Disclosure: we build custom ERP systems. We also recommend products more often than we recommend builds, and this page reflects that.
If you take one thing from it: the real choice is rarely “product or custom”. It is usually “which product, and what do we build around it”.
The question that actually decides it
How much of this product will you actually use, and what are you paying for the rest?
Products are built for the average of thousands of businesses. You pay for all of it and use perhaps a third. The two-thirds you do not use is not free — it is in the per-seat price, every month, rising at renewal.
The second question: what will you do about the parts that do not fit? The answer is almost always workarounds — a spreadsheet alongside, a manual step, an export and re-import. Those cost real hours, every week, for as long as you run the product.
A product is the right answer at genuinely low headcount with genuinely standard processes. Past roughly 25–30 users, the arithmetic shifts — and it keeps shifting with every seat you add.
The options
Odoo
Shape: modular open-source ERP. Pick the applications you need. Community edition free, Enterprise paid per user.
Typical UK cost: Enterprise roughly £20–£40 per user per month plus implementation. A 25-user implementation commonly lands £25,000–£80,000 depending on modules and customisation.
Strong when: you want broad functionality cheaply, you have a partner who knows it well, and some customisation is expected. The source is available, so deep modification is possible.
Weak when: customisation accumulates. Heavily modified Odoo becomes painful to upgrade, and upgrades are where the maintenance cost shows up. Quality varies enormously between implementation partners — the partner matters more than the product.
NetSuite
Shape: cloud ERP, Oracle-owned, aimed at mid-market. Comprehensive, mature, subscription-based.
Typical UK cost: base platform plus per-user licensing, commonly £20,000–£60,000+ a year for a mid-size business, plus implementation often £40,000–£150,000.
Strong when: multi-entity, multi-currency, international consolidation, strong financial reporting requirements. Genuinely capable at what it does.
Weak when: you are small, cost-sensitive, or your processes are unusual. Customisation is possible but specialised. Renewal pricing is a known negotiation point — model year three and year five, not year one.
SAP (Business One / S/4HANA)
Shape: Business One targets SMEs; S/4HANA is the large enterprise product.
Typical UK cost: Business One commonly £40,000–£150,000 all-in for a mid-size implementation. S/4HANA is a different category entirely — six and seven figures.
Strong when: you are in a supply chain where partners expect SAP, you operate at genuine scale, or you need deep manufacturing capability.
Weak when: you are an SME without those drivers. The implementation effort and specialist cost are substantial. For most UK businesses under 250 staff, SAP is heavier than the problem requires.
Microsoft Dynamics 365 Business Central
Shape: Microsoft’s SME ERP. Strong integration with the Microsoft estate.
Typical UK cost: roughly £55–£80 per user per month, implementation commonly £25,000–£100,000.
Strong when: you already run Microsoft 365 and want things to connect. Large UK partner network, so you can change partner without changing product — a genuinely underrated advantage.
Weak when: your processes diverge significantly. Customisation is capable but adds cost and upgrade friction.
Custom build
Typical UK cost: operational layer over existing accounting £25,000–£60,000; mid-size custom ERP £60,000–£150,000; large multi-site £150,000–£400,000+. Plus hosting and maintenance.
Strong when: the process that makes you competitive is unsupported by any product, you have ended up with a product surrounded by workarounds, or licence costs for functionality you half-use have become indefensible.
Weak when: your processes are standard. You would be rebuilding, at your own expense, something you could buy — and inheriting the maintenance.
Five-year cost comparison
Scenario: UK wholesaler, 30 users, stock and order management plus accounting integration.
| Year 1 | Years 2–5 | 5-year total | |
|---|---|---|---|
| Odoo Enterprise | £45,000 impl. + £10,800 lic. | £10,800/yr + £4,000/yr maint. | ~£115,000 |
| Business Central | £55,000 impl. + £23,400 lic. | £23,400/yr + £4,000/yr maint. | ~£188,000 |
| NetSuite | £80,000 impl. + £35,000 lic. | £35,000/yr + £5,000/yr maint. | ~£275,000 |
| Custom build | £110,000 build + £3,000 hosting | £3,000/yr + £8,000/yr changes | ~£157,000 |
| Hybrid (keep Xero + custom ops layer) | £45,000 build + £2,000 | £2,000/yr + £5,000/yr changes | ~£75,000 |
Indicative only. Licence pricing varies by negotiation, module mix and contract length. Model your own figures.
