When a client waits weeks for expert dashboards, only to build their own “concept” and asks you to make it look professional.
1. The Visionary Kick-Off Phase: Where Buzzwords Roam Free
Our collaboration began with what management would lovingly call a transformative cross-functional engagement model aimed at orchestrating holistic data value streams.
In simple terms:
“Please build us the data structures, the flows, the KPIs, and the Power BI dashboards that finally show what’s going on in the business.”
Perfect.
Clear.
Actionable.
We gathered the architects, defined governance, drafted templates, whiteboarded data lineage diagrams, and prepared a beautiful, scalable semantic model. We had conventions, naming patterns, mitigation strategies, refresh logic, error-handling routines – everything shiny and ready for a productive journey.
Everyone was enthusiastic. Everyone had ideas.
And, most importantly, everyone was aligned.
Well… everyone on our side, at least.
2. Weeks of Preparation, Zero Customer Feedback
We worked diligently, refining our approach as we waited for the customer to join the process.
But silence is also a form of communication.
Unfortunately, it rarely means “Great job, keep going!”.
It usually means “A surprise is coming.”
Still, we pushed forward.
We designed robust models in line with industry best practices.
We prepared rehearsal dashboards to validate assumptions.
We built reusable templates and processing flows.
We were ready for the big review.
And then the review came.
And it came hard.
3. The Moment of Cognitive Dissonance
The customer appeared in the meeting with the kind of cheerful energy that usually precedes a catastrophe.
“Well, we’ve already started on our side. Here’s the schema we’d like you to follow.”
You know these moments.
That tiny mental pause between “Sure, let us take a look” and
“Oh no… what… what is this…?”
Their model featured:
- Multiple Excel sheets stitched together like a quilt made during an earthquake
- A table named
Data_New_Final_v3_fixed_final2.xlsx - KPIs that contradicted each other within the same visual
- Columns with names like
ValueTemp,ValueTemp2, and the classicTest_OK - Facts stored in dimension tables “for flexibility”
- And a Power BI file that had the structural integrity of warm pasta
None of it aligned with business logic.
None of it reflected real processes.
None of it matched the reality we had analyzed for weeks.
But the expectation?
“Just make it professional and integrate your work into ours.”
Ah yes.
The classic move: Build us a Ferrari, we brought the wheels from a shopping cart.
4. Enter the Heise Parallel Universe (aka: “A Day in the Life of a Software Architect”)
Reading that Heise article was like attending group therapy.
- Meetings that exist purely to create more meetings?
- Teams working for weeks without involving the people who actually own the data?
- Actionism disguised as contribution?
- Architectural decisions shaped by department structures rather than logic?
- People “just starting something” because they feel waiting is unproductive?
Check.
Check.
Check.
And check again.
Conway’s Law says:
“Systems reflect the communication structures that built them.”
And sometimes those structures resemble a bowl of spaghetti thrown at a wall.
The customer’s schema reflected exactly that.
Not out of malice – but out of organizational gravity.
5. The Technical Truth Behind the Satire
Humor aside, here’s the real issue:
You cannot build a solid BI on ad hoc, inconsistent, self-modeled datasets.
Business Intelligence is not:
- “Import everything and see what sticks”
- “We’ll fix it later in Power BI”
- “Visuals first, structure later”
- “Just make it pretty”
A good BI project requires:
- Governance
- Naming conventions
- Semantic modeling
- Data quality rules
- Proper dimensional modeling
- An understanding of actual processes
- And above all: communication
Otherwise, every dashboard becomes a random-number-generator with pretty colors.
6. The Closing Satirical Punch
To move forward productively, we have now established a dedicated working group:
The Taskforce for Harmonization of Divergent Prototyping Schemas.
Its first action item is to evaluate whether we need a
Taskforce for Evaluating the Taskforce for Schema Governance.
We expect the initial PowerPoint deck (approx. 37 pages) to be ready shortly.
Final Thought
Behind the humor sits a lesson many teams need to hear:
Good dashboards require good data.
Good data requires good structure.
Good structure requires collaboration.
And if that collaboration happens before everyone builds their own schema,
we might even avoid founding another taskforce.
PS: Any resemblance to real projects is purely intentional. The humor is optional, the experience unfortunately not.