Enterprise resource planning (ERP) has a reputation as expensive software for multinationals. The name does not help. “Enterprise” sounds like “large corporation”, and “resource planning” sounds abstract. In plain terms, “enterprise” just means a business, and “resource planning” means planning and controlling the resources the business runs on: people, materials, machines, money, customers and time.
An ERP system brings the main records of a business into one connected system: sales orders, purchasing, inventory, production, finance and often customer and employee information. Instead of the same order being typed into a spreadsheet, an accounting package and a production board, it is entered once and flows through. The promise is consistency, accuracy, real-time information and less manual work.
Many small and mid-sized manufacturers can benefit. Many also have stories of ERP projects that ran over budget, frustrated staff and ended up used as an expensive invoicing tool. This article explains what an ERP does, how to tell whether you need one, how to choose a system starting from your business problems, and how to implement it without the common failures.
What an ERP system typically covers
Systems differ, but most include some combination of these modules:
| Area | What it handles |
|---|---|
| Finance and accounting | General ledger, accounts payable and receivable, tax, reporting |
| Sales | Quotes, orders, pricing, customer accounts, invoicing |
| Purchasing | Suppliers, purchase orders, receipts, supplier invoices |
| Inventory | Stock levels, locations, movements, stocktakes, valuation |
| Production and planning | Bills of materials, routings, work orders, capacity, material requirements planning |
| Quality | Inspections, non-conformances, traceability, certificates |
| Customer relationship management | Leads, opportunities, contact history, service |
| People | Timesheets, labour costing, sometimes payroll and HR |
| Reporting and dashboards | Real-time views of sales, margin, stock, work in progress and cash |
For manufacturers, the planning functions matter most. Material requirements planning (MRP) takes demand (orders and forecasts), the bill of materials for each product and the current stock and lead times, and works out what needs to be bought or made, and when. Done manually, this is where many small manufacturers lose money, either through shortages that stop production or excess stock that ties up cash.
The business case: what problems ERP solves
ERP is worth considering when the cost of the problems it solves outweighs the cost and disruption of implementing it. Typical problems include:
Matching supply with demand. If you cannot see demand clearly, you cannot plan materials, production, staffing or cash. Manufacturers of fast-moving products, such as food and beverage, consumer goods and spare parts, live or die on this. A system that combines order history, forecasts, stock and lead times lets you plan for the next three, six or twelve months. It also informs how much working capital, borrowing and staff you will need.
Too much manual re-entry. If the same information is typed into three systems, errors are guaranteed and staff time is wasted.
No single version of the truth. When sales, production and finance each have their own numbers, meetings become arguments about whose spreadsheet is right.
Slow decisions. If monthly results take weeks to compile, problems are found too late. Real-time reporting shortens the time between a problem appearing and someone acting on it.
Weak traceability. Some customers and industries require you to trace materials and batches from supplier to finished product. Paper systems make that slow and unreliable.
Inaccurate costing. Without accurate labour, material and overhead data per job, it is hard to know which products and customers actually make money.
Growth limits. Processes that work for ten orders a week can collapse at fifty. Systems allow volume to grow without a matching growth in administration.
Large organisations use ERP for similar reasons at a larger scale. A multinational with tens of thousands of employees may use it to understand its workforce, link appraisals to defined responsibilities, and cut decision cycles from months to weeks. A beverage manufacturer may use it to match production with demand. A high-volume online service may use connected systems to capture customer signals in real time and launch new offers faster. The scale differs, but the logic is the same as in a small factory: better information, sooner, with less manual handling.
Do you need an ERP yet?
Not every small business needs a full ERP. An accounting package with a few add-ons, such as inventory, job costing or a CRM, can serve a small operation well for years. Signs that you are outgrowing that setup include:
- Several disconnected systems and spreadsheets for orders, stock and production.
- Frequent stock-outs or excess stock despite effort.
- Regular errors from re-keying information.
- Inability to answer basic questions quickly, such as “what is our margin on this product?” or “what is in work in progress?”
- Customer or regulatory requirements for traceability you struggle to meet.
- Plans for significant growth, another site or more complex products.
If only one or two of these apply, a targeted add-on may be enough. If many apply, it is time to look at an integrated system.
Start with problems, not software
The most important principle in ERP selection is that no single system is right for every business. Systems are built with particular industries and problems in mind. Some are strong in discrete manufacturing (making distinct items such as machines and fabrications), some in process manufacturing (food, chemicals, recipes and batches), some in distribution, projects, services or retail. A system that excels at warehouse logistics may be weak at job-shop scheduling.
So start by writing down your problems and requirements, not by watching demonstrations. A good requirements document includes:
- Your main business processes, from quote to cash, procure to pay and plan to produce, described briefly.
- Your top ten problems, in order of cost or pain.
- Must-have requirements, such as multi-level bills of materials, batch traceability, job costing, integration with your accounting or e-commerce platform, or multiple sites.
- Nice-to-have requirements.
- Volumes: numbers of users, orders, products, suppliers and transactions.
- Constraints: budget, timeline, internal IT capability and any existing systems to keep.
Share this with vendors or implementation partners and ask them to show how their system handles your scenarios with your examples, not their standard demonstration.
Cloud or on-premise
Most small-business systems today are offered as cloud services, with a monthly subscription per user. The vendor hosts the software and infrastructure, and you access it through a browser or app. Advantages include low upfront cost, no server hardware, automatic updates, remote access and predictable costs. Disadvantages include ongoing subscription fees, reliance on internet connectivity, and less control over upgrade timing and data location.
