Power BI Licensing UK: A Plain English Guide

Power BI licensing is easy to get wrong. Microsoft rearranges the menu every couple of years, and the sales conversation almost always steers towards the top of the range. This is the plain English version, aimed at a UK finance manager, IT lead or head of data trying to work out what to actually buy.

The three licence tiers

For most UK businesses the choice comes down to three tiers.

• Power BI Pro. A per user licence, currently around ten pounds a month. Every builder needs one, and every viewer of shared content needs one unless the workspace sits on a Premium or Fabric capacity.
• Power BI Premium Per User. Around twenty pounds a month per user, includes Pro features plus larger dataset sizes, more refreshes per day and paginated reports. Still a per user licence, so every viewer needs one.
• Fabric capacity, historically Power BI Premium capacity. A capacity based licence starting at roughly two hundred and fifty pounds a month for the smallest F2 SKU, and rising quickly from there. Viewers on a workspace hosted on a Fabric capacity only need a free Fabric licence, not a Pro licence.

When Pro is enough

Pro is the right answer for the majority of UK small and mid market businesses. If the total number of people who need to view content is under a couple of hundred and none of the datasets are enormous, Pro is cheaper than any capacity SKU. The maths tips the other way somewhere around two hundred and fifty active viewers, though the exact break even depends on the Fabric SKU you would move to.

When Premium Per User makes sense

Premium Per User is a good fit for a small analyst team that needs the Premium features but does not want a capacity. It gets you paginated reports, deployment pipelines, XMLA endpoint, larger dataset sizes and more refreshes per day, without the commitment of a capacity SKU. It is a per user licence, so it does not solve the viewer licence question.

When Fabric capacity earns its keep

Fabric capacity is worth the money when one of three things is true. First, your viewer count is large enough that the per user Pro licences add up to more than the capacity. Second, you need the Fabric workloads beyond Power BI, such as OneLake, Data Factory pipelines, notebooks or a warehouse. Third, you have very large models that cannot fit inside the ten gigabyte Pro dataset limit even after tuning.

For most UK mid market businesses arriving at this point, an F8 or F16 SKU is the sensible starting size. F2 and F4 are cheap but tend to be slow in practice, and F32 upwards is often oversized for the actual workload.

Power BI Embedded

Power BI Embedded is a separate SKU family, priced hourly, aimed at software vendors embedding reports into their own product. It is unlikely to be relevant unless you are building a customer facing product. We cover this in more detail on the Power BI Embedded page.

The viewer licence trap

The single most common licensing mistake in UK businesses is forgetting that viewers of Pro workspaces also need a Pro licence. Somewhere around the point that a small pilot rolls out to the wider business, someone realises that the two hundred people who now want to see the dashboard each need a ten pound licence. That is the moment the Fabric capacity conversation becomes real, and it should be planned for early rather than sprung on a finance director.

Practical rules of thumb

• Under fifty viewers and a small analyst team. Pro across the board.
• Fifty to two hundred viewers with straightforward reporting. Pro across the board, with an eye on the viewer count.
• Two hundred to five hundred viewers. Run the numbers on Pro versus F8 or F16 Fabric capacity. Fabric usually wins on cost and adds room to grow.
• Five hundred plus viewers, or large datasets, or Fabric workloads beyond Power BI. F16 or F32 Fabric capacity.
• Embedding in a customer product. Power BI Embedded, sized to the peak concurrent load.

What to actually do next

Before you buy anything, do a short viewer count exercise. List every group of people who will need to see a report, roughly how many are in each group and how often they will look. That number, combined with a rough steady state dataset size, is enough to make the Pro versus Fabric decision honest.

If you would like a hand doing that, our Power BI audit page covers how we run licensing assessments as part of a wider review, and the Power BI pricing guide covers the total cost picture in more detail.

Frequently asked questions