Business intelligence consultancy: from scattered reports to numbers you steer on

Business intelligence (BI) is combining and visualising company data so you decide on numbers instead of gut feeling. Reality at many organisations: an export from the shop, a tab from accounting, a dashboard from the ad agency and a feeling that something is off. A business intelligence consultant solves that — not with prettier charts, but with one set of definitions and a source everyone looks at.

What a BI consultant does

A BI engagement almost always has the same four steps:

1. Collect the questions — which decisions do you take weekly and monthly, and which number do you need for them? This is the most important step and the one most often skipped.
2. Unlock sources — shop, CRM, ERP, accounting, ad accounts, email tool, analytics.
3. Model the data — clean, join and define. What is a customer, when is revenue revenue, how do returns and margin count?
4. Report — a dashboard in Power BI, Looker Studio or Metabase, for example, with a fixed set of KPIs and an owner per number.

Definitions matter more than tools

Most data discussions are not about technology but about meaning. Marketing counts sessions and attributed revenue, finance counts invoiced revenue after returns, sales counts pipeline. All three are right — and they still never match.

So first agree:

• Revenue — gross or after returns and discounts, including or excluding VAT and shipping.
• Margin — which costs count: purchasing, shipping, transaction fees, ad spend?
• Customer — per email address, per company or per account? And when is someone a new customer?
• Attribution — how you assign revenue to channels, knowing systems never match exactly. See attribution.

Without those agreements you build a dashboard on which every department applies its own filter.

Start small: one dashboard, five numbers

A full data warehouse is rarely the first step. What works more often: start with one dashboard holding the five numbers you steer on weekly. For a shop: revenue, margin, conversion, average order value and ad cost per order. For B2B: new leads, qualification rate, pipeline, average deal value and cycle time.

If the dashboard works, you expand. If nobody looks at it, you lost something small instead of something big.

Common mistakes

• Unlocking everything technically possible. Eighty metrics is not insight; nobody feels ownership.
• Reporting without a decision. Every number on a dashboard should have an action attached.
• No owner per number. If nobody is responsible for conversion, conversion will not rise.
• Leaving manual work in place. If someone has to merge four exports monthly, eventually it stops happening.
• Forgetting history. Keep raw data, even if you change definitions later. Otherwise you cannot look back.

Tools: what do you pick?

• Looker Studio — free, quick to build, strong on marketing data. Gets sluggish with many sources and heavy calculations.
• Power BI — the standard at many Dutch SME and industrial companies. Powerful model, licence per user.
• Metabase — pleasant when your data already sits in a database and teams want to explore themselves.
• A custom dashboard — sometimes we build a dashboard as an application, for example with Lovable, when you need specific logic or client-facing views a BI tool does not handle neatly.

The tool is the last choice, not the first. Those who start with the tool often build twice.

What we do — and what a specialist does

We set up marketing, e-commerce and customer data daily: measurement, definitions, connections and dashboards teams actually steer on. See also measuring online marketing. For a full data warehouse, heavy ETL work or BI inside production and supply chain we bring in a BI consultant from our network and stay your point of contact.

Frequently asked questions about business intelligence

What is the difference between business intelligence and analytics?

Analytics usually covers one channel or one system, your website data for example. Business intelligence combines sources into a picture of your whole business: marketing, sales, stock and finance in one set of numbers.

Do we need a data warehouse?

Not always. With a handful of sources and modest volumes you can get far with direct connections. Once you want to combine several systems, keep history and fix definitions, a data layer in between becomes cheaper than maintaining separate reports.

Where do you start with BI?

With one dashboard holding the numbers you steer on weekly. If that works and someone actually looks at it, you expand with more sources and possibly a data layer. How much work that is depends mostly on how clean your source data is.

Who needs to be involved on our side?

Someone who knows the processes (often finance or operations), someone with access to the systems, and one owner who decides on definitions. That last role is the most important.

Numbers that contradict each other?

We look at your sources, definitions and reporting with you — and honestly say whether you need a dashboard or something else first.

Get in touch or book a free strategy session.