An agency called us last quarter in a mild panic. Their monthly Power BI bill had quietly grown to €4,200. They had 400 viewers spread across 12 clients. Margins were getting eaten by Microsoft, not by their own delivery costs.
That story isn't rare. It's the standard pattern for BI agencies past the 50-viewer mark and most don't see it coming until they're deep in.
This post walks through what Power BI Pro actually costs at scale, why BI agencies feel it harder than in-house teams, and where the math flips toward Embedded. Concrete numbers, no jargon.
How Power BI Pro pricing actually works
Power BI Pro is licensed per user, per month. At list price that's €14 per user per month. Every person who opens a report, internal, external, viewer, editor, needs a Pro license.
For an in-house team of 20 people, that's €280 per month. Manageable. Predictable.
But Pro licensing scales linearly with every new viewer. There is no volume discount until you cross into Premium territory at 250+ users with capacity-based pricing. Until that line, every additional viewer adds €14 to your monthly bill, forever.
There's a second cost most teams forget: Pro requires a Microsoft 365 (or Power BI Pro standalone) account per user. For external users, your clients, that usually means either:
- They already have Microsoft 365 in their organization, in which case you can invite them as a guest (still needs Pro on their tenant)
- Or you provision accounts for them, which means managing identities, passwords, and offboarding
Both options add admin overhead. Neither is free.
For a complete licensing comparison, see our Power BI Embedded vs Pro vs Premium breakdown — but the short version is below.
The hidden costs beyond the license fee
The €14 sticker price is the visible cost. There are at least four hidden ones that BI agencies pay on top:
1. Onboarding friction per client. Every new external user needs to either get a Pro license through their employer, accept a guest invite to your tenant, or get a provisioned identity. Each path takes time. Multiply by the number of viewers per client, multiply by your client growth rate, and you're looking at meaningful weekly admin overhead.
2. Microsoft branding everywhere. Pro shows reports in Power BI Service. Your client sees Microsoft logos, Microsoft navigation, Microsoft chrome. Your agency's branding shows up nowhere. For agencies whose value is the polished delivery, this directly undercuts your premium positioning.
3. Lost adoption. When users have to log into a third-party tool (Power BI Service) just to see a dashboard, half of them stop opening it. We've seen 60% of Pro licenses go unused at multi-month audits. You pay for a viewer who never views.
4. Compliance and governance overhead. Managing dozens of external guest accounts across clients means managing Azure AD policies, conditional access, and audit logs at a scale Pro wasn't designed for. Time you should be spending on dashboards goes to license management.
None of these show up on the Microsoft invoice. They all show up in your team's calendar and your clients' satisfaction.
What Pro licensing costs at scale (real numbers)
Here's how the math compounds as a BI agency grows. The viewer count is the total across all clients combined.
These are bare license costs. Add Microsoft 365 for external users where needed, and the real number is 20-30% higher.
Now stack that against typical BI agency revenue. An agency billing €5,000 per month per client across 10 clients does €600,000 ARR. If those 10 clients have 30 viewers each (300 total), the Power BI bill alone is €4,200/month, €50,400 per year, or 6% of revenue going to Microsoft just to let clients see reports.
For an agency at 1,000 viewers across all clients, that's €120,000 per year. That's an engineer's salary, paid to license fees.
This is the math that catches BI agencies off-guard at scale.
Why this hits BI agencies harder than in-house teams
In-house teams have a fixed user base, the employees of the company. Growth is bounded. They can size their Power BI investment once and project forward.
BI agencies have a different problem: every new client adds users to the licensing math. Growth in revenue means growth in license cost. The two compound together.
Three patterns make this worse for agencies specifically:
Client-by-client expansion. Each new client onboarding adds a batch of new viewers. Unlike a SaaS company with one shared user pool, an agency runs many isolated user populations, one per client. Per-user licensing scales perfectly badly with this model.
White-label requirement. Agencies sell polished, branded delivery. Pro forces every report to live inside Power BI Service with Microsoft branding. The product you're delivering looks like Microsoft's product, not yours. Premium pricing depends on premium presentation.
Tight margins. Most BI agencies operate on 30-50% gross margins. Licensing costs eat directly into that margin. An agency that pays €50,400/year in licenses needs to justify that against retained client revenue, every renewal cycle.
This is the structural reason BI agencies in particular run into the Pro wall sooner than in-house BI teams. (For the broader pain pattern, see why sharing dashboards becomes a headache as your service business grows.)
The alternative: capacity-based licensing
Power BI Embedded works differently. Instead of paying per user, you pay for compute capacity (Microsoft calls these F SKUs in Fabric). On top of that capacity, you can serve unlimited viewers. The licensing question becomes "how much compute do you need?" not "how many people will look at reports?"
