Key takeaways
- Business Central is licensed per user per month, essentially in the Essentials and Premium variants.
- Not every employee needs a full licence — Team Member licences cover read and light tasks far more cheaply.
- In mid-market projects, implementation typically costs a multiple of the annual licence spend.
- The biggest cost driver is not licences but custom development and interfaces.
- Monthly terms are noticeably more expensive than annual licences.
The licence side: Essentials, Premium and Team Member
Business Central has two full licences. Essentials covers most of what mid-sized companies need: financials, purchasing, sales, inventory, projects, basic warehousing. Premium adds manufacturing and service order management. Within one tenant all full users must be on the same variant — mixing Essentials and Premium is not provided for.
Alongside these is the Team Member licence for employees who mostly read, record time or maintain data in their own records. It costs a fraction of a full licence. In practice it is often overlooked, although it can save considerable amounts: in many companies only part of the workforce really needs full access.
Prices change; Microsoft has adjusted Business Central conditions recently. So don't rely on figures from an older blog post but on a current quote. One point stays stable, though: monthly terms cost noticeably more than annual contracts.
Implementation: the actual chunk
Anyone who only calculates licences is calculating the smaller part. Implementation covers process analysis and design, system setup, data migration from the legacy system, testing, training and go-live support. In mid-market projects this block regularly amounts to a multiple of the annual licence cost.
The range is wide and hangs on a few questions: how many business areas are affected? How clean is the master data? How many historical documents have to be migrated? How many interfaces to shop, bank, payroll, CRM or logistics exist? And above all: how much custom development is really necessary?
That last point matters most. Every customisation costs not only once during development but permanently — it has to be retested with every release wave. So it pays to seriously question every special case: is the process really a competitive advantage, or did it just grow that way historically?
What quotes tend to leave out
When comparing quotes, look at the items that are not in them. Frequently missing:
- Data cleansing before migration — almost always needed, rarely budgeted
- Effort on your side: key users need time for testing and coordination
- Training beyond go-live, when daily business raises the real questions
- Ongoing support and the semi-annual release waves
- Add-on solutions from the marketplace including their own licence costs
- The stabilisation phase after go-live, which experience shows generates rework
How to realistically keep costs down
The most effective lever is staying close to the standard. The closer you stay, the cheaper implementation, operation and every future update become. That doesn't mean forgoing sensible customisations — it means justifying them.
The second lever is the licence mix. Honestly check who really needs a full licence. The third is a staged roll-out: core first, extensions later. That spreads cost, lowers risk and lets your team grow into it.
And the fourth: clean master data before migrating. Data quality is not a side topic — bad data makes every ERP project more expensive, no matter which system sits underneath.
Frequently asked questions
- Business Central Online has no high entry barrier in terms of user count — even small teams can license it. An ERP project only becomes economically sensible once process complexity justifies it, though; below that, leaner solutions are often a better fit.
- Looking purely at licences, on-premises sometimes appears cheaper. Add servers, backup, maintenance windows, updates and your own staff time, and in most mid-market setups the picture turns in favour of the cloud. On-premises remains sensible where regulatory or technical reasons call for it.
- For clearly delimited projects, yes. The precondition is an upfront analysis documenting scope, customisations and interfaces. Without that basis, any fixed price is either a risk premium or an invitation to argue about change requests.