Latest episode
A data assessment for finance, read entity by entity before anything is built.
We read each entity's ledger, mappings and workbooks, then hand back findings with an owner and an order against each one.
A data assessment for finance tells a multi-entity group with a date for AI which of its finance data a model can read today. Constancia, an EPM consultancy, reads it entity by entity, because a group that closes in 6 ledgers has 6 answers to that question.
- Reads
- Each entity's ledger, its mapping to the group chart and the workbooks between them.
- Produces
- A list of findings, each with a named owner and a place in the order of work.
- Then
- A fixed price for whatever the findings say has to be built first.
We read 6 things in every entity before we write a single finding.
Each entity is read in the order it closes, and the group office is read last. A group of 12 entities on 3 ERPs takes 3 reads of the export format and 12 of the mapping.
- 01
The entity ledger
Each ERP releases its trial balance at its own grain, and we record what the export holds and what it leaves behind.
- 02
The mapping to the group chart
Every local account maps to a group account somewhere, and we find whether that mapping lives in a system or in a workbook.
- 03
The intercompany balances
Each entity's intercompany balance is matched against its counterparty's, and the unmatched ones are listed with the entity that owes the explanation.
- 04
The customer, product and cost centre lists
3 entities that book the same customer under 3 names give a model 3 customers, and we count every case of that kind.
- 05
The workbook at the end
The last mile of most entity closes runs in 1 person's spreadsheet, and we name the person and the spreadsheet.
- 06
The owner of each figure
We ask who would sign each entity's numbers if the group auditor challenged them, and write the name or the gap beside the figure.
An entity with 2 ERPs is read twice, once for each ledger.
The estate looks ready from the group office and differs in every entity beneath it.
Our consultants have opened entity ledgers on 58 implementations over 35 years, and we count the names 1 customer carries first.
Where 3 subsidiaries book the same customer under 3 different names, the group report shows 3 customers and the sales director recognises 1.
The group controller finds the third name when a model is first asked for revenue by customer, and the model reports it 3 ways.
We write down what the data assessment covers and what it hands to somebody else.
The boundary is agreed with the CFO before the first entity is opened.
In scope
Every entity's finance data, read where it lives rather than where the diagram says it lives.
We read each ERP, each mapping and each workbook in the entity that runs it. Every finding names an entity, an owner and its place in the order.
- The trial balance from every ERP, at the grain it exports
- The mapping of each local account to the group chart
- The customer, product and cost centre lists in each entity
- The workbook that finishes each entity's close
Out of scope
A maturity score, a platform choice and any data that never reaches the ledger.
A score against a generic scale gives the board a number and gives nobody a task. Each finding we hand back carries a name and a place in the order.
- A maturity rating against a published scale
- The choice between OneStream (EPM software) and Abacum (FP&A software), which the findings inform later
- HR, sales and operations data that never reaches a ledger
A maturity score and a list of findings send the finance team in 2 different directions.
Both outputs answer the board's question, and the second one also starts the work.
| A maturity score | A list of findings with owners | |
|---|---|---|
| What the board receives | A number between 1 and 5 against a published scale. | A list of what fails in which entity, and who fixes each item. |
| What happens the following week | A workshop to agree what the number means. | The first 3 findings start, each with a named owner. |
| How the entity teams read it | As a verdict on the group, with no line about their own ledger. | As a page about their ledger, their mapping and their workbook. |
| How it prices a build | The estimate from the budget round stands, because a score prices nothing. | The findings set the order and the fixed price covers them. |
| How it ages | The number is repeated a year later and compared. | Each finding is closed by its owner and dated. |
A data assessment usually leads to a roadmap, and both belong to the AI advisory pillar.
AI readiness assessment
The survey of the whole data model, from the consolidation down, before any price is agreed.
Read about the readiness assessmentFinance roadmapping
The findings put in the order the data model can bear, with a date against each phase.
Read about roadmappingOneStream
Where the findings point at the consolidation, OneStream (EPM software) is 1 of the 2 platforms we build on.
Read about OneStream
See more on our
Subscribe on YouTube
Every Sunday, 6am
The Sunday Close.
Our weekly newsletter on AI in finance.
Yours to enjoy with your Sunday morning coffee.
- AI in finance observations from Alex (CEO) and Brad (COO)
- Newest video from our YouTube channel
- AI in finance trends we have spotted each week
Tell us the date you are working to.
The people who would do the work read every message and reply inside 2 working days.