A single Power BI report, built properly on data that is already reasonably clean, is a low-thousands job in the UK. A full reporting system that several people depend on, with a proper model underneath and documentation to hand over, is around £9,500 at my published rate. A one-off report converted from an existing Tableau, Qlik or SSRS report is £600.
Those numbers are the easy part. The reason quotes for "a Power BI dashboard" come back ranging from £1,500 to £50,000 is that the dashboard is rarely the work. The work is the state of your data, the number of definitions nobody has agreed yet, and how many people have to sign off on what a customer is.
This post is about project-shaped pricing: what you get for each band, and what pushes you from one to the next. If you want the day-rate view instead (what UK Power BI people charge per day and why), that is covered in what Power BI work costs in the UK.
What do the different project shapes cost?
| Project shape | What it includes | Published anchor |
|---|---|---|
| One hour on your data | Free dashboard running on your figures, questions answered | £195 fixed |
| An existing estate reviewed | 48 checks across six areas, scored, prioritised, costed | £950 fixed |
| One report converted from another tool | Like-for-like rebuild, numbers reconciled to the original | £600 per report |
| A full reporting system | Model, reports, documentation, handover, training | £9,500 |
| Ongoing changes and monitoring | Rolling monthly, no minimum term | From £250/month |
Those are my own published prices, which is why I can print them. Everything on that list is fixed or per-unit rather than "it depends", because I would rather compete on being predictable than on being cheapest.
For comparison against the market: contractor day rates for Power BI in the UK had a median of £490 with a range of £320 to £656 across 1,237 quoted rates in the six months to 26 August 2026, according to ITJobsWatch. That gives you a rough conversion. A £9,500 system build is somewhere around nineteen days at the market median, which is a useful sanity check in both directions when you are comparing quotes.
Why do quotes for the same thing differ so wildly?
Because "a dashboard" describes the output, not the job, and two suppliers can hear completely different jobs behind the same sentence.
Ask three firms to quote for "a sales dashboard" and one will price a single report page against a spreadsheet you email over, one will price a proper model built from your ERP with a scheduled refresh, and one will price a six-month programme including a data warehouse. All three answered the question you asked. Only one of them matches what you had in your head, and it is your job (or a decent supplier's) to work out which before anyone quotes.
The fix is not to get more quotes. It is to define the job more tightly, which is what the briefing questions are for.
What actually drives the cost up?
In rough order of how much they move the number.
The state of your data. This is the big one and it is not close. If your figures live in one system, exported cleanly, with consistent codes, the build is fast. If they live in four systems that disagree about customer names, with a spreadsheet in the middle that someone maintains by hand, most of the project is reconciling that, and the report at the end is the small bit. This is why a fixed quote given before anyone has seen your data is a guess wearing a suit.
Unagreed definitions. Every business has a handful of terms nobody has ever formally settled: active customer, net revenue, on-time delivery, headcount. It never matters until a report has to print one number for it, and then it matters enormously, because the answer needs a decision from someone with the authority to make it. Projects do not stall on technology, they stall waiting for that ruling. Every week of waiting is a week on the invoice.
The number of audiences. One report for one team is one design conversation. The same report for finance, operations and the board is three, and they want different things at different levels of detail. Cost scales with audiences far more than it scales with pages.
Access. The number of days lost to waiting for credentials, a gateway, a firewall rule or an IT ticket is genuinely one of the largest hidden costs on a small project. It is also the one you have the most control over.
What brings the cost down?
Starting smaller than feels natural. A single report answering the two questions the business actually asks, built on a model that can grow, will teach you more about what you need than six months of requirements gathering. It also produces something people use in weeks rather than quarters, which is what keeps the budget for phase two alive.
Bringing clean inputs. If you can export one reliable extract instead of granting access to five systems, the first version gets cheaper and faster. You can always connect properly later.
Deciding your definitions before the build, not during it. Half an hour with the right person in the room, writing down what "active customer" means, saves days later.
Buying an assessment before a build. It sounds like an extra cost, and for a straightforward job it is. For anything with unknowns in it, spending £950 to find out what you are dealing with beats discovering it halfway through a £9,500 build, because the discovery does not stop the clock.
Is a fixed price or a day rate better for a build?
Fixed price protects you when the scope is genuinely knowable: an audit, a report conversion, a defined system with agreed definitions. Day rate is more honest when nobody yet knows what is in the data, because a fixed price against unknown data is either padded to cover the worst case or destined for a change request.
The pattern I use, and would recommend regardless of who you buy from, is to fix the price of finding out and then fix the price of the build once you know. That is why the assessment is a separate fixed-price product rather than free discovery baked into a bigger number. Free discovery is never free, it is just priced into the build where you cannot see it.
What should be included that often isn't?
Check for these before signing, because their absence is what makes a cheap quote cheap.
Documentation of the model, in language a non-specialist can follow. Ownership of the source files, in writing. A refresh that runs on a schedule with someone watching it. A handover session recorded, not just delivered live once. Training for the people who will maintain it, or an honest statement that they will need support.
A build that arrives without those is not finished, it is transferred. The handover checklist covers what to insist on in more detail, and it is worth reading before you sign rather than after.
What does it cost to keep running?
Reports are not a one-off purchase, which is the thing most first-time buyers underestimate. Source systems change, someone adds a product category, a credential expires, and the report either breaks loudly or, worse, quietly goes wrong.
Budget something for the year after go-live. My own care plans start at £250 a month for monitoring and diagnosis and run to £1,500 for a standing development relationship, all rolling with no minimum term. If you would rather not commit, at minimum work out who in your business will notice when a refresh fails, and make sure they know it is their job.
Common questions
How much does a simple Power BI dashboard cost?
If the data is already clean and in one place, and the requirements genuinely are simple, a single report is a low-thousands job. Below about a thousand pounds you are usually buying someone's time to connect a spreadsheet to a chart, which can be exactly right for a proof of concept and is rarely right for something a business will depend on.
Is it cheaper to hire a freelancer or a consultancy?
Freelancer day rates are typically lower, with the UK median at £490 a day in the six months to August 2026. Consultancies cost more per day and bring continuity, cover and a wider skill set. Which is cheaper overall depends on how long you need the relationship and what happens if one person is unavailable. The freelancer versus consultancy comparison works through the trade-offs.
Why won't anyone give me a fixed price?
Usually because they have not seen your data and know that the state of it drives most of the cost. That is a defensible position, though the better answer is to fix the price of a short assessment and then fix the build price once the unknowns are gone. If someone refuses to fix any price at any stage, that is worth asking about.
How long does a Power BI dashboard take to build?
Days for something focused on clean data, weeks for a proper reporting system, longer if the data needs work first. Build time is rarely the constraint. Waiting for access, decisions and definitions is. The project timeline post covers typical durations by project shape.
Do I need a data warehouse first?
Usually not, and being told you do before anyone has looked at your data is a flag worth noticing. Plenty of businesses run good reporting straight from their source systems for years. A warehouse becomes the right answer when you have several systems that must agree, or volumes that make direct queries painful, not by default.
What is the most common way this goes over budget?
Discovering mid-build that two systems disagree about something fundamental, like which customers exist. It stops the build, needs a business decision, and often needs data fixing at source. Finding that out during an assessment costs a few hundred pounds. Finding it out in week three of a build costs the build.