Before you sign a Power BI support contract, get clear answers to four things: who personally answers when you raise something, whether you are buying a bucket of hours or the right to ask, what happens to time you do not use, and how you exit. Everything else, including the monthly figure, is easier to compare and matters less.
I run support plans, so this is an interested view. It is also the checklist I would use as a buyer, and several of the traps below are ones I have deliberately designed my own plans to avoid, which I will say plainly rather than pretend to neutrality. If you want the pricing comparison instead, published UK support prices are collected in what a Power BI support retainer costs.
The questions, and what a good answer sounds like
| Question | Good answer | Warning sign |
|---|---|---|
| Who actually does the work? | A named person you have met | "One of our consultants" |
| Hours or availability? | Clearly stated, either way | Vagueness that resolves in their favour later |
| What happens to unused time? | Stated plainly in the contract | "It rolls over" without a cap or expiry |
| What counts as support vs new work? | A worked example of each | "We'll be reasonable about it" |
| Response time commitment? | A number, with what it's measured from | "As soon as possible" |
| How do I leave? | Rolling, or a stated notice period | 12-month tie-in with auto-renewal |
| What do I get if nothing breaks? | Something proactive, or a lower tier | Nothing, and they see no problem with that |
| Who owns the files and workspaces? | You, in writing | Anything that isn't a clear "you" |
Who actually answers?
The most important question and the one people forget, because it feels rude.
In a small firm the answer is usually the person selling to you, which is what you want. In a larger firm the person in the meeting is frequently not the person who will respond, and that is not automatically a problem, but you should know it before rather than after.
The follow-up that gets the real answer: what happens when that person is on holiday? Every arrangement has a gap somewhere. A supplier who describes theirs honestly ("my colleague covers, here is what she can and cannot pick up") is more trustworthy than one who implies there is no gap at all.
Hours or availability?
This is the structural question, and the two models behave very differently on the day you need them.
An hours-based retainer sells you a number of hours per month. It is easy to compare on price and creates a specific bad incentive: you hesitate before raising something small because you are spending the allowance. Support you are reluctant to use is not worth much, and the hesitation lands hardest on exactly the small early warnings you would want raised.
An availability-based arrangement sells you the right to ask, with a limit on how much can be in progress at once rather than on total hours. You can raise as many things as you like; the constraint is how many are being worked on simultaneously. That removes the hesitation, which is the point.
My own plans work the second way, and I will explain the trade-off honestly rather than just advocate for my model. Availability-based pricing is harder to compare against competitors on a spreadsheet, and it needs trust on both sides that neither party will take advantage. Hours are more transparent. What I think hours get wrong is that they optimise for the wrong thing: the value of support is that problems get raised early, and any pricing that discourages raising them is working against its own purpose.
Either model can be run fairly. What you should not accept is not knowing which one you have bought.
What happens to time you do not use?
Ask directly, and get the answer in the contract rather than in an email.
Common arrangements: unused time expires monthly, rolls over with a cap, or rolls over indefinitely. Expiry is the most honest of the three even though it sounds least generous, because it prices what you are actually buying, which is availability rather than a stockpile.
Indefinite rollover is where the traps live. You accumulate a large balance across a quiet six months, then discover the balance expires on renewal, or can only be spent at a certain rate, or does not survive a change of plan. The pattern to watch for is a growing balance that feels like an asset and turns out to be a marketing device.
What counts as support and what counts as new work?
Every support arrangement has this boundary, and it is the most common source of disagreement.
Fixing a broken refresh is support. Building three new reports is a project. Between them is a wide grey area: adding a column to an existing report, changing a definition, adding a new source to an existing model.
Do not accept "we will be reasonable". Ask for a worked example on each side of the line. A supplier who has thought about this can describe the boundary specifically. One who has not will describe it vaguely, and the vagueness will be resolved in their favour when there is money attached.
How do you leave?
Ask before you join. It tells you a great deal about how the relationship is designed.
Rolling monthly with no minimum term is the most buyer-friendly arrangement, and it is what I offer, from £250 a month at the monitoring tier up to £1,500 for a standing development relationship. The commercial logic is straightforward: if I have to keep earning it every month, the incentive to stay useful is permanent. A twelve-month tie-in produces the opposite pressure once the contract is signed.
Long terms are not automatically wrong. A supplier committing dedicated capacity has a fair reason to want commitment in return. What should concern you is a long term combined with automatic renewal and a short notice window, which is a structure designed around inertia rather than value.
Also ask what leaving actually involves practically. Do you get your files, your documentation, your workspace access, cleanly and without a final invoice for the privilege? The handover expectations apply at the end of a support relationship just as much as at the end of a project, and the patterns to avoid are in who owns your Power BI estate.
What do you get in a month when nothing breaks?
If the honest answer is nothing, you are buying insurance and should price it as insurance, at the cheapest credible option.
Better arrangements include something proactive: monitoring that catches a failing refresh before anyone notices stale numbers in a board pack, a periodic review of what has changed, a scored report on the state of the estate. That is what makes a quiet month worth paying for, and it is why my lowest tier is built around monitoring rather than around a small number of hours.
The failure this guards against is worth naming because it is the expensive one: a refresh that fails silently and serves last week's figures without any error. Nobody raises a ticket, because nothing looks broken. The causes are covered in why your Power BI refresh keeps failing, and it is the strongest argument for paying something rather than nothing.
What is the market rate?
For context on whatever you are quoted, published UK support prices as of August 2026 cover a wide range. Independent consultants publish retainers from a few hundred pounds a month. Larger suppliers list managed Power BI support on public procurement frameworks running to several thousand a month. My own care plans are £250, £750 and £1,500 a month, rolling.
That spread is real and mostly reflects what is included, the size of the estate covered, and whether you are buying an individual or an organisation with formal cover. A £250 plan and a several-thousand-a-month plan are not competing for the same buyer, and comparing them on price alone tells you almost nothing.
Common questions
How much does Power BI support cost per month?
Published UK prices range from around £250 a month for monitoring-focused plans up to several thousand a month for enterprise support from a large supplier, as at August 2026. What drives the difference is the size of the estate, the response commitment, and whether you are buying an individual's time or an organisation's cover.
Should I buy support hours or a retainer?
Hours are easier to compare and discourage you from raising small things, which is where support earns its keep. Availability-based arrangements remove that hesitation but are harder to compare on price. Either is defensible; what matters is knowing which you have bought and what happens to anything unused.
What should a Power BI support contract include?
At minimum: a named person or a stated cover arrangement, a response commitment with a number attached, a clear boundary between support and new work with worked examples, stated treatment of unused time, defined exit terms, and written confirmation that you own the files and workspaces.
Do I need Power BI support if nothing is broken?
The main risk is not visible breakage, it is a refresh failing silently and serving stale numbers that nobody questions. If someone in your business is genuinely watching for that and knows what to do, you may not need a contract. If nobody is, monitoring is the part worth buying even at the lowest tier.
What is a fair notice period for Power BI support?
Rolling monthly is the most buyer-friendly and is available in the market, so a long tie-in should come with a specific justification such as dedicated committed capacity. Be particularly careful with long terms combined with automatic renewal and a short cancellation window.
Can I get support from someone who didn't build the system?
Yes, and it is common. Expect a paid familiarisation exercise first, since supporting an unfamiliar model without understanding it is guesswork. A supplier who offers to take it on with no review at all is either very confident or not planning to look closely until something breaks.