Three things this shows:
- Custom beats NetSuite by around £118,000 over five years at this scale — and you own the system at the end rather than facing a renewal.
- Odoo is genuinely cheap at this size. It is the serious product contender, and the risk sits in implementation partner quality and upgrade friction once customisation accumulates.
- The hybrid wins outright — and almost nobody proposes it, because it is the smallest invoice available to everyone in the room.
None of the product columns include workaround labour. Add 30 minutes a day of someone moving data between systems — roughly £1,900 a year — and every product column rises by about £9,500 across five years.
Change the headcount and the picture moves. At 8 users, buy. At 150 users, per-seat licensing across a five-year term makes a custom build look conservative.
The hybrid nobody proposes
Keep the accounting package you already run. Build only the operational layer that does not exist. Integrate them properly.
Most businesses convinced they need a full ERP actually need:
- Xero, Sage or QuickBooks — untouched, working, compliant, cheap
- Three integrations so systems stop disagreeing
- One custom module for the process no product supports
Typically a quarter the cost of a full build, with a fraction of the risk. You keep a tested, supported, HMRC-compliant product for the 80% that is standard, and build only the 20% that is actually yours.
Nobody proposes this because: ERP vendors want a licence, implementation partners want an implementation, and custom agencies want a build. The hybrid is the smallest invoice available and frequently the right answer.
Ask every supplier whether it would work for you before accepting that a full ERP is necessary.
Evaluating a product properly
Most “no product fits” conclusions come from a sales demo and an assumption. Most “this product fits” conclusions come from the same place. Both are expensive.
Run a real trial, not a demo. Two weeks, your data, your people, your actual processes. A demo is a performance given by someone who knows exactly which buttons to avoid.
Load your worst data, not your cleanest. The customer with four duplicate records. The product with the awkward unit conversion. The order with the contract pricing exception.
Give it to the people who will use it. Not the project sponsor. The person entering orders will find three blockers in an afternoon that nobody else would notice in a month.
Test the five hardest things, by name. Write them down first so the evaluation cannot drift toward what the product does well:
- Your most complex pricing scenario
- Your most awkward stock allocation
- The report you currently build by hand
- The integration you most depend on
- Your month-end close
Count the workarounds. Every time someone says “we’d just export that to Excel” or “we’d keep a note of that separately”, write it down. Three or more on core processes means it does not fit — and each one is 20 to 40 minutes a day for someone, for ever.
Check the exit before you enter. Can you export your data in a usable, structured form? What format? Does it include history and attachments? Ask during the trial, when they want your business, not in three years when they do not.
How to decide
1. Count your users and model five years. Per-seat licensing is the single biggest variable. Low headcount favours products decisively.
2. Trial at least two products with your real data. Not a demo — a trial, with your records and your people, for two weeks. Most “no product fits” conclusions come from a sales meeting rather than an evaluation.
3. List exactly what does not fit. Be specific. “Stock allocation across three warehouses using our priority rules” is a requirement. “It’s not flexible enough” is not.
4. Ask whether the gaps could be built around the product. This is the hybrid question and it is where the money is.
5. Judge the partner, not just the product. For Odoo, Business Central and SAP, implementation quality varies more than the products do. A good partner on a mediocre product beats a poor partner on an excellent one.
6. Check the exit. Can you export your data in a usable form? What does leaving look like? Ask before signing, not after.
The partner matters more than the product
For Odoo, Business Central and SAP Business One, implementation quality varies more than the products do. A good partner on a mediocre product beats a poor partner on an excellent one, reliably.
What separates them:
Sector experience with your specific complexity. A partner who has implemented for three wholesalers with multi-location allocation knows where it goes wrong. One who has done retail and professional services will learn on your project.
Willingness to say a module does not fit. A partner who answers every requirement with “yes, it does that” is either not listening or planning to bill the gap as customisation later.
Customisation discipline. The best partners resist customisation and look for configuration first, because they know what heavy customisation does to upgrades. A partner enthusiastic about customising everything is building you a maintenance liability with their logo on it.
Who owns the customisations. Ask directly. Some partners retain the code they write on top of the product and licence it back. Same question as with any supplier.
