Off-the-shelf or custom ERP: what should an SME of 20 to 50 people choose?

Quentin Debray
Written by Quentin Debray
Published 21 October 2026 · 9 min read

In brief

An off-the-shelf ERP, a custom ERP at €150,000, or a business tool linked to your accounting: three options, three budgets, and one question to ask before anything else.

You run a company of 20 to 50 people. Your quotes come out of Excel, your orders travel by email, and someone re-enters everything at the end of the month for invoicing. Sooner or later, the word ERP lands on the table. And as soon as you search for "ERP for SMEs", you find two opposing pitches: on one side, vendors promising an integrated tool for €30 per user per month; on the other, custom projects priced at six figures. Between the two there is a third way that nobody names, and it is often the one that suits an SME of 30 people. This article takes the three options apart, using the public benchmarks available.

What an ERP is, and what an SME of 30 people really needs

ERP stands for "enterprise resource planning". The original idea: a single application covering the whole company, with data entered once and reused everywhere.

The classic building blocks of an ERP

  • Sales: prospects, quotes, orders
  • Purchasing and suppliers
  • Stock and logistics
  • Production or project management
  • Customer invoicing
  • General and cost accounting
  • Payroll and human resources

Seven blocks. On a vendor's website, they are all shown with the same importance. In the reality of a company with 30 employees, they carry very different weight.

Where it really hurts

When you look under the bonnet of an industrial or services SME, the pain almost always sits in two or three places: the move from quote to order, tracking the progress of jobs, and re-keying before invoicing. Payroll is already handled by a provider or dedicated software, and nobody wants to touch it. Accounting is kept by the accountant, with their own tool, and again there is no point in reinventing the wheel.

In other words: an SME of this size needs a business management tool that runs from the first contact to the invoice, not an overhaul of its whole management system. That distinction changes the budget by a factor of ten. If the notion itself is still fuzzy, we have explained it in our article on what management software is.

ERP for SMEs: three options, not two

The debate is almost always presented as a duel between "buy off the shelf" and "have it built". In reality there are three paths, and the third is the one that gets explained least.

Option 1: the standard ERP from the market

Odoo, Sellsy, Axonaut, Microsoft Dynamics 365 Business Central, Sage, Divalto. Complete, proven solutions, maintained by a vendor, with an ecosystem of integrators.

Public benchmarks. Kolonell, which publishes ERP cost comparisons, puts licences between €25 and €75 per user per month, and initial integration between €30,000 and €90,000 for an SME. Over three years and twenty users, the total cost of ownership comes to around €102,000. That figure is not from a Magram client, it is a published industry benchmark: use it as such.

What makes up that total, and what price lists rarely show on their home page:

  • The initial configuration, which is the bulk of the cost
  • Migration of existing data (customers, items, history)
  • Team training
  • Add-on modules billed separately
  • Version upgrades and yearly adaptations

When it is the right choice. If your processes look like those of your sector, a standard ERP is unbeatable. Trading with stock and a classic catalogue, distribution, activities where regulation already sets the framework: the vendor has solved the problem a thousand times, and you are buying that experience.

Where it gets stuck. The breaking point comes when your business does not fit the boxes. A design office that tracks its jobs in phases, with hourly rates that vary by contributor and amendments during the project, will spend months bending a general-purpose ERP. The result is quickly recognisable: the tool is in place, and the teams keep their real tracking in a binder next to it. We devoted a whole article to this symptom, because it is far more common than people think: your software works, but your teams still work in Excel.

Option 2: the full custom ERP

This is the most misunderstood option, and often the worst costed by those who consider it.

Cegid, a leading French vendor, quotes a range of €150,000 to €800,000 on its blog for a custom ERP aimed at an SME. That range is not excessive when we are talking about a real, complete ERP: rebuilding accounting, payroll, stock management, purchasing and invoicing, with the regulatory compliance that goes with them.

And that is exactly why this path is rarely justified for a company of 20 to 50 people. Developing a custom payroll module means rebuilding something that already exists, works well, and will need maintaining with every change in legislation. The same reasoning applies to general accounting.

There are cases where it holds up: a unique business model, a size well beyond a hundred employees, a competitive advantage carried directly by the tool. For an SME of 30 people who wants to stop re-keying its quotes, no.

Option 3: the custom business tool, linked to what you already have

The middle way. We build to order only the part that makes up your business, and leave the rest to the tools that already do it well.

