One of the fastest ways to make an ERP project expensive is to treat every current process as a requirement.
Businesses develop workarounds for understandable reasons. A system could not handle something, so a spreadsheet was added. A manager needed control, so an extra approval appeared. A report was unclear, so someone created a manual check. Over time, the workaround becomes familiar and is described as the way the business must operate.
If all of that is copied into a new platform, the business pays to preserve the limitations of the old one.
Best practice is a starting point, not a refusal to listen
Established workflows exist because many organisations have already solved common problems in sales, purchasing, inventory, manufacturing, finance, projects and customer service.
A capable consultant should understand the current process, identify the real control or outcome behind it and then show how the platform handles that requirement using standard practice. The aim is not to force every business into an identical model. It is to avoid inventing a special process where no genuine difference exists.
Ask why before asking how
When someone asks for a custom field, approval or report, useful questions are:
- What decision will this support?
- What risk is the control reducing?
- Who uses the information and what do they do next?
- Is the step required by the business, a customer, regulation or the old software?
- Could a standard workflow produce the same outcome more clearly?
These questions separate genuine differentiation from inherited complexity.
Customisation is sometimes the correct answer
Some businesses have processes that create real competitive advantage or operate in sectors with specialist requirements. Commercial art, complex manufacturing, connected devices and unusual commercial models can all justify focused development.
The difference is that customisation should be deliberate. It should have an owner, a business case, acceptance criteria, documentation and a clear understanding of how it will be maintained through future upgrades.
Standardise before automating
Automation and AI can accelerate a good process. They can also accelerate inconsistency. If ownership, data and exceptions are unclear, adding automation may make the problem harder to see.
First establish the workflow. Make the responsibilities visible. Let users operate it and discover the genuine exceptions. Then automate the repetitive parts with evidence that the process works.
Good implementation requires judgement
A software demonstration can show what a feature does. It cannot decide what the business should do. That requires someone to understand the organisation, challenge assumptions and take responsibility for the wider design.
Cloudsaber uses Discovery, gap analysis, phased implementation and user testing to align the platform with best practice while protecting the requirements that genuinely matter. See how Cloudsaber works or explore our software development capability for requirements that need a focused extension.