Solutions · Group consolidation
Numminen Advisory builds and repairs group consolidation for investor-backed international startups and technology groups. Group consolidation combines the accounts of a parent company and its subsidiaries into one set of figures, with the transactions between them removed.
The work here is the structure underneath it: the chart of accounts, the intercompany model, the elimination logic, and a group close calendar that runs monthly without heroics.
Shapes we see often, not a specification. A group that matches none of them can still have exactly this problem.
Two to five entities, usually a parent with subsidiaries in the United States, the Gulf, the United Kingdom or another EU country.
An investment round on the horizon, or close to closing. You need to ensure group financial records are investor ready.
Reporting under local statutory rules, plus IFRS for investors. US GAAP appears when the counterparty is American.
A competent controller in place who has not built a group consolidation before.
In group consolidation, AI now does the matching and the arithmetic. It pulls each entity’s trial balance, matches intercompany lines and prepares FX translation and elimination entries. At Numminen Advisory, Joanna Numminen reviews and approves every entry, so the consolidated numbers carry a finance leader’s judgement and a full audit trail.
| Finance task | What AI does now | What stays with the CFO |
|---|---|---|
| Trial balances | Pulls each entity’s trial balance from NetSuite and the local ledgers. | Mapping decisions and the group chart of accounts. |
| Intercompany | Matches intercompany lines across entities and lists the differences. | Resolving differences with each entity’s finance team. |
| FX translation | Translates entity results at closing and average rates. | Rate policy and the treatment of unusual movements. |
| Eliminations | Prepares elimination journals for review. | Approval before anything is posted. |
Seven patterns show up in almost every group we look at.
One spreadsheet. The consolidation lives in a spreadsheet that one person maintains and nobody else can follow. When that person is on holiday during diligence, the group has no numbers.
Intercompany that does not agree. Intercompany balances do not agree, and the difference is written off without explanation. Both sides were posted by different people in different systems on different days, and nobody owns the reconciliation.
Annual only. The group position is available annually, months after the year ends. Adequate for filing, useless for running a company or populating a data room.
Two treatments, no decision. Development cost capitalisation, grant income and leases are treated one way in the statutory books and another way under IFRS, and nobody has decided which number the company is managed against.
Unclear loans. Intercompany loans with no agreement, no interest or unclear repayment terms, so it is hard to say whether they form part of the net investment.
Wrong rates. Income statement or balance sheet items are translated at the wrong rate, for example a closing rate used for the P&L.
Untracked CTA. The cumulative translation adjustment is not tracked by entity.
Statutory filing requirements and commercial reality are different questions. Whatever an obligation does or does not require of a group your size, your board needs to see the group, your investors need to see the group, and an eventual acquirer will want three consistent years of it.
Numminen Advisory is founded and run by Joanna Numminen. In an applied AI group scaling across several countries she led end-to-end group consolidation across every entity and owned the full close cycle, through the period in which the group was acquired. In a B2B SaaS group before that she owned consolidation across a multi-entity structure while three reporting frameworks ran at the same time.
Both are fixed in scope and quoted before the work starts. The diagnostic tells you what is wrong; the build fixes it.
A fixed-scope diagnostic over three to five days. It examines the consolidation method, intercompany balances, elimination logic and whether the structure will survive the next entity or the next auditor. It ends in a written findings memo with a ranked remediation list.
Project, four to eight weeks, with a defined end. It covers the chart of accounts designed for the group you will be in two years, an entity dimension, the intercompany model written down before the invoices start, elimination logic that runs monthly, and a group close calendar everyone works to.
Both end with something your team operates without us. A consolidation only one person understands is the problem, so rebuilding it around a single new dependency would be a poor trade.
Three things get confused with this work, and it is worth being direct about the difference.
There are good tools and we will help you choose one, but a tool applied to an undesigned chart of accounts makes the same mess faster. The design comes first.
They keep the books. This sits above that, and it usually improves what you ask of them.
If your audit firm builds your consolidation, it is later auditing its own work. Manageable while the group is small, and a real problem exactly when the group starts to matter.
Diligence looks back two to three years. If the group only consolidated properly from year three, years one and two get rebuilt under time pressure, from records never designed to be assembled that way, while three hundred other questions are waiting.
The design work is roughly a week when there is no history behind it. After two years of transactions it becomes a project, and it runs concurrently with something more important.
Yes, more than once, in groups with entities across Europe, the Gulf and the United States.
Three kinds of provider exist. Consolidation software vendors sell the tool. Audit and accounting firms will build it as a year-end exercise. Independent practices like Numminen Advisory design and build it as a project, then hand it to your controller to run monthly. The third is usually right for a group of two to five entities, because the work is a design job with an end.
A person first, then software if the group justifies it. The expensive mistake is buying a consolidation tool before anyone has designed the chart of accounts, the entity dimension and the elimination logic it is supposed to run.
They can, and many do. The cost arrives later, when the same firm is auditing a process it built. Build it in-house or buy it from somebody who is not your auditor, not urgently, but knowingly.
A health check is three to five days, and a build is typically four to eight weeks. Both are priced against scope and quoted before the work starts, because the effort depends heavily on how many entities and reporting frameworks the group runs.
A monthly group close, produced from your own systems, with intercompany balances that agree and eliminations that run the same way every period. You also get the structure documented, so a new controller or an auditor can follow it without an interview.
The first question is not how to set up the new entity. It is what should be settled before the certificate arrives, which is a short list and takes about a week when there is no history to unpick.
