You likely need a new ERP, or your first one, when the business depends on spreadsheets and re-keying between systems, month-end takes longer each quarter, nobody trusts stock or margin figures, and the current system cannot support new entities, sites or processes without workarounds. Any one sign is a reason to investigate; several together usually mean the cost of staying is rising.
A new ERP is a large commitment, and not every frustration justifies one. Sometimes the right answer is an upgrade, better configuration or a small integration. This guide helps you separate the two: ten signs, a way to measure what staying costs, and a decision table for the options.
What are the signs a business has outgrown its ERP?
The signs fall into three groups: data you cannot trust, work that does not scale, and a system that cannot change. Score each one honestly; the scorecard further down turns the list into a decision.
Data you cannot trust
1. Stock figures are wrong often enough that people check the shelf. If buyers or sales staff routinely confirm stock physically before promising it, the system has stopped being the record. Causes include movements not recorded, no location or batch tracking, or integrations that fail silently.
2. Different reports give different answers. Sales says one revenue figure, finance another, and the difference takes days to explain. This usually means data lives in several places with no single source.
3. You do not know true margin by product, customer or job. Costs such as freight, labour or rebates are allocated in spreadsheets after the fact, so pricing decisions rely on estimates. Our guide to the ERP reports finance teams need lists what a sound system should produce directly.
Work that does not scale
4. Spreadsheets sit between every system. Orders exported from the web shop, re-keyed into accounting, copied into a production plan. Each hand-off adds delay and error, and depends on one person who knows the macros.
5. Month-end takes longer as you grow. Manual reconciliations, inter-company journals typed by hand, and accruals rebuilt each month. A close that grows with volume rather than staying stable is a sign the system is not doing the work.
6. Adding a branch, entity or warehouse is a project. If opening a second site or acquiring a company means a new copy of the system, or manual consolidation in spreadsheets, the system was built for a smaller business. See multi-entity ERP governance for what group-ready looks like.
7. Approvals happen in email and chat. Purchase approvals, discounts and credit overrides are agreed outside the system, so there is no audit trail and no limit enforcement.
A system that cannot change
8. The version is unsupported, or close to it. Vendor support ends, security updates stop, and tax or regulatory updates may not be delivered. Check the vendor's lifecycle page for your exact version.
9. Customisations block every upgrade. Years of modifications mean each upgrade is a re-implementation, so the business stays on an old version. Each modification is one more thing to retest and rewrite at upgrade time, and the debt compounds.
10. Only one person, or one outside contractor, understands it. When the knowledge of how the system works sits with one individual, every change is slow and the business carries a key-person risk.
If several of these apply and your processes are distinctive, our custom ERP software service builds an ERP around your workflows, module by module.
How do you score the signs?
Score each sign from 0 to 2: 0 if it does not apply, 1 if it causes occasional pain, 2 if it is a weekly cost or a real risk. The total is not a formula; it is a way to make the conversation concrete with the leadership team.
| Sign | Score 0 to 2 | Evidence you can gather |
|---|---|---|
| 1. Stock not trusted | Count adjustments per month; stock-outs on items the system showed in stock | |
| 2. Reports disagree | Time spent reconciling sales and finance figures | |
| 3. Margin unknown | Number of costs allocated manually | |
| 4. Spreadsheets between systems | Count of re-keying steps; hours per week | |
| 5. Slow month-end | Days to close, trend over four quarters | |
| 6. New sites are projects | Effort for the last expansion | |
| 7. Approvals outside the system | Share of spend approved by email | |
| 8. Unsupported version | Vendor end-of-support date for your version | |
| 9. Customisations block upgrades | Versions behind current; last upgrade effort | |
| 10. Key-person dependency | People who can change the system |
A low total with one critical item, such as an unsupported version, can still justify action. A high total spread across many items usually points to replacement rather than patching.
What does staying on the current system cost?
The cost of staying is the time, risk and lost opportunity of keeping things as they are. It is real even though it never appears as an invoice.
Cost-of-staying worksheet
- Manual hours: re-keying, reconciliation, report building. Multiply hours per week by the loaded cost of the staff involved.
