Buying business software tends to start with a slightly unhelpful question:
Should we buy something off the shelf or have our own system built?
Sometimes the answer is custom software.
Quite often, it is SaaS.
Occasionally, the spreadsheet everybody complains about should stay exactly where it is for another year.
And there are businesses shopping for software when their current problem does not need software at all.
That last option does not get much space on development-company websites, for fairly obvious reasons.
It probably should.
Choose SaaS when the problem is common
SaaS, or Software as a Service, is software you subscribe to rather than commissioning for your business alone.
Think accounting systems, CRMs, project-management platforms, payroll products, appointment systems and email marketing tools. The provider builds one product that serves many customers, usually for a monthly or annual subscription.
SaaS makes a great deal of sense when your problem looks much like everybody else’s.
If you need accounting software, there are already mature products built around accounting. If you need project management, online bookings or email campaigns, starting a software-development project before looking at existing products would be a strange first move.
You can usually sign up quickly, start with a relatively low monthly cost and let the provider worry about hosting, updates and much of the technical maintenance.
The trade-off is that your business works within the product’s rules.
You choose from the workflows, fields, reports and integrations available. Sometimes there is plenty of flexibility. Sometimes your staff ends up changing the way they work because the software expects something different.
If those compromises are minor, SaaS is probably the sensible answer.
Consider custom software when the business itself is unusual
Custom software starts making more sense when the process you are trying to manage is specific to your company.
Perhaps pricing depends on rules particular to your industry. Several departments need to work through a workflow that no commercial system handles properly. You may have data scattered across older systems, or staff could be capturing the same information repeatedly because existing products do not talk to one another.
At that point, subscribing to more software can create another system instead of solving the original problem.
Custom software lets the system follow the business rather than asking the business to fit a standard product.
Custom Coding has built systems across legal, beauty, veterinary, laboratory and other specialist environments where the workflows are specific enough to justify that approach. Its current desktop application work includes new database systems, integrations, software takeovers and conversions from older desktop products to cloud systems.
The price is naturally higher upfront. You are paying for analysis, design, development, testing and implementation rather than subscribing to software whose development cost is spread across thousands of customers.
Custom therefore needs a better reason than “we would like our own system.”
Stay with spreadsheets when they still work
Excel does not need rescuing merely because custom software exists.
If one or two people maintain a manageable spreadsheet, the formulas are understood, the information is easy to find and mistakes are rare, replacing it may achieve very little.
A R60,000 software project that replaces a spreadsheet doing the job perfectly well is still R60,000 spent.
Spreadsheets become candidates for replacement when they start acting like infrastructure. Multiple people need the same data, version conflicts appear, formulas become dangerous to touch, reports require manual rebuilding or one employee holds most of the knowledge about how the file works.
We cover that progression in From Excel chaos to cloud business systems, including the move towards structured databases, user permissions and shared reporting.
Until those problems exist, Excel may be the cheapest and most practical system available.
Do absolutely nothing yet when the process is still changing
This is the option businesses tend to skip.
If you are still figuring out how the company should handle a process, automating it can simply make an unsettled process harder to change.
A new service launches and management immediately wants software for it. Three months later, the pricing model changes. Then responsibilities move between departments. Six months after that, customers are buying it differently from what everybody expected.
Software built during month one may faithfully preserve decisions nobody wants by month six.
There are cases where the better technology decision is to run the process manually for a while, learn where the real friction appears and only then decide whether software should be involved.
You may discover that an existing SaaS product handles it perfectly well.
You may discover that a small integration solves the problem.
You may even discover there was no software problem in the first place.
That is useful information to obtain before paying developers.
Cost should include what happens over several years
SaaS feels cheaper because the starting cost usually is cheaper.
R500 a month is easier to approve than a custom system costing tens of thousands upfront. Over several years, subscription fees accumulate, particularly when pricing is per user or several products are needed to run one workflow.
Custom software reverses the pattern. Development costs more at the beginning, followed by hosting, maintenance and future development where required.
Neither model automatically wins.
Custom Coding already has a detailed article comparing bespoke software and SaaS subscriptions for growing SMEs, which looks more closely at the cost question. The useful calculation should include subscription fees, implementation, staff time, integrations, support and the cost of repeatedly working around software that does not quite fit.
Existing software deserves a fifth option
There is another category worth mentioning: improve what you already have.
Older software is often treated as if there are only two choices. Keep suffering or replace everything.
Custom Coding’s work on legacy system modernisation takes a different approach. Existing data and useful processes can sometimes be retained while outdated sections are rebuilt or moved gradually.
A system does not become worthless simply because the interface looks old.
If it contains twenty years of business logic that still works, throwing all of that away can be the expensive option.
Start with the problem
SaaS is usually a strong choice when the requirement is common and the existing products fit comfortably.
Custom software becomes more compelling when the process creates a competitive advantage, has unusual rules or keeps being forced into products designed for somebody else’s business.
Spreadsheets deserve to stay while they remain manageable.
Waiting deserves serious consideration when you do not yet understand the process well enough to automate it.
Buying software should come after understanding what is wrong.
Otherwise you can spend a remarkable amount of money solving the wrong problem very efficiently.
