A side-by-side comparison of hiring Power BI skills, buying them from a consultant, or combining the two, and how to decide.
In short
Use a Power BI consultant when you need architecture, a fast rebuild or skills you lack, and an in-house team when reporting is continuous and close to daily decisions. Most mid-sized companies do best with a hybrid: internal owners of models and priorities, and a consultant for the data platform, peaks and coaching, with a planned handover.
Power BI is easy to start with and hard to scale. One analyst builds a useful report, ten more follow, and two years later finance and sales disagree about revenue and nobody knows which semantic model is the right one. At that point leaders ask whether to hire a Power BI consultant (Power BI-konsult) or build an internal team. This comparison from Devies, a Gothenburg-based product and engineering consultancy, is for CFOs, CIOs and data leads making that call.
What does a Power BI consultant do that an in-house team does not?
A Power BI consultant brings experience from many implementations, so they recognise patterns, such as model design mistakes, licensing traps and governance gaps, that an internal team meets for the first time. An in-house team brings what no consultant can: daily contact with the business, knowledge of how numbers are actually used and continuity over years.
The work behind a reliable reporting estate is broader than building reports. It includes:
- Data engineering: pipelines from ERP, CRM and other systems into a warehouse or lakehouse.
- Semantic modelling: star schemas, relationships, DAX measures and row-level security in shared semantic models.
- Report design: dashboards that answer the questions leaders ask.
- Administration: workspaces, capacity, deployment pipelines, refresh schedules and access.
- Adoption: training, support and a community of report builders.
Few companies need all of these full time. That is the real reason the choice is rarely "consultant or employee" and more often "which parts do we own".
How do a consultant and an in-house team compare?
The table compares the three common set-ups across the criteria that usually decide the matter.
Criterion
Power BI consultant
In-house team
Hybrid
Time to start
Power BI consultant
Weeks
In-house team
Months to recruit and onboard
Hybrid
Weeks, then gradual handover
Breadth of experience
Power BI consultant
High, across many clients
In-house team
Depends on hires
Hybrid
High at first, growing internally
Business knowledge
Power BI consultant
Must be learned for each client
In-house team
Deep and growing
Hybrid
Held internally
Cost profile
Power BI consultant
Variable, higher hourly rate
In-house team
Fixed salaries, lower marginal cost
Hybrid
Fixed core plus variable peaks
Continuity
Power BI consultant
Risk when engagement ends
In-house team
Risk when key people leave
Hybrid
Shared, if documented
Architecture and platform skills
Power BI consultant
Usually strong
In-house team
Often thin in small teams
Hybrid
Bought where needed
Governance and standards
Power BI consultant
Can set up quickly
In-house team
Must be built over time
Hybrid
Set up together, owned internally
Best for
Power BI consultant
Rebuilds, audits, platform work
In-house team
Continuous reporting and support
Hybrid
Most mid-sized companies
Licensing affects the decision too. As of October 2026, Microsoft's Fabric licensing documentation states that on Fabric capacities smaller than F64, every user who views Power BI content needs a Pro or Premium Per User licence, while on F64 or larger, users with a free licence and a viewer role can view content. Someone on the team, internal or external, must own those choices, because they change both cost and architecture.
When should you choose a Power BI consultant?
Choose a Power BI consultant when the problem is structural, time-bound or requires skills you do not need permanently.
Choose a consultant when
- Reports contradict each other and you need someone to find out why; our article on why Power BI shows different numbers lists the usual causes.
- You are moving to a new data platform, such as Microsoft Fabric, or redesigning the data model.
- The original developers have left and nobody understands the pipelines.
- You need an independent review of security, licensing or performance.
- You want an internal team but need someone to set standards and coach it first.
Devies met the inherited-platform case at Stena Line. When the original development team left, Devies first mapped how data moved into the warehouse and where the business rules lived, then stabilised reporting, including restoring NPS reporting, before optimising pipelines and migrating to a medallion architecture on Azure. The case reports roughly a 25 percent reduction in pipeline runtime.
When should you build an in-house Power BI team?
Build in-house when reporting is a continuous part of how the business runs and the main work is close to users. Signs that you are there:
- New report requests arrive every week, and most need business context rather than new architecture.
- Several departments rely on shared semantic models that need an owner who attends their meetings.
- You have a stable data platform and need people to operate and extend it.
- You want report builders in the business, supported by a small central team.
Microsoft's Fabric adoption roadmap describes three ways to own content: business-led self-service, managed self-service and enterprise BI, usually supported by a Center of Excellence. The model you pick tells you which roles you need. Managed self-service, where a central team owns shared semantic models and the business builds reports on top, suits most mid-sized companies and needs only two or three internal specialists.
For hiring, the Microsoft Power BI Data Analyst Associate certification (exam PL-300) is a reasonable baseline for report builders, covering data preparation, modelling, visualisation and security. It says little about platform engineering, so test that separately. If you are unsure which data roles you need at all, compare data science, data engineering and data analytics.
How does a hybrid model work in practice?
A hybrid model gives ownership to internal staff and uses a consultant for defined parts of the work, with knowledge transfer written into the plan. It works when responsibilities are explicit:
- Internal: a product owner for reporting, owners of each shared semantic model and the people who talk to the business every week.
- Consultant: platform architecture, pipelines, performance work, migrations and an agreed share of the build.
- Shared: standards for naming, measures, version control and deployment, written down once and applied by both.
- Handover: pairing on real work, documented models and a date when each area moves fully in-house.
Define every important measure once, in a shared model, so the next person does not rebuild it. A semantic layer with one definition per metric is what keeps a mixed team consistent.
Review the split once a year. If the consultant still does most of the routine changes after twelve months, the handover has stalled: either the internal team is too small, or the work was never documented well enough to move. If the internal team keeps postponing platform work, the opposite is true, and a short, bounded engagement for that part is cheaper than letting the debt grow.
Keep reading
From our work
Client Case: Stena LineStena Line modernised its Azure data platform - improving Power BI reliability, reducing pipeline runtimes and scaling with medallion architecture.Read the caseDevies
Want a second opinion?
Devies, a Gothenburg-based product and engineering consultancy, works on questions like this with Nordic product and IT teams. Tell us where you are and we will suggest a sensible first step.
Frequently asked questions
How much Power BI work can one person handle?
One experienced developer can usually maintain a handful of shared semantic models and a steady flow of report changes, but not also run pipelines, administration and user support. Once several departments depend on Power BI, plan for at least two people with overlapping knowledge, so holidays and resignations do not stop reporting. Add platform engineering skills internally or from a partner as data volume grows.
Is it cheaper to hire a Power BI developer than a consultant?
Per hour, yes: a salaried developer costs less than a consultant. Over a year, it depends on how steady the workload is. A consultant is cheaper for a three-month rebuild than a permanent hire who is under-used afterwards. An employee is cheaper for continuous work. Include recruitment time, onboarding and the cost of losing the only person who understands the models.
What should a Power BI consultant hand over at the end?
The consultant should hand over semantic models in version control, documentation of every shared measure and its business definition, data source and refresh dependencies, workspace and access design, deployment steps and a list of known issues. Ideally your staff have already made changes to the models under supervision before the engagement ends, so the handover is tested rather than only written.
Do we need Microsoft Fabric to use Power BI?
No. Power BI can run with Pro or Premium Per User licences without a Fabric capacity. Fabric capacity becomes relevant when you want lakehouses, warehouses, notebooks or pipelines inside the same platform, or when many people only view reports: on F64 or larger, viewers with a free licence and a viewer role can access Power BI content, according to Microsoft's licensing documentation.
