Business intelligence engagements are sold on the dashboard and spent on everything behind it. The visible deliverable is a set of charts; the work is finding where the numbers live, reconciling versions that disagree, agreeing what each measure actually means, and building something that keeps refreshing after the consultants leave. Buyers who price the dashboard and discover the rest are the reason this market has a reputation for overruns. This page covers where the cost really sits and how to scope a first engagement that pays for itself.
Data visualization services come last: most of the work is upstream of the chart
A dashboard is the last ten per cent. Before it come extracting data from systems that were never designed to share, reconciling records that disagree, handling history when a customer changes name or a product is recategorised, and agreeing definitions so that two departments reporting revenue produce the same number. Any proposal that prices visualisation without pricing that work has priced the part you can see. Ask explicitly what happens when two source systems disagree, because the answer describes the real engagement.
Definitions are a governance decision, not a technical one
The hardest question in most BI projects is what a measure means: which date counts, what is excluded, how refunds and cancellations are treated. Those are business decisions and they need someone with authority to make them. Consultants can facilitate; they cannot decide. Identify that person before the engagement starts. Projects without one produce technically correct dashboards that nobody trusts, which is the most common failure in this market and the most expensive to discover late.
Scope a first slice that answers one question
The engagements that succeed start with one decision somebody actually makes regularly and build the smallest thing that informs it, end to end, including the refresh. That produces working evidence in weeks, exposes the data problems early while the budget is intact, and creates internal appetite for the next slice. Programmes that begin by modelling the whole business spend months before anyone sees a number, and are cancelled at the point where the value would have started.
Plan for who runs it afterwards
Reports break when a source system changes, and something has to be done. Decide before the engagement whether your own team will own the result, whether the consultancy will support it and at what cost, and what documentation and access transfer at the end. Ask what training is included and who receives it. A BI deployment nobody inside the company can change is a subscription to the consultancy, whether or not that is what appeared on the proposal.
Questions people ask about business intelligence consulting services
What should a business intelligence service include?
Typically discovery of data sources, extraction and modelling, agreeing definitions with the business, building reports or dashboards, and setting up refresh and access. How much of the middle is included is the main difference between quotes, and it is where most of the cost sits.
How do BI service providers price an engagement?
Usually time and materials for discovery and modelling, because the size only becomes knowable once someone has looked at the data, then a fixed price or a retainer for delivery. Be cautious of a fixed price given before anyone has seen the source systems.
Do we need a data warehouse?
Not always. Small organisations with two or three clean sources often do not. You need one when the sources disagree, when history has to be preserved as records change, or when reporting queries interfere with the systems that run the business.