Custom software: bespoke applications compared with off-the-shelf software

Share

Glossary

What is custom software?

Custom software is software built for one particular company and the way it works. This article covers the definition, the difference from off-the-shelf software, typical examples, costs and how such projects run.

Custom software in brief

Custom software is software developed for a single company and its specific processes. It is not sold as a product but built for one client unlike off-the-shelf software, which an entire market uses in much the same way.

The terms bespoke software, tailor-made software and custom application development mean the same thing. In each case the feature set follows one client rather than the average of all possible customers.

The practical difference shows in the direction of adaptation. With off-the-shelf software the company adapts its processes to the product; with custom software the product is adapted to the processes. Either can be right what matters is whether the process in question is part of the business model or simply has to be administered.

Definition and classification

By common definition, custom software is a program created on behalf of a user for that user's specific requirements. Usage rights and normally the source code belong to the client, and the feature set comes out of their requirements list.

Between purely custom and purely off-the-shelf software lies a broad middle ground that accounts for most projects in practice:

Most companies run all four forms at once. The useful question is therefore not "standard or custom?" but "which process belongs in which category?"

  • Off-the-shelf softwareA finished product for many users, such as accounting or office programs. Adapted through settings.
  • Configured off-the-shelf softwareA standard product brought closer to the company through customising and add-on modules typical for ERP systems.
  • Extended off-the-shelf softwareA standard product with individually developed modules or interfaces, such as a shop with its own pricing logic.
  • Custom softwareAn application developed entirely for the client, often connected to existing standard systems.

Custom software vs. off-the-shelf software

Off-the-shelf software is cheaper, available immediately and proven in the field; custom software fits more precisely and does not tie you to a vendor. The comparison turns less on price than on how much idiosyncrasy the process in question can tolerate.

Off-the-shelf software

A product many companies use in much the same way.

  • Low entry cost, ongoing licence or subscription fees
  • Usable at once, tested by many users
  • The vendor handles updates and security
  • Adaptation limited to available settings
  • Dependence on the vendor's roadmap and pricing
  • Your own processes have to fit in

Custom software

An application developed for one business.

  • Higher initial investment, no per-user licence cost
  • Development time before first use
  • Maintenance and development sit with the client or their provider
  • Processes are modelled as they actually are
  • Source code and data stay in house
  • The feature set stays limited to what is needed

A frequent error is comparing purchase prices. Off-the-shelf software costs per user per month, custom software costs once plus maintenance with many users over several years, that ratio often reverses.

Advantages and drawbacks

The main advantage of custom software is precise fit: no unused features and no workarounds for processes the product never anticipated. The main drawback is that the effort and responsibility for maintenance and further development remain with the client.

The advantages that matter in practice:

Against this stand real drawbacks: a higher initial investment, development time before any benefit, dependence on the quality of the documentation, and the risk that unclear requirements produce a system nobody uses. The first three can be planned for; the fourth is only avoided by gathering requirements properly.

  • No bending of processesThe workflow is modelled as it functions, including its exceptions.
  • No per-user licencesCost does not automatically rise with headcount.
  • Independence from a vendorSource code and data stay in house; a price rise or discontinued product does not hit the core.
  • Integration instead of an islandThe application is connected to existing systems from the start.
  • A scoped feature setWhat is not needed does not get built, which keeps the software easy to operate.

When custom software is worth it

Custom software is worth it when a process is part of the business model or only works in standard software through detours. As a rule of thumb: the more a process distinguishes a company from its competitors, the less a standard product suits it.

Clear signs that building something pays off:

Conversely, off-the-shelf software is almost always the better choice for accounting, payroll and office work: every company needs the same thing there, and mature products are hard to beat on price. If you are facing that decision, our custom software development page sets out how we assess a specific case; for a neutral review with no delivery attached there is software consulting.

  • Spreadsheets as a production systemCore processes run on spreadsheets nobody fully understands any more, with versions in several inboxes.
  • Duplicate entryThe same data is typed into two or three systems by hand.
  • Workarounds as the normal routeThe established path through the standard software consists of exceptions only insiders know.
  • Licence cost without benefitYou pay for a feature set of which a small part is actually used.
  • The process is the advantageThe workflow is the reason customers come and that is exactly what no product covers.

Cost, duration and approach

A scoped internal application usually starts in the mid five figures, while larger systems with several interfaces run into six figures. The range is this wide because the effort depends less on the interface than on the number of connected systems and the quality of the existing data.

Three factors drive cost more than anything else: the number of interfaces, the volume and cleanliness of the legacy data to migrate, and the number of exceptions in the process. An estimate made before requirements are gathered is therefore worth little.

The usual approach runs in four stages: gathering requirements, architecture and data model, development in short stages with working intermediate builds, then operation and further development. A usable core should go live early an application first used after twelve months has not been tested until then. How this runs with us is described under custom software development.

Typical technologies and architectures

Custom software is mostly built as a web application today, because it then runs on any device without installation and can be updated centrally. Common choices on the back end are Node.js, PHP with Laravel, Python and Java together with relational databases such as PostgreSQL or MySQL; on the front end React, Next.js, Vue and Angular with TypeScript.

The connection to existing systems runs through an API, usually as a REST or GraphQL interface. This is where it is decided whether an application stays an island or becomes part of the system landscape which is why interfaces need logging, retry logic and monitoring, not just a connection.

On architecture: a well-structured monolith is the more sustainable choice for most business applications, because it creates less operational overhead than a distributed solution. Microservices only pay off once parts genuinely need to scale independently or be run by separate teams. If an internal application turns into a digital business model with external users, it moves towards a platform.

Frequently asked questions about custom software