Interview with Mark K., Director of Operations
A year after introducing a business intelligence system, I can say one thing clearly: it didn’t matter which tools we used. What held us back was everything we hadn’t sorted out before we called anyone in to implement anything.
This isn’t a post about technology. It’s about organization, processes, and — honestly — about us as a management team. Here’s what every experienced BI consultant should tell you at the first meeting, but rarely does.
1. Agree on what you’re actually measuring — and why
Sounds obvious. It isn’t. When we asked our sales director and our finance director to name their key performance indicators for the same area of the business, we got two different answers. Both were legitimate. Neither was aligned.
If your team doesn’t share a common definition of what “a good week” looks like, a BI system won’t fix that — it will just make the confusion more visible and more expensive. Before engaging anyone externally, sit down and define: which decisions matter, what data they require, and how you’ll track progress.
- Agree on the 5–10 metrics management reviews every week
- Align on definitions: what is “revenue”, what is “conversion”, what is “cost per order”
- Assign data ownership — not just who uses it, but who’s responsible for it
2. Know the state of your data — even if it’s messy
This cost us the most time. We had data spread across five systems, some of them inconsistent with each other, and one where fewer than two-thirds of records had mandatory fields filled in correctly. When that surfaced, the project stalled.
Your data doesn’t need to be perfect — it never is. But you need to know what you’re working with. Before any partner conversation, do an internal review: where do your key data sources live, how reliable are they, and who’s accountable when something doesn’t add up.
From the field: Our controller spent three weeks mapping the journey a single data point takes from our CRM to a finance report. That document turned out to be more valuable than any onboarding material we received from the software vendor.
3. Assign internal ownership — genuinely
A BI rollout isn’t a project you hand off and wait to receive. Without someone internal who understands the business, communicates with the implementation team, and is accountable for making sure the reports actually answer the right questions — things get lost in translation.
That person doesn’t need a technical background. Ours was an analyst who knew our processes well and was curious enough to learn the basics. His job was to translate: from “I need to see how sales are tracking by segment” into something the implementation team could actually build. Manager probably isn’t the best person to be the project owner. He is too “polluted” with details – you need a clear perspective that will do the job between the lines that the owner drew.
4. Start small — but start with a real problem
We tried to do everything at once: finance, sales, operations, HR. The result was months of work on something nobody used, because no one knew where to begin.
What works is picking one problem that management feels every single week. For us it was a sales activity view by region — simple, but painful without it. Once we solved that and saw how much faster we could make decisions, conviction in the rest of the project followed naturally.
“The success of that first module wasn’t technical. It was organizational — people started trusting the data and actually changing decisions based on it.”
5. Culture matters more than the tool
A BI system is just infrastructure. What makes it valuable is a culture where decisions are driven by data — not instinct, habit, or what someone heard at lunch. That culture isn’t installed. It’s built.
A year on, the difference is tangible. Problems surface faster, we understand what’s actually driving results, and the hours once spent compiling reports are now spent on interpretation and action. That was the real return — not the software, but the capability it unlocked.