PHASE 05
BUY, CONFIGURE,
OR BUILD.
Three options, not two. Most small businesses choose between buying and building when the answer is usually the middle one, which is buying something adequate and configuring it properly.
The three options
Buy
Off-the-shelf software used close to how it ships. Fastest to value, lowest cost, and you adapt your process to the tool's assumptions.
Configure
Off-the-shelf software set up deliberately around your process, with fields, stages, automations and integrations built to fit. Where most small businesses should land.
Build
Something made for you. Fits exactly, costs the most, and you now own it permanently — including every future change and every fix.
The middle option gets skipped because it is unsatisfying. Buying feels decisive. Building feels serious. Configuring feels like homework, and it is where the actual leverage sits for a business under a few dozen people.
When off-the-shelf is right
Buy when the work is genuinely standard. Accounting. Email. Scheduling. Document storage. Payroll. Basic CRM. These are solved problems, the products are mature, and your version of the requirement is not meaningfully different from everyone else's.
Buy also when speed matters more than fit. Something adequate running next week usually beats something perfect running in six months, because the adequate version teaches you what you actually need.
The real discipline here is resisting customization of things that do not need it. Every deviation from how a product ships is a thing to maintain, a thing that can break on an update, and a thing to explain to the next person who uses it. If the standard version works, run the standard version.
For the full context behind this, see the full story behind this work.
When configuration is right
This is the right answer more often than either alternative, and it applies when the work is common in shape but specific in detail.
Your sales pipeline has stages that are recognizable to any CRM and named according to how your business actually works. Your intake needs particular fields. Your follow-up sequence has your timing and your language. Your job tracking needs the stages your delivery process uses. All of that is configuration inside products that already exist.
The cost people underestimate is not money. It is that configuration requires you to have made decisions. You cannot configure a pipeline until you know what your stages are and what done means at each one. This is why the mapping and handoff phases come first — they produce the decisions that configuration implements.
Budget real time for it, and give it an owner. Configuration done in an afternoon by whoever had a free afternoon produces a system that fits nothing in particular.
When custom is justified
Rarely, and for narrow reasons. The honest list is short.
- The process is genuinely differentiating. Not just yours — actually part of why customers choose you, and no product supports it.
- The volume makes the math work. Custom pays on repetition. A process that runs a few times a month almost never justifies it.
- Nothing off the shelf comes close, which you can only know after seriously evaluating what exists rather than assuming.
- You can carry it permanently. Someone maintains it, changes it as the business changes, and fixes it when it breaks. Forever.
That last one is the one that ends most custom projects badly. The build gets funded and finished. Then the business changes, the person who built it is gone, and the thing everything depends on is understood by nobody. A custom system you cannot maintain is worse than a product that fits imperfectly, because at least the product has a support line.
The cost of leaving
Every option carries an exit cost and almost nobody prices it at the time of decision, which is exactly when it is cheapest to think about.
| Option | What leaving costs |
|---|---|
| Buy | Export the data, learn a new interface, rebuild simple settings. Usually the cheapest exit. |
| Configure | All of the above, plus rebuilding the configuration and automations you invested in. |
| Build | There is no exit. You rebuild, or you keep maintaining something nobody currently understands. |
Before committing to anything, ask three questions and get real answers. Can I export my data in a format another system can read, and can I test that this week. What is the contract term and what happens at renewal. If I cancel, how long do I keep access to my own records.
A vendor confident in their product answers all three without friction. Hesitation on the export question is a finding, not a formality.
Frequently asked
Questions people actually ask
Should a small business ever build custom software?
Occasionally, and later than most people think. The test is whether the process is genuinely differentiating, runs at real volume, is unserved by existing products, and can be maintained permanently by someone who will still be there.
What is the most common mistake here?
Buying before the process is defined. The tool then gets configured around confusion, and six months later the conclusion is that the software was wrong, so a different tool is bought and configured around the same confusion.
How much should configuration cost?
Less in money than people expect and more in decisions than people expect. The expensive part is the clarity it demands: stages, owners, definitions of done. That work is yours whether you pay someone to implement it or not.
Is an all-in-one platform better than several tools?
Sometimes, for the integration it saves. The tradeoff is that all-in-one products are usually adequate rather than good at several of the things they include. Judge it against the two or three functions you actually depend on.
What if we outgrow the tool?
That is a normal and healthy outcome, and it is why the exit questions matter at purchase time. A tool you outgrew in three years with a clean export was a good decision, not a mistake.
Make your next move
A year from now, what will you be glad you started today?
You don't need another promise that everything will be easy. You need something useful to learn — and a next step you're willing to take.