BI is the art of asking data the right question, then answering it on a single page.
The setting: Synerlink
A global manufacturer of filling and packaging machines, Synerlink designs integrated lines for dairy, food and beverages. A Barry-Wehmiller company.
Visit the Synerlink website- 180+ customers worldwide
- 55 customer countries
- 96% of business from exports
- 550 employees
- 9 sites in 5 countries
- 40+ years of expertise
The problem to solve
The ERP, the CRM and the finance and HR tool each have their own database, naming and way of counting. Before any report, this data has to be gathered and reconciled in a Datalake. My job: maintain it, improve it, and deliver the reports end to end.
One dataset, three readings
Choosing the right visual is already answering the question. The same sales tell three different stories depending on what you want to know.
How much did we sell each month?
Fictional data, for illustration only.
My method, report after report
Listen
A colleague brings me a specific problem, or a broader question to explore.
Frame
I gather their needs and pin down the real question, so I can offer the best-suited solution.
Build
I design the report and its visuals together with them, so it stays simple to use.
Check
Final visual tweaks, then a data check: consistency, correct results, nothing missing.
Deliver
The report goes into production and is published to every user concerned.
Measure
Reading the report, an action plan can be decided. That is where it proves its worth.
The toolbox
-
Reporting
Power BI
Interactive dashboards published for teams: filters, exploration, sharing.
-
Data
SQL Server (SSMS)
Storing, transforming and making reliable the tables that feed the reports.
-
Automation
Power Automate
Flows that trigger actions with no manual step, such as sending e-mails automatically.
-
Checks
Excel
One-off exports and verification calculations to check data integrity.
What I take away
-
A report serves a decision
Before the first chart, one question: what will the reader decide by looking at it?
-
Data first
A beautiful visual on wrong data does more damage than a plain table on correct data.
-
Deliver, then listen
A report is never finished on delivery: users are the ones who say whether it is useful.