Thinking long term - what kind of bespoke software are you building?

Our experience at SkillsLogic is that there are two broad categories of bespoke software, each with their own challenges.

Bespoke software for internal use because someone's decided that an off-the-shelf product can't do the job. This might be a bespoke web application that tackles an immediate business need but then grows over time as the business changes and different internal stakeholders request new features. The challenges with big internal bespoke systems seem to be:

1. How do you know when it makes more sense to go bespoke rather than configure something that's off-the-shelf?


2. How do you stop internal stakeholders with short-term 'urgent' requirements from adding features that have limited long-term value.


3. How do you avoid becoming completely dependant on bespoke software that comes without any long-term product roadmap?

A new software product or service that the you want to sell on to your customers. Often this is a software as a service (SaaS) solution. There are challenges here too:

1. You often need higher levels of usability - better user experience design. Your paying customers may well have higher expectations around look and feel.


2. You need long-term vision - a process to capture and prioritise new feature requests from your customers and a software architecture that will allow the system to grow without it becoming expensive and difficult to maintain.


3. Software products often need to be more configurable - you want to switch features on or off for different clients because you absolutely don't want to support different bespoke versions for different customers.

And there's a third category that can be a real problem - software that becomes a hybrid of the other two types. It starts out as an internal system, grows over time (in ways you didn't anticipate) and then at some point you want to sell it to external clients. And because you've got external customers who want some of your functionality but not all of it you start to think about retrospectively adding configurability into a system that's already very bespoke. The difficult choice then is between more development of the original system and starting again on a new system for external clients.

Thinking hard about the long-term at the start of the development is the answer. It also helps to think about where you don't want to be in 5 years (that seems to be the start of middle age for most web applications). Finally, take time to create a product road map - even for an internal bespoke system - that explains the broad goals of the system and sets out the limits of the scope of the bespoke development.