Business Central Fabric Mirroring vs Jet Analytics

As part of FabCon Europe in Barcelona, Microsoft announced Business Central mirroring into Microsoft Fabric. Selected BC tables replicate into OneLake within minutes, without touching BC's APIs and with no charge on the BC side. Microsoft is also rebuilding its nine free Power BI apps for Business Central on Fabric. Together, these remove most of the reason mid-market BC customers bought packaged reporting stacks such as Jet Analytics. The argument: the basics are now close to free, and the money worth spending sits in three places the free kit doesn't reach: data from outside BC, governance of a copy that includes every field, and your own customisations.

Do I still need to keep paying the t(r)oll?

Business Central mirroring is the most consequential thing to happen to BC reporting in years, and most BC customers won't have noticed it yet as the announcement is super fresh. This brings to BC level a capability that's been available in it's big brothers F&O toolkit for some time.

A couple of questions for Jet users now arise. How much of your reporting budget goes on getting data out of Business Central, before anyone has analysed anything? And when did someone last question the reporting licence renewal that makes that happen?

Why did anyone pay the troll in the first place?

Fundamentally, getting data out of BC was genuinely hard. Every extract went through BC's APIs, with their limits, their refresh times and their habit of slowing production if pulling too much data. The road out of BC was a goat track, and the only decent crossing was a bridge with a troll under it, collecting a toll on every trip. Paying the troll was a perfectly rational decision (if you remember the Three Billy Goats Gruff, the goats didn't have many options either).

Comparisons published earlier this year by BC partners still framed Fabric as the months-long, high-overhead enterprise option, with Jet as the sensible mid-market choice. That was a fair read prior to the FabCon accouncement. It isn't now, because Microsoft has just built a free bridge downstream.

What does Business Central mirroring actually do?

It replicates the tables and companies you choose into OneLake, with latency of a few minutes. It uses an internal replication service rather than the BC APIs, so it doesn't eat your API limits and isn't expected to slow production. There's nothing to pay on the BC side. The cost is Fabric capacity, and F2 runs roughly A$400 a month pay-as-you-go in Australia East, or about A$250 on a one-year reservation.

Then there's the part that matters to a CFO. Microsoft already gives away nine Power BI apps for Business Central, covering everything from finance and sales to manufacturing and sustainability, and it has released the source code for all of them. Their weakness has been that each company in your environment needs its own copy of each app, in its own workspace. Microsoft is now rebuilding all nine on Fabric, with a multi-company version of each.

The usual caveat applies. Public preview is expected in BC version 29.1, and the BC-specific documentation isn't written yet. Whether this holds up in the wild is another question, and I may well be revising this article by Christmas.

Should you cancel Jet?

Not on Monday. But I'd look hard at the next renewal (Full disclosure – Talos is a Microsoft partner with no formal relationship with Jet, albeit some experience)

The pitch for Jet Analytics rests on two things: a pre-built BC data model, and automated consolidation across companies. Microsoft already gives away the first, and it's about to give away a credible version of the second. In my opinion, the warehouse layer is now the hardest part of that stack to defend.

The usual objection is "but our finance team lives in Excel". It doesn't hold up the way you'd expect. Under the hood, Jet Analytics serves its data through an SSAS cube. Swap that for a Power BI semantic model and Excel connects to it in much the same way, through PivotTables and cube formulas. Your finance team can stay exactly where they are.

The real switching cost is the workbooks. Years of Jet Reports are built on Jet's own formulas, and none of them will point at a semantic model, so every one has to be rebuilt. That's the number to get before you renew. Not the licence fee, but how many reports you'd have to rebuild, and how many of them anyone still opens.

Equally, most BC customers run a single company, so consolidation was never their real problem. For the groups that do need intercompany eliminations and currency translation, I wouldn't expect the free apps to do that work for you.

So where should the money go instead?

The free kit gets you good standard reports on standard BC data. Most businesses don't run on standard BC data alone, and that's where the real work is.

The first gap is everything that isn't in BC. Project margin needs hours from the timesheet system. Revenue per head needs payroll. Pipeline needs the CRM, and the budget almost certainly lives in a spreadsheet. The nine apps know none of it exists. The whole point of OneLake is that this data can sit beside your mirrored BC tables without being copied around. But someone still has to land it, model it and make it agree with the ledger.

The second is governance. Mirroring lets you pick tables and companies but not fields, so every column comes across. Once it's in OneLake, Fabric workspace permissions decide who sees it, not BC's permission sets. The careful role design in your ERP doesn't follow the data. Vendor bank details, employee records and margin by customer all sit in a copy that needs its own security model before the first report is shared.

The third is your customisations. Very few BC sites are vanilla. Most carry ISV apps, extensions and custom fields that reflect how the business actually works. Mirroring brings extension tables across, but the standard apps only model standard tables, so your custom data lands in OneLake and sits there unused. Because the apps are open source, extending them is a scoped job rather than a rebuild. The catch is that Microsoft keeps updating them, and someone has to fold each new version into yours.

To Close & next steps

Business Central mirroring turns BC reporting from a plumbing purchase into a design decision. Microsoft has built a free bridge, and the troll is welcome to stay under his. The value now is in what you build on the other side. I would suggest:

• Don't sign a multi-year reporting renewal until you've seen mirroring running on 29.1.
• Count your Jet Reports workbooks, and how many of them anyone still opens. That's your real migration cost.
• Spend on the gaps: outside data, governance and your customisations.

If you've got a BC reporting renewal on the horizon, we'd be glad to look at what the free kit covers for you, and what it doesn't, before you sign.

Cheers, James