Concretely:
- The smallest Fabric capacity (F2) starts around €180/month
- One F2 can comfortably serve hundreds of viewers across multiple reports
- F4 (~€350/month) handles thousands of viewers
- You scale capacity up only when performance demands it, not when viewer count grows
For an agency with 400 viewers, that's roughly €350/month on an F4 versus €5,600/month on Pro licenses. The savings start showing up around 50-100 viewers and grow exponentially from there.
Power BI Embedded was designed for exactly this pattern: one organization (your agency) embeds reports for many external users (your clients' teams), without each viewer needing a Microsoft license.
The catch: setting up Embedded directly requires building the embed authentication flow, the multi-tenant separation, the white-label portal, the row-level security wiring, and the user provisioning. We covered four architecture patterns for multi-tenancy in a separate post, it's not trivial work. That's why most agencies either stay stuck on Pro (paying the bill) or use a platform like DataTako that handles the Embedded layer for them.
For the deep technical comparison, see our Power BI Embedded in 2026 guide.
How to calculate your real Power BI cost
Here's a quick framework to see where your agency sits today:
- Total viewer count. Add up every person across every client who has access to a Power BI report you've built. Include people who barely use it — they still have licenses.
- Annual license cost. Multiply by €120 (12 months × €14).
- Hidden overhead. Estimate the hours per month your team spends on user provisioning, guest invites, license issues, and support tickets about login problems. Multiply by hourly cost.
- Lost margin. Express the total as a percentage of agency revenue. Anything above 3% is worth a serious look at alternatives.
If you want the math done for you with current pricing and apples-to-apples comparison to Embedded, the Power BI cost calculator gives you a real number in 30 seconds.
For agencies serving multiple clients with branded portals, see how MeerMetData uses DataTako to deliver white-labeled Power BI without per-user licensing. The pattern they describe — branded environment per client, no Pro fees, fast client onboarding — is the alternative path most BI agencies don't realize exists.
When Pro still makes sense
Pro isn't always wrong. Three cases where it's still the right call for a BI agency:
- You have fewer than 25 total viewers and aren't growing fast
- All your viewers already have Microsoft 365 in their organizations and don't mind Microsoft branding
- Your delivery model is project-based (one-off reports) rather than retained client portals
Below 25 viewers, the math favors Pro by a small margin. Above that, the gap widens fast.
FAQ
At what viewer count does Power BI Embedded become cheaper than Pro?
The breakeven sits around 50-100 total viewers, depending on your capacity sizing. Above 100 viewers, Embedded is consistently cheaper. At 500+ viewers, the difference is large enough that staying on Pro costs an engineer's salary per year.
Can I migrate existing Power BI reports to Embedded without rebuilding them?
Yes. Existing reports built in Power BI Desktop can be embedded directly through the Power BI Embedded "App owns data" pattern. No rebuilding, no PBIX changes. DataTako handles this layer specifically so agencies don't have to write embed code themselves.
Do my clients need Microsoft accounts with Power BI Embedded?
No. Embedded uses application-level authentication, which means viewers can log in with any identity system you choose, SAML, Google, email/password, MFA, without a Microsoft account. This is the main reason agencies move to Embedded: clients don't need to onboard into Microsoft to see reports.
What about Power BI Premium?
Premium is a capacity-based model like Embedded, but it includes additional features (paginated reports, AI capabilities, larger model sizes) that most BI agencies don't need to deliver client dashboards. For pure distribution to external users, Embedded is the cleaner fit. See our Pro vs Premium vs Embedded comparison for the detailed feature breakdown.
How long does migration from Pro to Embedded typically take?
For a BI agency with under 10 clients, plan one week of work, mostly configuring multi-tenancy, setting up branded portals per client, and migrating users from Pro accounts to the Embedded auth flow. Reports themselves move over in hours, not days.
Will Microsoft sales tell me about Embedded?
Rarely, in our experience. Pro is per-user recurring revenue — Microsoft's preferred motion. Embedded is a capacity-based one-time setup that doesn't grow with your client count the way Pro does. The math favoring Embedded for agencies is real; the sales motion explaining it is not.
Stop paying per viewer
If your Power BI bill grows every time you onboard a new client, you've hit the Pro wall. The next step is doing the math — accurately, with your real numbers.
Calculate your Power BI savings →
Or see how DataTako works if you want the full picture before you start.
If you're a BI agency specifically and want to see this in action, our BI agencies page walks through the typical client-portal setup. Or book a 30-minute demo for a walkthrough against your specific situation.
