Scoping & cost
What actually drives the cost of a Power BI implementation
We do not publish day rates or package prices, and we would be suspicious of any consultancy that does — a number quoted before anyone has looked at your data is a guess dressed up as a proposal. What we can do is be completely transparent about what makes a Power BI or Microsoft Fabric engagement large or small, so you can size the problem yourself before you speak to anyone.
Six cost drivers
The variables that move the number
In practice, two projects that look identical on a brief can differ several-fold in effort. These are the reasons why.
How many source systems — and how clean they are
One tidy SQL database is a fundamentally different engagement to an ERP, a CRM, three spreadsheets maintained by different people and a SaaS platform with a rate-limited API. The number of sources matters less than their condition: undocumented tables, inconsistent customer codes and no reliable date of record are what actually consume days.
Depth of ERP integration
Reading a handful of tables out of Sage 200 Evolution is straightforward. Reconstructing accurate financial logic — cost of sales, inventory valuation, multi-company consolidation, period locks — is a build in its own right, because that logic lives in the ERP's application layer rather than its tables.
Refresh frequency and latency
A dashboard refreshed overnight is inexpensive. Near-real-time reporting means incremental refresh, change data capture or Direct Lake on Fabric, and infrastructure that has to be sized and monitored. Ask for real-time only where a decision genuinely depends on it — this is the single most common place budget gets spent for no operational gain.
Semantic model complexity
The number of reports is a poor predictor of effort. The number of distinct business questions is a good one. Time intelligence, budget-versus-actual, allocations, slowly changing dimensions and multi-currency each add real modelling work — and each one, done properly, makes every subsequent report cheaper.
Governance, security and compliance
Row-level security by branch, region or sales rep; workspace and deployment-pipeline structure; POPIA-conscious handling of personal data; audit requirements. Governance designed in from the start costs a fraction of governance retrofitted after the business has already built on top of an ungoverned model.
Handover model versus ongoing support
Do you want a build handed to an internal team with documentation and training, or a platform we continue to operate and evolve? Both are legitimate. They price and staff very differently, and deciding which one you want early keeps the engagement honest.
How we scope
From first conversation to fixed-scope proposal
01
Free scoping session
A working session, not a sales call. We map the questions the business needs answered, the systems that hold the data, who consumes the output and what currently happens manually. No cost, no obligation.
02
Data reality check
We look at the actual sources. This is where the surprises live — and where a quote either becomes trustworthy or becomes fiction. We would rather find the broken customer master now than three weeks into a build.
03
Fixed-scope proposal
You get a written scope: what gets built, what data it uses, what it does not cover, the delivery timeline and the commercial terms. Fixed fee against a fixed scope, so the risk of the unknown sits with us rather than with you.
04
Build in increments
First useful dashboard early, then expand. You see working output within weeks rather than at the end, which means scope changes get caught while they are still cheap.
What a first Power BI engagement usually covers
Most South African businesses we work with start narrow and deliberately so: one department, one clearly painful reporting problem, one or two source systems. That first engagement typically includes source connection and a data-refresh pattern, a semantic model with documented measures, one or two production dashboards built for how the audience actually reads them, row-level security where it is needed, and handover documentation with a training session.
What it usually does not include — and should not, on a first engagement — is a full Microsoft Fabric lakehouse footprint across every system in the business. That is a second-phase conversation, once the first build has proved which data the business actually acts on.
Why not just publish a price list?
Because the honest range is wide enough to be useless, and the narrow number is only achievable by quietly cutting scope. We would rather spend an hour understanding your situation and give you a figure we can stand behind in writing.
What about Power BI licensing?
Licensing is separate from implementation and paid to Microsoft, not to us. Per-user, Premium Per User and Fabric capacity have very different economics depending on your audience size. We model that with you during scoping so the running cost is known before the build starts, and we have no reseller margin either way.
Do you quote fixed fee or time and materials?
Fixed fee against a fixed scope for build work. Ongoing support and evolution are handled separately, on a retainer sized to how much change you actually expect.
Is the scoping session really free?
Yes, and there is no obligation attached to it. If we conclude your problem does not need us, we will tell you that — it is a cheaper outcome for both of us than a badly matched project.
Want a number you can actually budget against?
Tell us what you need to see and which systems hold the data. We'll come back with a scope and a fixed fee.