ERP selection is one of the most consequential decisions a manufacturing business makes. The platform you choose will shape your operational capability for the next decade. It will affect how you hire, what processes you can run, how fast you can scale, and how much technical debt you carry.
Most businesses get it wrong. Not because they do not try hard enough, but because the selection process they run is not designed to reveal the things that matter.
The Problem With How ERP Selection Usually Works
The standard approach to ERP selection involves:
Writing a requirements document. Inviting vendors to respond. Sitting through demonstrations. Scoring responses on a matrix. Selecting the highest-scoring system.
This process is almost entirely designed to surface what vendors want you to see. A requirements document that is too high-level lets vendors claim compliance for things they barely support. A vendor-controlled demonstration shows you the system at its best, on sample data, run by someone who has spent years making it look good. A scoring matrix rewards vendors who are good at responding to RFPs.
What you actually need to know is not in the RFP response.
What Actually Determines Whether an ERP Works
Fit to your manufacturing model, not your industry
ERP vendors sell to industries. Their marketing materials will tell you they are perfect for aerospace, or food manufacturing, or industrial equipment. This is often true in the sense that they have clients in those industries. It does not mean their system fits the specific way you manufacture.
The questions that matter are operational:
Does the system support your production model: make-to-order, make-to-stock, engineer-to-order, or a hybrid? How does it handle your BOM structure? Can it manage the revision control your engineering team needs? How does it handle lot and serial number traceability? What does multi-site or multi-currency operation look like if you need it?
These are not features that appear on a checklist. They are capabilities that need to be demonstrated in a configuration that resembles your actual operation, not in a standard demo script.
Total cost of ownership over ten years
The licence cost of an ERP system is not the cost of the ERP system. The implementation cost typically runs two to four times the first-year licence cost for a mid-market system. Ongoing maintenance, upgrades, and support run fifteen to twenty percent of the licence cost per year. Customisation, when the system does not fit your process, can dwarf all of these.
The vendors whose systems require the most customisation are also the vendors who make the most money from customisation. This is not a conspiracy. It is an incentive structure you need to account for.
When you are evaluating total cost, ask for the fully loaded ten-year cost from businesses similar to yours. Not the vendor's estimate. References.
Integration capability
An ERP does not operate alone. It needs to connect to your PDM system, your Atlassian environment, your HRIS, your customer systems, and potentially a range of manufacturing equipment and MES systems.
The integration architecture available to a given ERP platform determines how clean those connections can be. A platform with a well-documented API and a robust integration marketplace is a different proposition from one that requires custom middleware for every integration.
Ask specifically: what does the integration path to your existing systems look like? Who has built it before? What is the ongoing maintenance burden?
Upgrade stability
ERP platforms that require significant customisation for basic operation also tend to make upgrades expensive. Every time the vendor releases a new version, your customisations need to be reviewed, tested, and often rebuilt. Some businesses on heavily customised ERPs effectively cannot upgrade and run version-stale systems for years.
This is a support and security risk as well as a capability risk. Ask vendors what the upgrade experience looks like for their typical implementation. Ask references whether they are current on the platform version.
The Traps
Buying for where you are, not where you are going
The most common ERP selection mistake is selecting a system that is right for your business today and inadequate for your business in three years.
A fifty-person manufacturer selecting an ERP should be thinking about what the system needs to support at one hundred and fifty people. The data model, the multi-site capability, the reporting depth, the API ecosystem. Scaling an ERP implementation is expensive. Replacing an ERP because you outgrew it is very expensive.
Letting the IT manager decide alone
ERP selection affects every operational function of the business. Finance, manufacturing, engineering, service. The selection process needs input from the people who will use the system, not just the people who will implement it.
The most dangerous outcome is a system that the IT team can implement but the manufacturing team will not adopt. A QMS built for the audit, not the operation. An ERP selected for its technical architecture rather than its operational fit.
Operations leadership needs to be in the room, with real weight, during evaluation.
Being sold on the roadmap
ERP vendors frequently respond to gaps in their current capability with promises about upcoming features. Version 2.3 will have the functionality you need. It is on the roadmap.
Buy what the system does today. If the feature does not exist in a released version you can evaluate, do not make your decision on the assumption it will arrive.
Underweighting implementation quality
The implementation partner matters as much as the platform. A strong ERP on a poor implementation will fail. A solid ERP on a strong implementation will succeed.
When you evaluate vendors, evaluate their implementation capability independently of the platform. How many implementations of a similar size and complexity have they completed? Can you speak to references? What does their project management methodology look like? How do they handle scope changes?
A Better Selection Process
Start with architecture, not selection. Before you talk to a single vendor, document your operational processes and what you need the system to do. Not at the feature level. At the business process level. A clear operational requirements document that a manufacturer wrote, not a requirements template downloaded from the internet.
Shortlist based on model fit, not brand. Tier-one platforms are right for some manufacturers and wrong for many. The question is whether the platform fits your operational model, integrates with your existing systems, and can be implemented within your budget and risk tolerance.
Run scripted demonstrations. Give shortlisted vendors your data and your scenarios and ask them to demonstrate the system against your actual requirements. Not their standard demo. Your processes, your products, your workflows.
Check references properly. Talk to businesses similar to yours in size and operational complexity. Ask specifically about the implementation experience, the support quality after go-live, the total cost relative to the initial estimate, and whether they would choose the same platform again.
Include integration architecture in the evaluation. Your ERP does not operate alone. The integration path to your Atlassian environment, your PDM, your HRIS, and your identity management needs to be evaluated as part of the selection, not solved after go-live.
When to Walk Away
If a vendor cannot or will not demonstrate their system against your specific processes on your data, walk away. A vendor confident in their system has no reason to refuse this.
If the implementation estimate is dramatically lower than market rates, treat it as a sign of either inexperience or a plan to recover margin through scope changes. Either is a problem.
If references are unavailable or reluctant to speak, treat that as a signal.
Big Finish is system agnostic. We do not have vendor relationships that influence what we recommend. If you are selecting an ERP and want an objective view of your options, book a discovery call.