On-premise systems run on your own servers. They need more upfront investment and in-house IT support, but they can offer more control and customisation. For most small manufacturers, cloud is now the practical default unless there is a specific reason otherwise, such as unreliable connectivity or particular data requirements.
Understanding the real cost
The subscription or licence fee is often the smallest part of the cost. Budget for:
- Implementation services: configuration, process design and project management by the vendor or partner.
- Data migration: cleaning and loading customers, suppliers, items, bills of materials and opening balances.
- Customisation and integration: links to other systems and changes to standard functionality.
- Training: for every user, not just super-users.
- Internal time: your staff will spend significant time on design, testing and training, often while doing their normal jobs.
- Ongoing support and improvement.
A realistic budget and timeline, with a contingency, prevents panic decisions halfway through.
Implementing without the classic failures
ERP projects fail for predictable reasons. Most are about people and process, not technology.
1. Fix the process before automating it. Automating a messy process produces a faster mess. Map and simplify your key processes first. Standard operating procedures written before implementation make configuration and training far easier.
2. Clean your data. Duplicate customers, inconsistent part numbers, wrong units of measure and inaccurate bills of materials will undermine the new system from day one. Allow time to clean data before migration.
3. Appoint an internal owner. Someone inside the business must own the project and have the authority to make decisions. Without an internal champion, the project drifts and the vendor fills the gap with their assumptions.
4. Prefer configuration to customisation. Heavy customisation makes upgrades expensive and locks you into a particular partner. Where the standard process is reasonable, adapt your process to the system.
5. Phase the rollout. For a small business, a phased approach, such as finance and sales first, then inventory and purchasing, then production planning, reduces risk compared with switching everything on in one weekend.
6. Test with real scenarios. Run real orders, jobs and month-end processes through the system before go-live. Involve the people who will use it.
7. Train properly and support after go-live. Productivity usually dips for a few weeks after go-live. Plan for extra support, quick fixes and refresher training.
8. Measure the benefit. Define before you start what success looks like, such as stock accuracy, on-time delivery, time to close the month or quote turnaround, and measure it afterwards.
Questions to ask vendors and references
Demonstrations are designed to impress. These questions help uncover how a system will actually work for you:
For vendors and implementation partners
- How many businesses of our size and type have you implemented in the past two years? Can we speak to three of them?
- Can you show our scenarios, such as a multi-level bill of materials, a rush order, a part rework or a supplier delay, rather than a standard demo?
- What is standard functionality and what would need customisation?
- What is included in the subscription, and what costs extra, such as additional users, modules, storage, integrations or support?
- Who will do the implementation, what is their experience, and will the same people stay through go-live?
- How are upgrades handled, and how will customisations be affected?
- How do we get our data out if we leave?
- What does post-go-live support look like, and what are the response times?
For reference customers
- How long did implementation take compared with the plan, and how much did it cost compared with the budget?
- What went wrong, and how did the vendor respond?
- What would you do differently?
- Is the system still used the way it was intended, or have workarounds crept in?
- Would you choose the same system and partner again?
A small-manufacturer scenario
A family-owned manufacturer of steel storage products has thirty staff, an accounting package, a separate quoting spreadsheet, a whiteboard for production scheduling and a paper-based stores system. Stock counts rarely match the books. Production runs out of particular components several times a month, while other items gather dust. Month-end takes ten days, and margin by product is unknown.
The owner documents the main processes and the top problems: stock inaccuracy, component shortages, no product costing and slow month-end. Requirements include multi-level bills of materials, material requirements planning, barcode stock movements, job costing and integration with the existing e-commerce site.
After shortlisting systems suited to discrete manufacturing at their size, the business runs scripted demonstrations with its own products and asks each vendor to plan a shortage scenario. It chooses a cloud system and a local implementation partner, and phases the rollout: finance, sales and purchasing first, then inventory with barcoding after a full stocktake and data clean-up, then production planning.
Six months after the final phase, stock accuracy is above 95%, component shortages are rare, month-end takes three days, and the owner can see margin by product family for the first time. That visibility leads to price increases on two unprofitable lines. The project ran about two months over plan, mainly because cleaning the bills of materials took longer than expected, a common lesson.
A simple selection process
- Weeks 1–3: document processes, problems and requirements. Agree the budget and an internal owner.
- Weeks 3–5: search for systems and partners suited to your industry and size, and shortlist three or four.
- Weeks 5–8: run scripted demonstrations using your scenarios. Check references from similar businesses.
- Weeks 8–9: compare total cost of ownership over three to five years, implementation approach, support and fit.
- Week 10: decide, agree a phased plan and start with process mapping and data cleaning.
Summary
An ERP system connects the core records of a business, including orders, purchasing, inventory, production and finance, so information is entered once and decisions are based on real-time data. It helps most when disconnected systems, poor demand visibility, manual re-entry and slow reporting are limiting the business. Choose by starting from your problems and processes, because systems are built for particular niches. Budget for the full cost, clean your data, fix your processes first, appoint an internal owner and roll out in phases. Done well, ERP turns a business that runs on memory and spreadsheets into one that runs on reliable information.
Sources: small-business training notes on business automation and enterprise resource planning, including case examples from large international businesses, together with general systems-implementation practice. No software vendor is recommended or endorsed.
