You're reading one of the top pains on Qlik Community — a free sample of the catalog. Create a free account to browse all Shopify & Atlassian pains and the basic Chrome catalog.
Power BI analysts struggle with low-skill developers and poor data models
.25
Demand — how loud & how many (score)
Willing to payImplicit
Momentum (7d)▼ 0.5%
Cluster37 posts · 27 people
MarketplaceQlik Community
Updated2d ago
The story
The pain
We're drowning in reports built by people who barely know Power BI — bidirectional many-to-many relationships, fact tables filtering other fact tables, no dimension tables, and colossal DAX measures. Management thinks it's drag-and-drop, so they hire cheap cert-holders who churn out broken dashboards that take 10+ seconds to load. Then we get called in to fix the mess, but IT drags their feet on delivering the right data, and leadership does nothing about it.
Evidence
What people said · 37 complaints, 27 people
I blame these 10$ online certs that were like if you know how to CSV into power bi then you're top data scientist because of which we have this influx of inexperienced people hogging every job opening and willing to work cheaper than an experienced individual and management thinking pbi is just drag and drop work.
If there was a list of dumb things to do in PBI, this lad hit bingo as he nailed them all e.g bi-directional many:many, fact filtering fact, no dims, colossal dax measures, page loads that were 10 seconds or more, and so on
He was pulling massive amounts of data and entire tables from sql server that he only needed one column for, because he thought he figured out a workaround to learning sql and could just brute force pbi to give a realistic result or the "right" answer but only for one or two example scenarios. Schema was a mess, tons of calculated columns and clearly thinking dax is just excel formula mechanics.
I was hired in to bridge the gap. Take the stakeholder requirements, I translate them into Medallion Architecture with the end purpose of creating SQL Views for Power BI to ingest and create the data model within the Power BI semantic layer (no more SSAS bullshit)....but IT doesnt like that...so they ignore my requirements take forever to deliver and use their own code to produce the tables I need...doing thus completely redundant work and pushing delivey ETA 3 to 4 months past the original delivery ETA. It's frickin clown show. All I hear is "everyone knows IT is a problem", without ever hearing anything about okay...then what is leadership doing to fucking fix it?
Export the pain — problem, evidence and market signals — as Markdown, PDF or JSON. The Markdown export ships with a built-in briefing you can hand straight to a developer or an AI agent.