Measure the symptoms before naming the cause
SCSB is often approached by Malaysian finance teams who suspect their accounting software is holding the business back but cannot yet evidence it. That gap matters. "The system is too small" is a conclusion, not a finding, and a replacement project built on it tends to reproduce the original problem in a more expensive platform.
The six symptoms below are deliberately measurable. Work through them over one full reporting cycle and record numbers rather than impressions: hours spent, error counts, days elapsed, transactions affected. At the end you will either have a defensible case for change or a clear view that the constraint is process rather than software. Both outcomes are useful.
1. Consolidation is assembled by hand each month
Where a group operates several companies, the test is not whether consolidation is difficult but how it is produced. Record who exports each entity's trial balance, how balances are converted and mapped when charts of accounts differ, how intercompany balances are identified and eliminated, and how long the assembled pack takes to reproduce when a late adjustment arrives.
A monthly consolidation rebuilt manually in a spreadsheet carries a recurring cost and a version-control risk that grows with each additional entity. It is worth noting that consolidation software reduces the assembly effort rather than removing the judgement: in Dynamics 365 Business Central, for example, Microsoft documents eliminations as a manual process posted through the general journal. The gain is in structure and repeatability, not in automation of the accounting decisions.
2. Spreadsheets have become part of the system of record
Every finance function uses spreadsheets, and that is not a symptom by itself. The symptom is a spreadsheet that holds data existing nowhere else, that another process depends on, and that only one person can maintain.
- Stock valuations, landed-cost calculations or work-in-progress schedules maintained outside the ledger.
- Customer or supplier terms, pricing and discount matrices kept in a workbook rather than on the master record.
- Fixed asset registers, accruals or prepayment schedules rekeyed each month.
- Departmental or project reporting rebuilt from an export because the ledger cannot produce the required split.
For each one, record who maintains it, what happens when that person is unavailable, and whether a reviewer can trace a reported figure back to a posted transaction. Where the answer to the last question is no, the reporting risk is already material.
3. Month-end is getting longer, not shorter
Track the elapsed days from period close to management pack, and the split between waiting for data, correcting data and analysing it. A close that lengthens as volume grows, while headcount and process stay constant, indicates work that scales with transactions rather than with complexity — the pattern most often caused by system limitations. Four figures are worth recording for a single cycle:
- Days from period close to a management pack that finance is prepared to sign.
- Hours spent reconciling subledgers to the general ledger, recorded per person.
- Adjustments raised after the first close, separated into data corrections and genuine accounting judgements.
- The proportion of the cycle that repeats identically each month and carries no decision content.
The last figure is the most revealing. Repetitive work with no decision content is exactly what a system should absorb, and its volume is a reasonable proxy for what a better-structured platform would return in capacity.
Then separate the causes honestly. A close delayed because supplier invoices arrive late is a procurement discipline problem, not a software problem. A close delayed because three people must reconcile subledgers to the ledger by hand every month is a structural one. Only the second category is solved by a platform change.
4. Inventory records are no longer trusted
The practical measure is whether the business acts on its own stock figures. Record the variance at the last two counts, how often a sales order is confirmed against stock that turns out to be unavailable, whether stock is held at more than one location and can be seen separately, and whether costing is understood well enough that a margin figure can be defended.
Where staff routinely check physically before committing to a customer, the record has stopped functioning as a control. That is a strong indicator, particularly for businesses moving towards assembly or production, where Microsoft documents assembly, manufacturing and warehouse management as areas requiring the Premium experience in Business Central rather than a standard entry-level configuration.
5. Approvals live in email and chat
Approval workflow is where growth is felt earliest. Record how a purchase order above a threshold is approved today, how a credit note or manual journal is authorised, and how someone reviewing the transaction six months later would evidence that approval.
Where the answer is an email thread, a chat message or an initial on a printout, the control exists in practice but not in the record. That may be acceptable at ten transactions a month. It becomes difficult to sustain when volumes rise, when responsibilities are delegated, or when an external reviewer expects the approval trail to sit alongside the transaction.
6. The audit trail and access model no longer reflect the organisation
Two related checks. First, can the system show who created, amended, approved and posted a given transaction, and can a posted entry be altered without leaving a record? Second, do user permissions reflect current job responsibilities, or have they accumulated as people changed roles?
Shared logins, a single administrative account used by several staff, and permissions that allow the same person to raise, approve and pay are common findings in growing businesses. They are also the findings most likely to be raised by an external reviewer. Note that stronger access controls in any platform support segregation of duties; they do not by themselves establish it, because that outcome depends on how roles are assigned and reviewed.
Reading the result — including when the answer is no
Score the six areas and count how many are structural rather than behavioural. One or two isolated symptoms usually point to a fixable process, a data-quality exercise or a targeted add-on. Four or more, sustained across a full cycle and supported by recorded numbers, indicate that the operating model has moved beyond what the current application was designed to carry.
Some conclusions should survive commercial pressure. If the evidence shows late supplier documents, unclear ownership of master data or an undocumented month-end process, a new platform will not resolve any of them and may make them more visible and more expensive. Fixing the process first is frequently the correct answer, and it also produces the requirement definition any future implementation would need.
Next step
Where the assessment produces evidence of a structural constraint rather than a process one, SCSB can review the findings, the data condition and the scope that a replacement would involve. SCSB's Business Central implementation service describes how that scope is defined; the decision on whether to proceed remains the organisation's own, based on its evidence and priorities.
This article is general information about assessing finance systems. It is not accounting, tax, legal, audit or regulatory advice, and it does not evaluate any particular organisation's systems, controls or obligations. Verify product capabilities and Malaysian requirements against current vendor documentation and the relevant authority.
Dynamics 365 Business Central and Microsoft are trademarks of the Microsoft group of companies.