When does a small business actually need custom software?
Custom software is not just for big companies. Here is how to tell when off the shelf tools are costing you, and when a bespoke build pays for itself.
Custom software has a reputation as a big company thing: expensive, slow, risky. So small businesses force their processes into off the shelf tools instead, or run on a sprawl of spreadsheets, and assume that is just how it is. Sometimes it is. But often there is a quiet, daily cost to making do, and at some point a bespoke tool is simply the cheaper option. Here is how to tell which side of that line you are on.
The signs off the shelf is costing you
You probably do not need custom software because you want it. You need it when the tools you have are getting in the way. Watch for these signs:
- You have workarounds for your workarounds. Every “almost fitting” tool spawns a manual step to patch the gap. When those steps stack up, they are a tax on every transaction.
- A spreadsheet runs something important. Spreadsheets are brilliant until the business depends on one. Then they become fragile, error prone and impossible to share safely.
- You pay per seat for features you do not use. Off the shelf software charges for everyone and everything, whether you use it or not, and the bill grows as you do.
- You are copying data between systems by hand. If a person is the integration between two tools, that is time and mistakes waiting to happen.
- You cannot get the report you need. When the answer to a simple business question takes an hour of manual digging, the tool is working against you.
If two or three of those ring true, the “almost fitting” tools are costing you more than you think.
What custom software actually fixes
Bespoke software does one thing off the shelf cannot: it fits your business exactly. The booking flow matches how you actually book. The dashboard shows the numbers you actually care about. The tool removes the manual steps instead of adding them. And you own it, so there are no per seat fees climbing every year and no platform deciding your limits.
The result is usually quiet. Things that used to take an hour take a minute. Mistakes that used to slip through do not. The friction you had stopped noticing because it was always there simply goes away.
We split the work on this site into app development when the thing is a web app customers or staff log into, and software development when it is a deeper internal system. The decision process is the same. The wrapper is not.
”But isn’t it really expensive?”
This is the fear that keeps businesses making do, and it is fair, because custom software has a history of runaway costs. The way to avoid that is scope discipline. Define the must haves first, build the simplest thing that solves the real problem, and fix the price before any work begins. That is exactly how we approach it, so a project cannot quietly balloon.
Against that fixed cost, weigh the daily tax you are already paying: the wasted hours, the per seat fees, the mistakes. For a lot of businesses the maths is clearer than they expect.
A useful napkin test: count the hours per week spent on the workaround, multiply by a conservative hourly cost, multiply by 52. If that number is in shouting distance of a scoped build, you are no longer “saving money” by staying on the sheet. Add the cost of the one bad copy-paste that already happened, or will.
We do not publish a fake “from €X” for software the way we do for websites, because a booking widget and a multi-yard operations tool are not the same job. Websites start at €2,500 because the shape repeats. Software does not. You get a number after we write down the must-haves. See pricing for how that conversation runs.
How to decide
Start by listing the manual workarounds in your week and the tools you are unhappy with. If the pattern is “the software almost works but,” you are a candidate. If your current tools genuinely fit, do not fix what is not broken.
Then sort the list into three buckets:
- A form and a few rules. That is often AI automation, not a new product. Instant replies, follow-ups, sheet updates. Read five tasks to automate first before you commission an app.
- A small web app. Quotes, bookings, a customer portal, a job board for the van. One audience, one job, a browser is enough. That is usually app development.
- A system the business would stall without. Stock, scheduling across people, permissions, history, reports. That is software, and it needs a harder scope.
Most people who write to us thinking they need 3 are in 1 or 2. We will say so. Building a platform to send a follow-up email is how small businesses waste a year.
When you should not build
Stay on the shelf, or on the spreadsheet, when:
- You cannot describe the process without saying “it depends” twelve times. Write the process down first.
- The pain is really marketing. No app will fill a calendar if nobody can find you. Fix the website and local SEO first.
- You want the software to invent a new business model. Custom software encodes how you already work. It is a bad place to discover the offer.
- You need it next week. Honest builds take discovery. Anyone who promises a complex system in seven days is selling a template with your logo.
Aviation, trades and multi-location operators often do need something custom, because the off-the-shelf tools were written for a different country or a different scale. Even then we start with the smallest version that removes the dangerous spreadsheet, not the full vision.
Automation, an app, or software
| You keep saying | You probably need | Why |
|---|---|---|
| ”I just want the form to hit the sheet and a reply to go out” | Automation | No new product, just a pipe |
| ”Customers need to book or see a quote without calling” | A small web app | One job, a browser, a login if needed |
| ”The sheet is the business and I am scared to touch it” | Software | History, permissions, the thing you cannot lose |
| ”Nobody can find us” | A website and SEO | Software will not fill an empty calendar |
If two rows feel true, do the higher one on the list first. Pipes before products. Products before platforms.
What a first version should include
A first version is allowed to be unglamorous.
- The two or three screens people touch every day
- Login for the people who need it, not a public playground
- The data you cannot afford to lose, backed up
- A way to export, so you are not locked to us
- No “phase 2” features hiding in the first invoice
If the first version needs a mobile native app, a customer community and an AI layer, the first version is not defined yet.
How we keep the price still
Discovery is written. Must-haves are listed. Nice-to-haves wait. The quote is fixed for that list. Change the list and we re-quote the change. You decide. That is the same discipline as our website work, applied to a fuzzier object.
You own what we write for you once it is paid, the same as a site. Third-party pieces keep their own licences. We are not interested in renting you a black box you cannot leave.
If you want the Dublin-shaped version of this conversation, app development in Dublin is the local page. The decision does not change with the county.
When you are ready to look properly, our software development starts by scoping the real problem and fixing a price, so you can decide with a clear number in front of you rather than a vague fear of cost.
Bring the sheet, or a screen recording of the workaround, to that first conversation. We can tell you in one sitting whether this is a weekend of automation, a small app, or a build you should not start until the process is written down. That honesty is the point of paying for the conversation.
Tired of making do? Start a project and we will scope the problem and give you a fixed price before you commit to anything.