- Is the pain about connecting systems, not capability? If your tools each work but do not talk to each other, you usually do not need new software – you need automation between what you have.
- What is the true total cost over three years? Compare per-seat subscriptions plus workaround labor on the buy side against one-time build cost plus maintenance on the build side. Count the hidden costs, not just the sticker price.
- How fast do you need it, and how much will it change? Need it next week and unsure of requirements? Start by buying or automating. Stable, well-understood, and core to the business? Building becomes worth it.
A worked example
Say invoices are always late. Walk the questions: invoicing is generic (lean buy), and good accounting tools exist (lean buy) – but the delay is not the accounting tool, it is that job data lives on paper and never reaches it in time. The real pain is connecting the field to the office. That is an automation answer, not a new-accounting-software answer. Buying a second tool would not have touched the actual bottleneck. This is exactly the situation in our Post Falls construction case study.
A rule of thumb
Buy the commodity. Automate the glue. Build the thing that is actually your business.
The hybrid reality
Most healthy businesses end up with a mix: bought tools for solved problems, automation stitching them together, and one or two custom pieces where they genuinely differentiate. The mistake is defaulting to one answer for everything – building what you should have bought, or buying five subscriptions when a day of automation would have done it. The goal is not to be a “custom software company” or an “off-the-shelf company.” It is to put each dollar where it does the most good.
See the framework in action
Everything in this series is an application of this decision. The hidden costs of off-the-shelf software are what you weigh on the “buy” side. Whether you need a CRM – and which kind is this same question in miniature. And the case for a local developer is really about having a partner who will give you an honest answer here.
Common pitfalls
- Building too early. Custom software before you understand the process just locks in your guesses. Automate or buy first, learn what you actually need, then build the part that has earned it.
- Buying for the demo. The polished sales demo solves the vendor’s tidy example, not your real, messy workflow. Evaluate every tool against the way you actually work.
- Automating a broken process. Automation makes a process faster, including a bad one. Fix the process first, then speed it up.
- Forgetting maintenance. Everything you build or buy needs upkeep. Budget for it from the start so it does not ambush you in year two.
If you are staring at this fork right now and want an honest second opinion – one with no incentive to oversell you a build – that is exactly what we do. DevWharf helps Spokane and North Idaho businesses make the build-buy-automate call and then execute it. Start the conversation here.

Leave a Reply