Production and quality on one screen, from machine data and SharePoint.
At a 140-person manufacturer, machine output logs lived in SQL, quality checks in SharePoint lists and downtime notes in a notebook by the line. Monthly production reviews argued about whose numbers were right — instead of deciding what to fix.
The review argued about the numbers, not the fixes.
The plant was not short of data — it was short of agreement. Machine output logs landed in a SQL database, shift after shift, exactly as the machines reported them. Quality checks were recorded separately, in SharePoint lists maintained by the quality team. And downtime — the thing everyone most wanted to explain — lived in a notebook next to the line, in whatever shorthand the shift that wrote it happened to use. Each source was reasonable on its own terms. Nothing had ever been built to make them agree.
The gap showed up wherever the sources met. Output as SQL counted it and output as the quality lists implied it did not reconcile cleanly; downtime existed in the notebook but nowhere the machine data could see it. So anyone preparing for the monthly production review pulled their own extract, applied their own assumptions and walked in with a version of the numbers that looked authoritative — right up until it met someone else's version across the table.
That turned the monthly production review into a reconciliation meeting. Instead of asking what to fix — which line, which product, which recurring stoppage — the room spent its energy on whose spreadsheet was closest to reality. Fixes that needed a shared baseline could not start from one, because there was no shared baseline: production and quality were describing the same plant from records that had never been joined.
One certified model — and a dashboard the whole room trusts.
We did not add another source of truth to the argument. We took the sources the plant already had — the SQL machine data and the SharePoint quality lists — made them agree in one certified Power BI model, and put the answers on one screen.
Machine output logs from SQL and quality checks from SharePoint lists were combined into a single Power BI model — so production and quality data finally live in the same place, joined on the same lines, products and time periods, and can be compared directly instead of argued about.
The model is certified: one agreed definition of output, scrap rate and downtime, endorsed as the plant's official numbers. When a figure appears in a review now, there is exactly one place it can have come from — which is what ends the debate before it starts.
On top of the model sits the production dashboard: OEE and scrap rate up front, with drill-through on downtime — from the plant-level picture down to the line and the reason behind a stoppage. Reviews start at the answer, then go straight to what to fix.
Quality measures now watch themselves. When a metric drifts past its threshold, an alert goes out to the people who can act on it — so drift gets caught between reviews, while there is still time to correct it, instead of being discovered in the monthly meeting.
What changed.
No vanity metrics — these are the three numbers the engagement was measured on.
Which numbers does your team still argue about?
Book a free 30-minute audit — we'll look at where your data actually lives, pick the report your meetings fight over, and show you what one trusted dashboard would take on the Microsoft licenses you already pay for.