The short answer: buy a ready-made product where your business works like everyone else's, and build where your differentiation lies. Accounting, payroll and email are almost always best handled by off-the-shelf software. Processes that make you better than your competitors, or data that has to flow between systems no existing product connects, are where custom software can pay for itself.
In practice the answer is rarely either-or. Most companies end up with a mix: ready-made products for the standard parts of the business, and custom integrations or extensions where the need is specific. Below is a framework for making the decision on the right premises, with total cost, risk and your exit route in view.
When off-the-shelf is the right answer
If the process you are solving looks the same at your company as at thousands of others, a ready-made product is almost always the smarter buy. Development, security updates and new features are shared across every customer, at a price that reflects that scale. You also get something that is easy to underestimate: a system others have already tested in real operations, with documentation, a support desk and people on the job market who already know how to use it.
Off-the-shelf is especially attractive when time matters. A subscription can be live in days or weeks, while a custom build needs discovery, design, development and testing. If you don't yet know what the process should look like, it is often wiser to start with an existing product, learn from real use, and revisit custom work later on firmer ground.
The rule is simple: buy off-the-shelf where you are like everyone else.
The hidden costs of off-the-shelf
Ready-made systems aren't free just because the subscription looks low. The cost shows up elsewhere, and often only once the system is in daily use:
- Per-seat fees multiply with headcount and tend to rise year over year, with little room for you to negotiate.
- Workflows get bent to fit the system instead of the system supporting the workflow. That cost never appears on an invoice, but it shows up in staff time and frustration.
- Integrations are limited, and your data is either locked inside the vendor's platform or only accessible on a more expensive tier.
- You pay for features you never use because they come with the package.
- Manual work grows around the system: spreadsheet exports, double entry and data transfers the product can't handle on its own.
None of this makes off-the-shelf the wrong choice. It means the comparison has to cover the full cost, not just the monthly fee on the vendor's pricing page.
When custom pays off
Custom software earns its place when the process is part of what makes your business different. Then a system that fits exactly isn't a luxury, it's an advantage. If you reshape your differentiator to fit a generic product, you are effectively making your business more like everyone else's.
Custom work is also worth considering when no ready-made product talks to your other systems, when your rules, data or calculations are too specialised for general software, or when several years of per-seat fees would exceed the cost of building and running your own. A third reason is control: you decide the priorities, the release pace and where the data is hosted.
Custom software has real drawbacks too, and they deserve to be said out loud. It takes longer up front, someone has to own operations and updates, and vague requirements make a project more expensive than it needs to be. A custom system nobody is responsible for ages badly.
Compare total cost over five years
The most useful comparison is total cost over five years. For off-the-shelf, that is users times the per-seat fee times 60 months, plus implementation, configuration, integrations and likely price increases. For custom, it is the build itself plus operations, hosting and maintenance for the whole period.
For illustration only, not a quote: a focused custom system that takes 120 to 200 hours to build comes to roughly ISK 2.1 to 3.6 million at our hourly rate of ISK 17,880 excluding VAT. Our System operations plan costs 59,900 ISK per month, or about ISK 3.6 million over five years, with maintenance and new features on top as needed. Actual scope depends on users, integrations, data and risk, and only becomes clear in discovery.
In many cases the ready-made product wins this comparison, and then that is the answer. But the question isn't only which option is cheaper, it is what you get for the money. A system that saves your staff time every day can pay for itself even if it costs more on paper.
The questions that settle it
- Is this process the same for us as for other companies? If so, a ready-made product is probably the right route.
- Is the process part of what makes us better than our competitors? If so, a custom build or a custom extension is worth exploring.
- What does the ready-made option cost over five years, including every seat, implementation and integrations?
- Can our data leave the system, in a usable format, if we want to switch later?
- Does the product offer an API so it can be connected to accounting, the website and other systems?
- Who is responsible for operations, security updates and backups once the system is live?
- If we build, who owns the source code, and could another provider take over the system if needed?
The hybrid route is usually where you land
The most common, and often the most sensible, outcome is to keep ready-made products where they work and build only what is missing between them. That might be an integration that moves web orders into accounting, a customer portal that pulls data from a system you already own, or a specialised calculation no generic product offers.
The benefit is lower risk and a more focused investment. You don't pay to rebuild what already exists, and the custom budget goes where it creates real value. The trade-off is that more systems mean more connections to monitor, so running those integrations should be part of the plan from day one.
Common questions
Is custom software always more expensive than off-the-shelf? Not necessarily. It usually costs more up front, but over several years per-seat fees, configuration and manual work around a ready-made system can add up to more. It depends on the number of users and how well the product actually fits.
Can we start with an off-the-shelf product and build later? Yes, and it is often a good approach. Just choose a product with an API and proper data export, so the door stays open.
What if we are already stuck with a system that fits poorly? Start by mapping what works, what is missing and whether your data can be extracted. Sometimes an integration or extension is enough; sometimes switching is the wiser move.
How to take the first step
Start with a simple list: which processes are standard, which are specific to you, which systems are already in use, and where staff do manual work around them. Then estimate the five-year cost of each option. That list alone often settles the question.
If you are weighing these options, an independent assessment before you sign anything is usually worth it, especially of total cost and integration capability. At Novamedia, our Discovery and planning engagement is designed for exactly this: a fixed price of ISK 214,560 excluding VAT for 12 hours, with prioritised recommendations, an indicative scope and a cost estimate. The conclusion may well be that a ready-made product is enough, and if one of our own products solves the problem, we will recommend it rather than propose a custom build.