Do we actually need Microsoft Fabric?

Often not. Fabric earns its place when the analytical problem is bigger than any single system: multiple operational platforms, large volumes, long history, complex transformation or shared enterprise data. If the requirement is one finance dashboard or a contained Power BI model, Fabric adds cost, capacity planning and skills you then have to keep.

Should the report come from the ERP or from Power BI?

Operational questions such as open sales orders, overdue invoices, stock availability and open cases are usually best answered in the system that owns them. Power BI earns its place when the answer needs data from more than one system, or context the operational screen cannot give.

Why does nobody trust our dashboards?

Usually the definition is unclear, the data is late, the source is disputed, or the exception has no owner. It is very rarely a design problem, which is why redesigning the dashboard tends not to fix it.

Will better reporting fix our data quality?

No. It will make the problem visible, which is useful, but the correction still needs named owners and a bit of unglamorous work. Where that is the real issue, we will say so rather than building something prettier on top of it.

Do you have reporting outcome figures we can look at?

Not published ones, and we will not invent them. We agree what the decision is and what it looks like today before starting, so the change can be judged against something real. Customers can speak to what we are like to work with, which is a different and more honest kind of evidence.

Power BI & Microsoft Fabric

When the meeting argues
about whose number is right.

Most organisations do not need more data. They need trusted numbers, delivered early enough to act on. Power BI solves that problem for many businesses. Fabric matters when the challenge moves beyond reporting into data integration, engineering, governance and scale.

How we decide what you need

Start with the decision. The architecture follows, and it is often smaller than expected.

From data to decision

01TransactionSomething happened in the business.
02InformationIt is recorded in a usable form.
03InsightIt is explained, not just displayed.
04ExceptionSomething needs attention.
05DecisionSomeone decides what to do.
06ActionSomething actually changes.

What supports it

ERP, CRM and other systems
Power BI, or Fabric with Power BI
AI and Predictive Intelligence, where justified

Start with the decision. Then decide what architecture it justifies.

Sound familiar?

Nobody asks for a dashboard.

They ask why the figures never match, why the pack takes a day to build, and why the red number has been red since March.

The meeting argues about the number
Two people, two spreadsheets, and twenty minutes gone before anyone discusses the business.
The pack is built by hand every month
One person exports, pastes and reconciles, and everyone hopes they are not on holiday.
There are dashboards, but people still ask for a report
The clearest sign nobody quite trusts what is on the screen.
It shows the total, not the thing that changed
So it gets looked at once, then quietly ignored.
Nobody owns the exception
The red number has been red for months and it is not anybody's job to fix it.

Where we start

The decision first. The platform later.

Four steps, in this order. Plenty of reporting work finishes at step two, because the problem was never the technology.

Stage 1 of 4

Ask what decision it serves

What we do

If nobody would do anything differently, the report does not need building.

Whose decision is it?

Pick the person and see the question.

Reporting gets useful when it belongs to somebody with a job to do.

Where is cash pressure building?

Explore Finance and Supply Chain

CFO dashboard

Financial overview and performance

Last 12 monthsIllustrative

Total revenue

£28.7m

+8.6% vs LY

Gross margin

32.4%

+2.1 pts vs LY

Operating profit

£6.2m

+12.7% vs LY

Cash balance

£14.3m

-4.3% vs LY

Revenue trend (monthly)

MayAugNovFebApr

Cost by category

Labour38%
Transport26%
Facilities15%
Technology10%
Other11%

Profit and loss summary

Metric	Actual	vs LY
Revenue	£28.7m	+8.6%
Cost of sales	£19.4m	+6.1%
Gross profit	£9.3m	+12.4%
Operating costs	£3.1m	+5.2%
Operating profit	£6.2m	+12.7%

Working capital

Inventory

£7.4m

Receivables

£5.2m

Payables

£3.1m

Cash cycle

42 days

Illustrative view. Exceptions route to a named credit owner.Explore Finance and Supply Chain

Built-in intelligence

From reading the number to asking what changed.

Once the definitions and the data are trusted, newer capability can shorten the gap between seeing something in a report and knowing what deserves attention.

Power BI1 of 4

Ask what changed.

Copilot can summarise a report, a page or a visual and answer questions grounded in the information already in Power BI, within whatever the person is allowed to see.

What this gives you

Why it matters. Less time reading every chart before reaching the movement or exception that needs attention.
How we help. We make sure the measures, labels and semantic model are clear enough that a question means the same thing to the business and to the system.

Availability

Microsoft status: Depends on Fabric capacity, tenant settings, region and permissions. We confirm current availability before we design around it.

Power BI2 of 4

Describe the view you need.

A report author can ask Copilot to suggest or create pages and visuals from a plain description of the question the report should answer. The first version still needs review.

