Three teams, one output, three different numbers - and none of them wrong. In a recent conversation, a customer asked why their production figures never match. Operations had one, finance another, management reporting a third. It looked like a reporting problem. It was a data-integration problem. The instinct is to hunt for the “right” figure and correct the rest, but that hunt never ends - each number is faithful to the system that produced it. The real question is not which one wins, but why the same word means three different things across the business.
In Brief
- Operations, finance and reporting routinely show three different figures for the same output - each correct in its own system.
- The cause is structural: different sources, timings, definitions and manual adjustments - not careless people.
- Gartner puts the cost of poor data quality at an average of $12.9M per year, much of it spent reconciling numbers.
- The fix is one shared, traceable definition every system links back to - not forcing everyone onto one identical view.
Part of the series , The Industrial Data Foundation
- Why your production numbers never match - you are here
- When reporting takes days (KPIs)
- When Excel becomes enterprise software
- From data silos to operational context
- Traceable industrial data
- Data foundation before advanced analytics
- AI readiness in industry
The setting
Everyone trusts their own number
The customer wanted to compare production performance across the business. Operations read quantities from the plant. Finance worked from posted ERP values. Management reporting used consolidated data that had already been adjusted, filtered and aggregated. Each team trusted its own figure, because each used the data its own work depends on.
The numbers came from different systems, at different moments, under different definitions. Nobody was careless, and no single report was simply “wrong”. The gap between them was structural - and structural gaps do not close in a meeting.
Three correct answers to a question no one has defined the same way. The discussion becomes about which number is right - instead of what each number means.
One object - three perspectives, one shared context
| Perspective | Definition | Value | Source |
|---|---|---|---|
| Operations (live) | gross output, before adjustments | 12,400 t | MES |
| Finance (ERP) | net of returns, valuated | 11,980 t | ERP |
| Reporting (adjusted) | non-saleable excluded, business rules | 12,150 t | BI / Reporting |
Same asset – three valid perspectives on one object, each traceable to its source and point in time. The question stops being “which number is right?” and becomes “what does each number mean?”
Why it happens
Same word, different meaning
The causes are always the same handful. Different source systems and different points in time: one figure is live production, another a posted ERP value, another a report generated after corrections. Different definitions: gross, net and saleable output, rework and scrap treated differently. Different aggregation: grouped by shift, batch, order, plant or period. And manual adjustments in a spreadsheet that no other team ever sees. Wherever data crosses a system boundary, it quietly loses the context that made it comparable.
And the cost is not abstract. Gartner has estimated that poor data quality costs organisations an average of $12.9 million a year - much of it burned exactly here, in reconciling numbers that never agree instead of acting on them.
The problem in one image
Ask three departments how big an apple is. Sales says “medium”, the warehouse says “180 grams”, the quality lab says “calibre 70-75”. Nobody is lying - each measures in the way its own work requires. Production numbers diverge for exactly the same reason.
The way out
Connect the perspectives to one shared foundation
The fix is not forcing everyone onto one identical view - operations, finance and management legitimately need different lenses. It is connecting those lenses to one shared, business-centric foundation, with data referenced at its source rather than copied around:
- One definition, referenced everywhere. Agree the shared meaning of “output” once; each system keeps its own wording, mapped to it.
- Link, don’t copy. Every figure remains traceable to its authoritative source, so versions do not drift apart through uncontrolled copies and exports.
- Keep the relationships. Orders, batches, materials, quality status and time travel with the number, so context survives across boundaries.
One object. One shared context. Different attributes. We no longer argue about which number is right - we discuss what each number means.
Stop asking which number is right. Start asking which definition, which source, which moment.
The lesson travels
When your reports disagree, look below the reports
If operations, finance and management show different numbers for the same reality, the problem is probably not the people preparing them - it is the missing shared foundation the numbers can be traced back to (the same traceability gap explored here). Giving three honest perspectives one honest source is exactly what the boTec DataBridge is built to be.