Most organisations will never need fully custom software, and most of those that do still rely on off-the-shelf products for a large part of their operation. The useful question is not “build or buy?” but “which parts of the business deserve software shaped exactly around them?”
Start with the process, not the software
Before comparing products or quotes, write down how the work actually flows today: who does what, where the information lives, and where it gets copied, checked or lost. Two things usually become obvious. Some processes are standard for your industry and a mature product handles them well. A few are specific to how you operate, and that is where the friction is.
When off-the-shelf is the right answer
- The process is standard. Payroll, general ledger accounting and email are solved problems. Building your own is rarely worth it.
- A mature product covers most of the need. If a product does 85% of what you need and the remaining 15% is preference rather than necessity, adapt the process.
- Speed matters more than fit. A subscription can be running this week. Custom software takes longer to plan and build.
When custom software is worth it
- Your process is an advantage. If the way you price, schedule, approve or deliver is what makes you better, flattening it into a generic tool gives that advantage away.
- You are stitching tools together by hand. When staff spend hours a week re-keying between systems, the cost of the workaround often exceeds the cost of building the missing piece.
- Constraints rule out the alternatives. Compliance obligations, data residency or integration with an existing system can make generic products unsuitable.
- The software is the product. If you plan to sell the software to others, you need to own it.
The answer is often both
The highest-return projects we see are frequently the smallest: keep the products that work, and build the integration, portal or workflow that connects them. A reliable link between your practice system and your accounting platform, for example, can remove more manual work than replacing either.
Build the part that is unique to you. Buy the parts that aren’t. Connect them properly.
Compare total cost, not the first invoice
Subscriptions look cheap per seat and expensive over years. Custom software looks expensive up front and cheaper once it is in place. Neither comparison is complete without the cost of the workarounds: the hours spent copying data, fixing errors and training people on processes the software doesn’t support. Estimate those honestly; they are usually the largest number in the comparison.
Questions to ask before deciding
- Which steps in this process are genuinely different from how our competitors work?
- How many hours a week do we spend moving information between systems?
- What would break if our main software vendor changed pricing or direction?
- Do we have obligations (privacy, data residency, audit) that limit our options?
- Who will maintain the software in three years, and how?
If the answers point towards standard products, choose well and configure them properly. If they point towards custom software, start with discovery: a short, focused piece of work to understand the problem and design the smallest system that solves it.