ERP Selection for Retailers and Wholesalers

Look at whether your problems are structural or operational. Structural signs include core commercial logic living outside your systems, such as contract pricing maintained in spreadsheets, multiple entities requiring manual consolidation, or an inability to cost inventory properly because the accounting package cannot represent your stock model. Those generally need a different class of system. Operational signs such as inaccurate stock, rekeying between systems and slow reporting frequently have cheaper causes: counting discipline, missing integration, or reporting tools. Establish which category your specific problems fall into before beginning a selection, because buying an ERP to fix an operational problem often leaves the problem intact and the budget spent.

We avoid quoting ranges because they mislead more than they help. The licence or subscription is usually the smaller component, while implementation, configuration, data migration, integration, training and internal time typically exceed it, sometimes substantially. The main drivers are how much customisation you require, how clean your data is, how many integrations are in scope, how many entities and locations you operate, and how much the implementation partner charges for a business of your complexity. Model your own figure over five years including ongoing support and integration maintenance, obtain quotes against the same documented requirement, and add contingency, because implementation estimates at this scale are optimistic more often than not.

Because whatever you choose will have to exchange data with a commerce platform, a point of sale, freight providers, marketplaces and reporting tools, and the cost and feasibility of that varies enormously between products. Common late discoveries include an interface that does not expose the objects or fields you need, rate limits that are fine at average volume and problematic at peak, functionality available only on a higher licence tier, or sparse documentation that turns a routine build into an expensive one. All of these are answerable during evaluation. Ask the specific questions, request documentation, and speak to a business already integrating that product at a similar scale.

GST treatment including mixed, exempt and input taxed supplies, and how tax codes map to the reporting you need. BAS preparation support and whether it produces what your accountant expects. Handling of the Australian financial year and end of year processes. Payroll if in scope, including superannuation and single touch payroll reporting, though many businesses keep payroll separate. Local support availability in Australian hours, since offshore support at your peak trading is a genuine operational risk. And where data is held, together with what your obligations under the Privacy Act 1988 and the Australian Privacy Principles require in respect of personal information.

In our experience more important. Most mid market products used in Australian retail and wholesale are capable of running a business like yours, and the difference between a good outcome and a poor one is usually the configuration, the process decisions and the quality of the people doing the work. Evaluate partners on implementations completed for businesses of similar size and model, insist on knowing which named consultants will actually be assigned and what else they are committed to, and take references you contact yourself. Ask those references directly what went badly and what they would do differently, because every implementation has a difficult chapter and the useful information is in how it was handled.

It depends on how coupled your processes are and how much risk you can absorb. A staged approach, such as finance first then inventory then sales, reduces the size of each cutover and lets the team build familiarity, but it means running parallel processes and temporary integration between old and new for a period, which carries its own cost and confusion. A single cutover is cleaner conceptually and much less forgiving if anything goes wrong. For most mid sized retailers and wholesalers, a small number of well defined stages works better than either extreme, and the timing of each should avoid peak trading and year end regardless of the approach chosen.

Selection is not irreversible, but it is expensive to reverse, which is the argument for doing it carefully rather than for agonising indefinitely. Two things materially limit the damage. First, keep customisation restrained, because a heavily customised system is far harder to move away from and harder to upgrade. Second, ensure you own and can export your data, know where configuration is documented, and understand what a migration away would involve, in the same way you would with any significant system dependency. Beyond that, most genuine mismatches emerge during scripted demonstrations with your own scenarios, which is exactly why that step is worth the effort.