In July 2026 we walked a small healthcare startup through a proposal for a sales system. The sharpest question of the call had nothing to do with sales. The person heading up their systems had already picked a different platform for the pipeline. They wanted to know how to get one view across it, the operational systems their team runs on, and whatever they buy next. So they stopped us mid-sentence and asked: what is Fabric?
It is a fair question. The name gives away nothing.
Fabric Basics
Microsoft Fabric is Microsoft's analytics platform. It gives you one place to put all of your data so that reporting can happen in one spot.
Right now you probably have a CRM here, an accounting system there, and an operations system somewhere else. Each one holds part of your data, and none of them holds all of it. Fabric sits underneath all of them. Its Data Factory tools copy data out of each system on a schedule and land it in OneLake, a single data lake for your whole organization. Your reporting then reads from that one store.
If it helps to place it against a name you may already know, the storage and warehouse side of Fabric sits in the same category as Snowflake or Google BigQuery.
Power BI is the part people see: the reports, the charts, the numbers leadership looks at on a Monday. Power BI is one of the workloads inside Fabric, and the rest of Fabric stores and prepares the data that feeds it.
Power BI is what people look at. Fabric is where the data lives so that Power BI has something trustworthy to look at.
Non-Microsoft Systems
Neither Power BI nor Fabric requires your source systems to be Microsoft products.
Power BI has built-in connectors for Salesforce, SQL Server and other databases, and Excel spreadsheets. HubSpot, QuickBooks, and Shopify have no built-in connector. Power BI reaches systems like those through a generic interface or a custom connector, such as the system's own application programming interface. Fabric's Data Factory connectors do the same work on a schedule, at larger volumes, and with somewhere to keep the results.
Running your sales process in HubSpot puts neither of them out of reach. Nothing here needs Dynamics 365 underneath it. We run Power BI and Fabric as a practice of its own, separate from our Dynamics 365 work.
Standalone Power BI
Plenty of businesses never need the warehouse. A rough guide to where the line falls:
| Situation | Usual Suggestion |
|---|---|
| One system, fairly standard reports | Power BI connected straight to the source |
| Two or three systems, light joins between them | Power BI with scheduled refreshes and a small data model |
| Several systems, or history your source overwrites, or heavy volume | Fabric underneath, Power BI on top |
Many operational systems only show you today: what the pipeline looks like right now, what inventory sits on the shelf right now. Each update replaces yesterday's figure. If a question you care about starts with "how has this changed since March", you need somewhere to keep the history. That is the job of a warehouse in Fabric.
Worked Example
The startup on that call described a common goal. Their sales pipeline lives in the CRM. Their volume data, broken out by location, lives in the systems operations runs on. Their cost data lives in finance. Three systems, three owners.
They want someone quoting a new customer to see the likely cost to serve before the number goes out. They also want leadership to see volume by site next to what each site costs to run. Answering either question means reading all three systems in one query and comparing this month against last. That is a warehouse problem, whatever the CRM happens to be.
Data Definitions
In these projects, the harder work is agreeing what the data means.
Your CRM has one definition of an active customer. Your finance system has another. Someone has to decide which one the dashboard uses, and that decision belongs to your team. Skip it and you get a polished number that two departments both refuse to believe.
Write down the 10 questions you want answered, in plain sentences, before anyone connects a system. Half of them usually need a definition nobody has settled yet. That costs far less to find out now.
Starting Point
If you run HubSpot or anything else outside Microsoft, and you want one view across all of it, the first conversation is short. Bring us your list of systems and the handful of questions you cannot answer today without exporting to a spreadsheet. That is usually enough for us to say whether it is a Power BI job, a Fabric job, or something a tidy-up of your existing reports would fix for far less. We are glad to have that conversation whether or not a Microsoft product ends up in the answer.