Portability. Business Central has a large UK partner network, so you can change partner without changing product. Odoo and SAP have smaller pools for any given specialism. This is an underrated factor — your relationship with a partner can end for reasons unrelated to the software.
Questions to ask any implementation partner:
- How many implementations in our sector, and can we speak to one?
- What have you told a client they could not do with this product?
- Who owns customisations you write?
- What happens at the next major version upgrade?
- Which modules will we need that are not in this quote?
Why ERP projects fail regardless of route
The Panorama Consulting 2026 report, surveying 170 organisations, found roughly 30% of ERP projects finished over budget and 22.3% exceeded their schedule. Among the over-budget projects, 54.9% needed technology discovered after the project began and 51% saw scope grow.
Gartner projects that more than 70% of recently implemented ERP initiatives will fail to fully meet their original business-case goals by 2027.
Note that none of those failures is about product versus custom. They are about scoping, data, adoption and ownership. The detail is here.
Choosing the right route matters far less than running the project properly.
The licence trap, specifically
The cost nobody models properly, because it does not appear until year three.
Renewal repricing. Introductory pricing is common, and renewal is where the real rate appears. Your negotiating position at renewal is weaker than at purchase, because your data, your processes and your training are now inside their product.
Per-user pricing and growth. A system priced at 25 users becomes a system priced at 40 when you grow. The software did not change; the bill rose 60%. Custom software costs the same at 40 users as at 25.
Module creep. The base licence covers the core. Advanced inventory, EDI, manufacturing, better reporting — each is a module with its own fee. The quoted price and the price in year two frequently differ substantially.
Mandatory version upgrades. Some vendors end support for older versions, making an upgrade compulsory. If you have customised, that upgrade has a cost you did not choose to incur.
The arithmetic over ten years — the realistic life of a business system:
| 10-year cost, 30 users | |
|---|---|
| Product at £45/user/month with 8% annual rises | ~£235,000 |
| Custom build £110,000 + £8,000/year | ~£190,000 |
And at the end of ten years: one scenario leaves you owning a system you can change, sell with the business, or hand to another firm. The other leaves you with a renewal notice and the same negotiating position, only weaker.
None of this makes products wrong. At low headcount they win comfortably. It makes the five-year and ten-year arithmetic worth doing before signing a three-year contract.
Frequently asked questions
Which is the best ERP for a UK SME?
For most, a product rather than a custom build. Odoo for cost-sensitive businesses with a good partner; Business Central if you already run Microsoft 365; NetSuite for multi-entity and multi-currency complexity. Custom only where a specific process is genuinely unsupported.
Is Odoo good enough for a serious business?
Yes, for many. It is capable and cost-effective at SME scale. The risks are implementation partner quality and upgrade friction once customisation accumulates.
Is custom ERP cheaper than NetSuite?
Over five years, often yes at 30+ users, because there is no per-seat licence. It carries considerably more delivery risk and you own the maintenance. Model your own headcount — the answer flips at lower user counts.
Should I use SAP?
Probably not, if you are a UK SME without a supply chain that expects it or genuine manufacturing complexity. Business One is capable but heavier than most sub-250-staff businesses need.
What is the hybrid approach?
Keeping your existing accounting package and building only the operational layer that no product provides, integrated properly. Typically a quarter of the cost of a full ERP build with far less risk, and rarely proposed because it is the smallest invoice.
How do I know if a product genuinely does not fit?
Trial at least two with your real data and real users for two weeks. If you are still designing workarounds for the same core process afterwards, the gap is real — and even then, ask whether building around the product beats replacing it.
How long does each option take?
Odoo or Business Central implementation typically 3–9 months. NetSuite 6–12 months. SAP Business One 6–12 months. A custom operational layer 2–4 months; a full custom ERP 5–9 months.
Written by Jas Singh, Founder. Last updated 8 October 2026. We build custom ERP and are therefore not a neutral party — the cost models are shown so you can check the arithmetic with your own figures.
Licence and implementation costs are indicative UK market observations and vary substantially by negotiation, module mix and partner. Project outcome figures from the Panorama Consulting 2026 ERP report (170 organisations, January 2025 – January 2026) and Gartner, reported via secondary sources and cited as directional.
