Why clean master data comes first
ERP projects often fail in the same quiet way. The software is installed, users are trained, and then the business discovers that 12 versions of the same supplier exist under different names, units of measure are mixed up, and stock codes have no clear structure. At that point, the problem is not the system. It is the data.
Item, customer and supplier records are the backbone of daily operations. If that information is inconsistent, every module that depends on it starts to slip. Purchasing orders go to the wrong supplier, stock counts do not match, delivery addresses are duplicated, and management reports become hard to trust. For SMEs in Johannesburg, Pretoria, Durban, Cape Town and Gqeberha, that can mean missed deliveries, poor cash flow control and extra admin that nobody has time for.
Clean master data means the business agrees on one version of the truth before migration begins. It is a practical step, not a technical luxury.
What usually goes wrong in small and medium businesses
Many South African businesses grow by adding people, branches and products faster than they tidy up their records. A supplier may be saved as one name in purchasing, another in accounts and a third in a spreadsheet. The same customer may appear as a company name, a trading name and a contact person. One item may be measured in boxes, another in units, while the warehouse uses metres or litres. That is where confusion starts.
Load-shedding can make matters worse when teams work offline and reconcile later. A rushed manual update during a power cut can create duplicates, miss mandatory fields or overwrite useful history. Coastal humidity in places like Durban and Gqeberha also affects stock control for goods with shelf-life concerns, so accurate records matter even more when batches, expiry dates or serial numbers are involved.
Typical signs of poor master data include the following:
- Duplicate customer and supplier records with small spelling differences
- Old contact details that no longer match current branches or delivery points
- Inconsistent item codes, descriptions and unit measures
- Missing tax numbers, payment terms or credit limits
- Stock items grouped differently across departments
How clean records improve ERP results from day one
When master data is cleaned before implementation, the ERP system can do its job properly. Users spend less time searching, checking and correcting. Managers get clearer reports because the same customer is not counted three times. Finance can reconcile invoices faster, and operations can trust stock levels when placing orders or planning production.
Good master data also supports better automation. If item names, supplier codes and customer categories are standardised, workflows can route approvals, generate purchase orders and trigger alerts without manual fixing. That is important for SMEs that need to control overheads and keep admin lean.
Businesses that prepare properly usually see these gains early:
- Cleaner migration into the new system
- Less duplicate entry across departments
- More accurate sales, stock and finance reporting
- Faster user adoption because staff trust the screens they see
- Fewer errors in customer communication and supplier invoicing
A practical example from a Centurion distributor
Consider a Centurion distributor serving sites in Pretoria, the East Rand and Durban. Before their ERP project, the same supplier had five records across purchasing and accounts. One branch used a local abbreviation, another used the full legal name, and one record had an outdated banking line. Items were saved with different descriptions depending on who entered them, and stock was counted in cartons in one area and each units in another. Reporting was slow and procurement kept chasing corrections.
Bizmatics worked with the team to map each record type, identify duplicates and agree on a clean set of rules. The business chose one naming standard for items, one supplier record per legal entity and one customer format for all branches. The result was not just tidier data. It gave the project a proper foundation. Training became easier, the go-live was calmer, and the finance team could trust the numbers coming out of the system.
The lesson is simple: if your records are unclear before the project, the ERP will carry that confusion forward. If the records are clean, the system has a much better chance of working as intended.
What a proper master data clean-up should include
A useful clean-up process goes beyond deleting duplicates. It should define rules that people can actually follow in day-to-day work. That includes coding structures, naming conventions, mandatory fields and ownership for future updates.
For South African SMEs, it also helps to align the records with VAT requirements, delivery areas, branch structures and credit control needs. If a supplier serves both Gauteng and the coast, the system should show that clearly. If a customer has multiple sites, each site should be linked correctly so deliveries and invoicing stay accurate.
Before an ERP project starts, these steps are worth doing:
- Review item, customer and supplier lists for duplicates and gaps
- Standardise codes, descriptions and units of measure
- Confirm legal names, tax details and contact fields
- Remove inactive records or mark them correctly
- Decide who may create or edit master data after go-live
Keeping the data clean after implementation
Master data clean-up should not be treated as a once-off exercise. If no one owns the rules after go-live, old habits return quickly. A new supplier gets created instead of checking for an existing record. A customer name is typed differently in sales and finance. A product description changes every time someone edits it. Within months, the same problems are back.
The best approach is to build simple controls into the process. Set approval steps for new records, train users on the naming standard, and review data quality regularly. Monthly checks are often enough for smaller teams. Larger groups with branches in Pretoria, Cape Town or the West Rand may need more frequent audits, especially where multiple staff members handle purchasing or customer onboarding.
At Bizmatics, the focus is always practical: clean inputs, clear rules and systems that support the way the business really works. That starts with master data that everyone can trust.
