Most groups do not decide to leave spreadsheet consolidation. They are pushed — by one more acquisition, one more currency, one more restatement the model was never built to carry. The workbook that served three entities is quietly straining under twelve, and everyone can feel it at close. The question is not whether Excel is capable. It is whether the risk you are now carrying is one a board would knowingly accept.
Below are seven signs a group has outgrown spreadsheet consolidation. You will not have all of them. Two or three, recurring every close, is usually enough.
How do I know we've actually outgrown it, and not just had a bad close?
The tell is repetition: the same failure returns every period, regardless of who runs the process. A bad close is an event. An outgrown process is a pattern. Read the seven signs as symptoms, and pay attention to the ones that recur — those are structural, and no amount of care will drill them out.
| Sign | The risk it carries |
| The close slips a day with every acquisition | The board pack arrives late, and later each year |
| Intercompany never quite eliminates to zero | A residual that hides a real error — or masks one |
| Multi-currency translation is done by hand | Translation and rate errors flow straight to reserves |
| There is no audit trail; the auditors ask for one | An audit finding, and a longer, costlier audit |
| Version chaos — "Consolidation_FINAL_v7b" | Someone reports off the wrong file |
| A restatement or re-org means rebuilding the model | Weeks of rework, and new errors introduced in the rebuild |
| Only one person understands the workbook | Key-person risk: if they leave, the close stalls |
Why does the close slip a day with every acquisition?
Because a spreadsheet grows linearly and a group grows structurally. Each new entity is another tab, another set of manual links, another mapping to maintain by hand — and the effort compounds rather than adds. A model built for a handful of subsidiaries does not scale to a portfolio; it degrades. If your close is drifting outward year on year, that drift is the cost of the tool, not the team. Our companion piece on how many days a group close should take sets out a realistic benchmark to measure yourself against.
Is intercompany that never nets to zero really a problem?
Yes — a stubborn residual is rarely cosmetic. When intercompany balances refuse to eliminate cleanly, it usually means two entities have booked the same transaction differently, at different rates, or in different periods. In a spreadsheet the fix is to plug the difference and move on. That plug is the problem: it either conceals a genuine misstatement or becomes a habit that quietly erodes the numbers' credibility. Good consolidation matches intercompany at source and shows you the mismatch, rather than letting you paper over it.
What's the real danger in translating currencies by hand?
Manual translation is one of the easiest places for a material error to enter undetected. Applying the wrong rate, using an average where a closing rate belongs, or leaving the cumulative translation reserve to a side calculation — each is a single keystroke away, and none announces itself. The office of the CFO ends up defending figures it cannot fully trace. A proper consolidation applies the right rate to the right line by rule, and keeps the translation reserve as an output of the model, not a manual afterthought.
Why do the auditors keep asking for an audit trail?
Because a spreadsheet cannot tell them who changed what, when, or why. Every consolidation adjustment sits inside a formula with no record of its origin or approval. Auditors increasingly expect to trace a reported number back through each adjustment to its source — and when they cannot, the audit takes longer, costs more, and may surface a finding on controls. An audit trail is not a nicety here; it is fast becoming the price of a clean opinion.
How much does version chaos actually cost us?
More than the annoyance suggests. When the definitive file is whichever copy was last emailed, someone eventually reports off the wrong one — in a board pack, a covenant certificate, or a regulatory return. "Consolidation_FINAL_v7b" is funny until it is the version the numbers went out on. A single governed model, with one source of truth and controlled access, removes the entire class of error rather than managing it.
What if a restatement or re-org means rebuilding the model from scratch?
That is the clearest structural warning of all. If changing your legal or reporting structure means unpicking hundreds of hard-coded links, the model has hard-wired assumptions it was never meant to hold. Real groups reorganise, acquire, and occasionally restate — and the consolidation should absorb that through configurable hierarchies, not a rebuild. When change forces a rewrite, you are one restatement away from a very long month.
How risky is it that only one person understands the workbook?
It is the risk boards understand fastest, because it has a name: key-person risk. When a single individual holds the logic of the group's numbers in their head, a resignation, an illness, or a holiday can stall the close outright. No documentation fully closes that gap, because the knowledge lives in the formulas and the habits, not on paper. A shared, transparent model turns a single point of failure into a repeatable process.
What does "good" look like, and how do I decide it's time?
Good is a single model that matches intercompany, translates currency by rule, keeps a full audit trail, and supports multiple hierarchies for legal, management, and statutory views. One version, traceable, that survives a re-org. Decide it is time when the signs above recur every close, when the risk has moved from inconvenience to something you would not want to explain to your auditors or your board, and when the manual effort is displacing the analysis the office of the CFO should be doing. The strongest signal is simple: you have started building controls around the spreadsheet to compensate for the spreadsheet. That is the moment to cost the alternative. Modernising consolidation is not only a risk trade — there is measurable value in the days and errors you recover. The score is the headline; the pounds are the point.
Common questions
When should a group replace spreadsheets for consolidation?
There is no single entity count that decides it. The practical trigger is recurrence: when the same failures — a slipping close, an intercompany residual, a manual translation error — return every period regardless of effort. That pattern signals a structural limit, not a bad month, and is the point to cost an alternative.
Isn't Excel still fine for group consolidation if we're careful?
Care manages symptoms; it does not remove the underlying risk. A spreadsheet cannot enforce intercompany matching, hold an audit trail, or absorb a re-organisation without a rebuild, however diligent the team. Once you find yourself building controls around the spreadsheet to compensate for it, the tool has become the risk.
What does good group consolidation actually look like?
A single model that matches intercompany at source, translates currency by rule rather than by hand, records a full audit trail of every adjustment, and supports multiple hierarchies for legal, management and statutory views. Crucially, it survives a restatement or re-org through configuration, not a rewrite. The result is one traceable version of the numbers.
How does key-person risk show up in spreadsheet consolidation?
It shows up when only one person truly understands the workbook's logic. Because that knowledge lives in the formulas and working habits rather than in documentation, a resignation or a period of absence can stall the close entirely. A shared, transparent model converts that single point of failure into a repeatable process.