Concretely, what a tool of this type covers:

  • First contact and sales follow-up
  • The quote, with your own calculation rules
  • The order and the tracking of the job or project
  • Time spent and purchases assigned to the job
  • Invoicing, including progress invoices and deposits
  • Margin dashboards, by job and by period

And what we do not touch: payroll, general accounting, VAT returns. Those blocks stay with the specialised tools, and the business tool passes them the invoicing data.

At Magram, this perimeter, which we call a "business ERP", falls in a range of €15,000 to €25,000. The factor of ten compared with option 2 does not come from a magic method: it comes from the perimeter. We do not build what already exists.

Comparing the three options over three years

Basis of comparison: an SME with 20 users, three-year horizon.

Standard ERPFull custom ERPCustom business tool
Initial investment€30 to 90k of integration€150 to 800k€15 to 25k
Recurring cost€25 to 75 per user per monthMaintenance and hostingMaintenance and hosting
3-year total cost≈ €102k (Kolonell benchmark)€150k and moreTo be completed
Time to go live3 to 9 months12 to 24 monthsA few weeks to a few months
Fit with the businessAverage, unless the sector is standardisedTotalTotal on the chosen perimeter
Payroll and accountingIncludedTo be builtKept as they are
DependenceOn the vendor and the integratorOn the product codeOn the product code

The point that surprises business owners most: the TCO of a standard ERP is not low. Kolonell puts modular custom development at a comparable level, around €104k over three years, for a broader perimeter than option 3. So the question is not "which is cheapest", but "what am I spending on". We pushed this reasoning all the way in buying software at €300 a month or building a tool for €15,000.

Five questions to decide

Before looking at demos, answer these five questions honestly.

1. Is your way of pricing standard in your sector? If your quotes are calculated like your competitors', a market tool handles them. If you have a pricing method you have refined over fifteen years, no vendor planned for it.

2. How many times is data re-keyed between the quote and the invoice? Count the real copy-pastes on a typical file. Three re-entries per file, multiplied by the annual volume, give an estimate of lost time that speaks far louder than a sales pitch.

3. Who keeps the accounting and payroll today? If it is your accountant and it works, take those blocks out of the perimeter. Immediately.

4. How many people will use the tool every day? Five daily users and fifteen occasional ones do not justify the same investment as thirty intensive users.

5. Have you already tried a tool that did not take? If so, find out why before starting again. The failure rarely comes from the technology, almost always from a gap between the tool and the way people really work.

These questions overlap with the decision grid we detailed in build or buy software: how to decide.

The honest limit: when to pick a standard

We build custom software. That does not make it the right answer in every case, and saying so is more useful than selling.

Choose a market ERP if:

  • Your activity is highly standardised (classic trading, distribution, retail)
  • You need accounting and payroll integrated in the same tool
  • You expect rapid growth to more than 100 employees in the short term
  • Your sector imposes heavy, evolving standards that the vendor follows for you
  • You have nobody internally to carry a project and arbitrate priorities

That last point is underestimated. A successful custom tool requires someone on your side (often the owner or the operations manager) to be available a few hours a week during the project. Without that relay, even the best builder delivers off target.

An example of a realistic perimeter

Take a building design office of sixteen people. The needs expressed: track jobs by phase, record time spent per contributor, trigger invoicing milestones, and see the real margin per job. Payroll stays with the accounting firm, and so does accounting. The useful perimeter fits in a single tool, and it looks like no ERP on the market, because the notion of phase and amendment is central there while it is incidental elsewhere.

Where to start, concretely

Before signing anything, do this work internally. It takes half a day and is worth several sales meetings.

  1. Draw the journey of a typical file, from the first call to the payment
  2. Note at each step who enters what, in which tool
  3. Circle the re-entries and the breaking points
  4. List what already works well and must not be broken
  5. Rank the irritants by real cost, not by annoyance

That document is what an integrator, like a custom builder, will ask you for. Arriving with it means arriving with the power to compare proposals on the same basis.

If you would like an outside view of that diagram and a costed order of magnitude for your case, an hour's conversation is usually enough to know which of the three options you fall into. Even if the answer is "take a market tool".

A question about your situation?

Ask Quentin, he replies personally.

Your details are used only to reply to you. Learn more

We use audience-measurement cookies to improve the site. Learn more.