What this gives you

Why it matters. Less effort on the first draft, so more attention goes on whether it answers the right business question.
How we help. We agree the decision, the measures and the semantic model first, then use the tooling to speed up authoring rather than letting it decide what the numbers mean.

Microsoft Fabric3 of 4

Ask governed data a question in plain English.

Fabric data agent can answer natural-language questions against approved sources such as lakehouses, warehouses and Power BI semantic models. A useful agent needs business meaning as well as data: clear names, a focused domain, instructions and example questions.

What this gives you

Why it matters. People can explore governed information without writing SQL, DAX or KQL for every question.
How we help. We choose the sources, define the business terminology, set permissions and instructions, and give it example questions so it understands what the organisation actually means.

Availability

Microsoft status: Generally available on Fabric capacity. Some newer source and knowledge extensions remain preview, so we check status per source.

Microsoft Fabric4 of 4

Do something when the number changes.

Activator watches for a defined condition in data and triggers an action: an email, a Teams message, a Power Automate flow or a Fabric activity. It is rule and event driven, not generative. Stock cover drops below the agreed threshold and the right person hears about it now, rather than in tomorrow's report.

What this gives you

Why it matters. The exception that matters does not have to wait for somebody to open a dashboard and notice it.
How we help. We agree which condition genuinely deserves action, who is notified, what may happen automatically and how to stop alerts turning into noise that people mute.

Swipe, drag or use the arrow keys to move between decisions.

These capabilities work best once the information underneath them is trusted. Where the requirement moves into prediction, custom AI, wider agent design or decisions spanning several platforms, that becomes a broader Data and AI conversation. Explore Data & AI

Microsoft Fabric

Bring your data, analytics and AI together in one platform.

Microsoft Fabric is Microsoft’s end-to-end analytics platform. It gives data professionals and business users a single environment to ingest, store, process and analyse data, rather than relying on separate tools and disconnected data platforms.

Microsoft Fabric brings data integration, engineering, analytics, AI and business intelligence into one unified platform. Different Fabric experiences are designed for different types of data work, but they operate together on the same shared data foundation.

One platform. Nine ways to work with your data.

Data Engineering

Build and prepare the data foundation. Microsoft Fabric Data Engineering helps teams collect, organise, process and transform large volumes of data. It supports lakehouses, notebooks, Spark jobs and pipelines so data can be prepared for analytics, reporting and AI.

Best for:

Lakehouses · Data transformation · Spark · Notebooks · Large-scale data processing

Data Factory

Connect, move and transform data at scale. Data Factory brings together data from databases, files, cloud services and on-premises systems. It supports data movement, orchestration and transformation, helping organisations turn fragmented data into information that is ready for analytics.

Best for:

Data integration · ETL/ELT · Pipelines · Data movement · Workflow orchestration

Data Science

Turn data into predictions and intelligent insight. Fabric Data Science provides an end-to-end environment for exploring and preparing data, training machine-learning models, tracking experiments, scoring models and making predictive insights available to the wider business.

Best for:

Machine learning · Predictive analytics · Model training · Experimentation · AI workflows

Data Warehouse

Create a governed SQL foundation for enterprise analytics. Fabric Data Warehouse provides an enterprise-scale relational warehouse built on a data lake foundation. It supports T-SQL, dimensional modelling, governed semantic models and high-performance analytics while integrating directly with OneLake and Power BI.

Best for:

Enterprise reporting · SQL analytics · Dimensional models · Data marts · Governed BI

Databases

Run operational data and analytics in the same platform. Fabric Databases support operational database workloads using SQL database and Cosmos DB capabilities. Organisations can manage transactional data while keeping it closely connected to analytics, engineering and AI experiences across Fabric.

Best for:

Operational applications · Transactional data · SQL · Cosmos DB · AI-enabled applications

Graph

Understand how your data is connected. Graph in Microsoft Fabric helps organisations model, visualise and analyse complex relationships across their data. It can uncover connections, dependencies and patterns that are difficult to identify using traditional relational analysis alone.

Best for:

Relationship analysis · Dependencies · Network analysis · Connected data · AI reasoning

Industry Solutions

Apply Fabric to industry-specific data challenges. Industry Solutions in Microsoft Fabric provide data capabilities designed around the requirements of particular sectors. These solutions help organisations integrate industry data, accelerate analytics and address sector-specific operational and data-management challenges.

Best for:

Industry data models · Sector-specific analytics · Accelerated implementation · Domain-focused insights

Fabric IQ

Give data business meaning. Fabric IQ provides a shared intelligence layer over organisational data. It brings together business entities, relationships, rules and semantic models so people and AI agents can understand data in the context of how the business actually operates. It can help create a common business language across data, analytics and AI.