- Errors: credit notes issued for wrong invoices, stock write-offs, missed supplier discounts.
- Close time: days finance spends closing that could go to analysis.
- Growth friction: a new channel, site or entity delayed because the system cannot support it.
- Risk: unsupported software, security exposure, a key person leaving, an audit finding.
- Workaround systems: licences and upkeep for side tools that exist only to fill gaps.
Use your own figures. The point is to compare the cost of staying with the total cost of owning a new ERP, not to justify a decision already made.
What are the options besides a full replacement?
There are four realistic paths: upgrade, extend, replace with a package, or build a custom ERP. Each fits a different pattern of signs.
| Option | Fits when | Watch out for |
|---|---|---|
| Upgrade the current system | Core processes fit; the main issue is an old version | Customisations may need rework; upgrade may not fix process gaps |
| Extend with integrations or add-on modules | The core works, but data moves by hand between systems | Too many add-ons create a new patchwork |
| Replace with a packaged ERP | Processes are mostly standard; you want vendor-maintained features | Configuration limits; heavy customisation repeats the old problem |
| Build a custom ERP | Processes are distinctive, or packages need heavy change to fit | Needs clear requirements, a capable partner and ongoing ownership |
For a general comparison of buying and building, read ERP versus custom development. For an old system specifically, our legacy ERP modernization guide compares upgrading, extending and replacing in more depth.
Who should be involved in the decision?
The decision belongs to the leadership team, but the evidence comes from the people who live with the system every day. A replacement decided only by IT, or only by finance, tends to fix one department's problems and create new ones elsewhere.
- Finance: close time, manual journals, reconciliation effort, audit findings
- Operations and warehouse: stock accuracy, picking errors, production planning gaps
- Sales and customer service: order entry time, stock and delivery promises, pricing exceptions
- Purchasing: off-system orders, approval delays, supplier invoice mismatches
- IT or the system owner: support dates, security patches, integration failures, key-person risk
Ask each group for two or three concrete examples from the last quarter, with dates and rough hours. Examples carry more weight in a board discussion than general complaints, and they become the first draft of your requirements if you do decide to replace.
Questions to answer before choosing a route
- Which three signs cost the business most, in hours or risk?
- Would an upgrade or better configuration fix them, or are they caused by how the system is designed?
- What changes are coming in the next two to three years: new entities, channels, sites or regulations?
- Who will own the system after the project, and do they have the time?
- What is the latest date by which something must change, for example a vendor end-of-support date?
What if you do not have an ERP yet?
If you run on accounting software plus spreadsheets, the same signs apply, but the question is whether you need an ERP at all. Many small businesses do not, yet. Our guide on whether you need an ERP helps you decide. The usual trigger is when stock, purchasing, production or multiple entities have to share data with finance every day.
Illustrative scenario: two businesses, two answers
These are illustrative examples, not client cases.
A wholesaler scores high on signs 4, 5 and 7, but its system is current and its processes are standard. The answer is to extend: integrate the web shop and add an approval workflow, then review again in a year.
A manufacturer scores high on signs 1, 3, 6, 8 and 9. It runs an old version with heavy customisation, has just acquired a second company, and costs jobs in spreadsheets. Patching would not address the underlying fit. The answer is replacement, and because its job costing is distinctive, it compares packaged options with a custom build before deciding.
When are these signs misleading?
Some symptoms come from process or training, not the software. If stock is wrong because receipts are not recorded on time, a new ERP will be wrong too, unless the process changes. Before replacing, check whether the current system is configured and used properly. An honest review by someone outside the team can save a replacement project.
A new ERP also brings its own disruption: migration, training and a period of lower productivity. If the business is in the middle of a major change, such as an acquisition or a peak season, the timing may matter as much as the decision.
How to start
Complete the scorecard and the cost-of-staying worksheet with your finance and operations leads. Then write down the processes that matter most using our software requirements brief template. If you plan to compare vendors, turn the same brief into a request for proposal so every supplier answers the same questions.
Timeline Digital builds 2 to 3 key modules as a free pilot before the full project starts, so you can test whether a custom approach fixes your highest-scoring signs before committing. Discuss your project with us.