Three systems, three revenue figures
Customer, product and finance identifiers were never designed to match, so meetings begin by arguing about whose number is right instead of deciding anything.
Single source of truth
One place the numbers come from
Your finance, customer and operational records joined into one dataset, with each figure defined once and traceable back to the system it came from.
The same number usually exists in three or four systems and nobody owns which one is right. We extract from each source, resolve the customer and product keys that stop the records joining, define every measure once, and reconcile the result against the figures your board already signs off. Where a business has no joined data at all, this is normally the entire first engagement.
Discuss this buildWhen this comes up
What this usually looks like from the inside.
One person holds the whole picture
The joins, the exceptions and the workarounds live in a single head. None of it is written down, and none of it survives that person taking annual leave.
Every report is rebuilt by hand
Extracts get pulled, pasted and adjusted each month. A number costs days to produce and there is no record of how it was reached.
The work
The work, step by step.
Scope varies with the state of your systems. These are the parts that recur.
Source mapping
We find every system holding a record that matters, establish what can actually be extracted from each, and sit with the people who use them.
Identity resolution
We fix the customer, product, site and period keys that stop records joining, then write the rules down so they are not rediscovered later.
The build
Extract, transform and load into one store, with each measure defined once and reconciled against numbers your team already accepts.
Handover
Definitions, documentation and a runbook, written for whoever inherits this rather than for us.
What you leave with
Everything you keep.
You own all of it: the code, the written definitions and documentation aimed at whoever maintains this after us.
- A map of every source system and what it holds
- One data store with each measure defined once
- Written definitions and join rules
- A runbook your team can operate from
How this runs
Scope, timescale and what happens afterwards.
Roughly the first week
We establish which commercial question is worth answering here, and what your data, systems and people can realistically support.
The bulk of the engagement
We build it, rather than specifying it for somebody else to build. Reconciled continuously against numbers your team already accepts.
The final fortnight
Documentation, definitions and a runbook, plus a walkthrough where your team drives and we watch.
Only if you ask
Ongoing technical support is quoted separately from the build, so declining it costs you nothing.
Worth asking
Questions this tends to answer.
- Which version of this number is the right one?
- What breaks if the data person leaves?
- Can any figure be traced back to its source?