Best for:

Business semantics · Ontologies · AI grounding · Data agents · Context-aware decision-making

Power BI

Turn trusted data into decisions. Power BI is a core Microsoft Fabric workload for analytics and visualisation. It allows users to create reports, dashboards and semantic models, share insight across the organisation and interact with data through familiar business intelligence experiences.

Best for:

Dashboards · Reporting · KPIs · Semantic models · Business intelligence

How Fabric works

From fragmented data to usable insight.

Connect
Bring together data from ERP, CRM, operational systems, files, cloud applications and other sources.
Ingest & Transform
Use Data Factory and engineering tools to move, clean and prepare the data.
Store
Use OneLake, Lakehouse or Warehouse depending on the type of data and analytical requirement.
Analyse
Use SQL, notebooks, machine learning and real-time analytics to explore and understand the data.
Visualise
Surface information through Power BI reports and dashboards.
Act
Use real-time intelligence, AI and automation to respond to events and support faster decisions.

Why Microsoft Fabric?

A more connected approach to data and analytics.

One shared data foundation

OneLake provides a common data layer across Fabric workloads, reducing the need to maintain separate analytical data stores for every team or tool.

Less movement between platforms

Fabric workloads operate on the same underlying data foundation, allowing engineering, science and BI teams to work with shared data rather than repeatedly copying it between systems.

Analytics and AI together

Fabric supports data engineering, data science, machine learning, reporting and AI workflows in one managed platform.

Real-time visibility

Real-Time Intelligence can process and analyse streaming information so organisations can react to changes as they occur rather than waiting for scheduled reporting.

Built for different data roles

Data engineers, analysts, data scientists and business users can work in the same Fabric environment using experiences designed for their roles.

Fabric + Power BI

Fabric brings the data together. Power BI turns it into decisions.

Microsoft Fabric provides the data integration, storage, engineering, science and real-time analytics capabilities behind the platform. Power BI then provides the reporting and visualisation layer used to consume that information across the organisation. Microsoft also supports Direct Lake, allowing Power BI to work directly with data stored in OneLake without duplicating that data into a separate semantic model.

Business Systems

Dynamics 365 · Business Central · CRM · Files · Operational Data

Microsoft Fabric

Data Factory → OneLake → Engineering → Warehouse → Data Science → Real-Time Intelligence

Power BI

Dashboards · KPIs · Analysis · Forecasting · Decision Support

Where Fabric can help

Use cases that match the platform.

Unify fragmented data

Bring data from different business systems into a common analytical environment.

Modernise reporting

Create a governed foundation for Power BI rather than building reports directly from disconnected source systems.

Prepare data for AI

Clean, organise and govern business data before applying machine learning and AI.

Build predictive models

Use Fabric Data Science to train, track and operationalise machine-learning models.

Analyse operational data in real time

Use Real-Time Intelligence for scenarios such as IoT, manufacturing, system monitoring and operational events.

Create a scalable data platform

Use OneLake, Lakehouse and Warehouse capabilities to support different structured and unstructured analytical workloads.

Customer experience

Customers describing what we are like to work with.

The strongest evidence of how we work comes from the organisations we work with.

80%

of our business comes from existing customers and referrals.

All customer stories

The best I’ve ever experienced in over 20 years of collaboration.
Programme Manager, Hill & Smith Plc

Before you call

The questions leaders ask us first.

A conversation, not a demo

Tell us which number nobody believes.

Describe the decision it should support and where the data sits today. We will give you an honest view on whether that needs a report, a definition everyone signs up to, or genuinely a data platform.

See customer stories

When the meeting arguesabout whose number is right.

Nobody asks for a dashboard.

The meeting argues about the number

The pack is built by hand every month

There are dashboards, but people still ask for a report

It shows the total, not the thing that changed

Nobody owns the exception

The decision first. The platform later.

Ask what decision it serves

Pick the person and see the question.

From reading the number to asking what changed.

Ask what changed.

Bring your data, analytics and AI together in one platform.

One platform. Nine ways to work with your data.

Data Engineering

Data Factory

Data Science

Data Warehouse

Databases

Graph

Industry Solutions

Fabric IQ

Power BI

From fragmented data to usable insight.

Connect

Ingest & Transform

Store

Analyse

Visualise

Act

A more connected approach to data and analytics.

One shared data foundation

Less movement between platforms

Analytics and AI together

Real-time visibility

Built for different data roles

Fabric brings the data together. Power BI turns it into decisions.

Use cases that match the platform.

Unify fragmented data

Modernise reporting

Prepare data for AI

Build predictive models

Analyse operational data in real time

Create a scalable data platform

Customers describing what we are like to work with.

The questions leaders ask us first.

Do we actually need Microsoft Fabric?

Should the report come from the ERP or from Power BI?

Why does nobody trust our dashboards?

Will better reporting fix our data quality?

Do you have reporting outcome figures we can look at?

Tell us which number nobody believes.

Related

When the meeting argues
about whose number is right.