The first entity is easy. Any ledger handles it.
The second is where the trouble starts, and it does not announce itself. Somebody incorporates abroad, the local accountant recommends the local system, and it gets set up in an afternoon because that is genuinely the fastest way to be compliant by Friday. Nothing about that decision is wrong on the day it is made.
By the fourth entity the company has 4 systems, 4 bookkeepers who cannot cover for each other, 4 definitions of the same account, and a group number that exists only after somebody spends a week assembling it. No single decision caused that. All 4 were reasonable.
ledgers at All Gravy before Light: e-conomic, Xero, Fortnox and Tripletex, one per market
ledgers at Tillo, 7 instances of QuickBooks and 1 of Xero, collapsed into a single system
countries supported out of the box, from US GAAP to UK MTD VAT
What actually breaks
Not the accounting. The accounting is fine in every one of those 4 systems.
What breaks is everything that requires looking across them. A group P&L. A vendor's total spend when that vendor bills 3 entities separately. An intercompany recharge that both sides recorded correctly and slightly differently. A board question about margin by product line, when 1 market encoded product in the account number and another used a tag.
The close breaks worst, because it becomes 4 closes and a consolidation rather than 1 close. Each entity has to finish before the group work can start, so the slowest market sets the group's timetable. Tillo lived this: nobody could close an entity, log out and start the next one, and the close ran 12 days. On 1 ledger the team started working by function instead, revenue and approvals and reconciliations each done once for all 4 entities, and the close came down to 5 days.
Six things to evaluate
Whether entity is a dimension or a database. The single most consequential question. If each entity is a separate instance, the group view is always assembled. If entity sits on the transaction alongside department and country, the group view is the default and a single entity is a filter.
Whether the chart of accounts is shared. 4 entities on 1 ledger with 4 charts of accounts has most of the problem still intact. All Gravy rebuilt its chart of accounts from the ground up during migration, with custom dimensions for department and country from day 1, replacing a structure that had faked departmental reporting through dedicated GL accounts.
Whether intercompany eliminates itself. Both legs of an intercompany transaction should be the same record seen from 2 entities, not 2 records that have to be reconciled. At Alva Labs, elimination across the Swedish, Norwegian and UK entities drafts as the transaction posts.
Whether it holds more than 1 standard. A group reporting under IFRS with a US subsidiary on US GAAP needs both, from the same transactions, without a second set of books.
Whether currency is native. Tillo revalues 23 currencies inside a 5 day close. That is only possible when the transaction keeps its original currency and rate, and every other view is derived rather than converted on entry.
How much a fifth entity costs. The real test. Adding a market should be a configuration decision, not a procurement decision, a hiring decision and a permanent addition to the close.
Where most teams are
One ledger per market
Each entity on the local system a local accountant recommended. Compliant everywhere, coherent nowhere, and the group view gets built by hand.
The enterprise answer
A multi-year implementation
Solves the architecture by installing one sized for a much larger company, which then takes a consultant to change.
The third option
One ledger, entity as a dimension
Every entity, currency and standard on a single ledger. Eliminations and group reporting run against transactions as they post.
What it looks like at this size
None of the companies here are enterprises. That is the point.
Officeguru runs Denmark and Germany with a finance team of 2. Omnea runs the UK and US on 1 ledger. Ocean.io scaled 2 entities without adding finance headcount and cut close time 60%. Oper Credits runs Belgium, Switzerland and the UK, and went from 8 systems to 3. Paebbl runs the Netherlands, Sweden, the UK and Finland after 3 separate ERP general ledgers made the group, in its VP Finance's words, "difficult and time consuming." KeyShot runs Denmark and the United States on $30M+ ARR with a CFO who reckons he can reach $100M "without adding many more resources to it."
"Going from logging into four different systems to seeing every entity in one place was the part that changed my day."
Linea Meldgaard Andersen, Finance Analyst, All Gravy
When to move
The honest trigger is rarely the entity count on its own. It is the entity count plus something else arriving at the same time: a round that wants monthly consolidated reporting, a first statutory audit, an acquisition, or a finance lead working out that nobody can answer a group question inside a day.
All Gravy's Head of Finance and Ops framed the test as whether the foundation scales 10 times, which is the right question to ask before the fifth market rather than after it. The cost of moving grows with every entity added, and the cost of not moving is paid every month, quietly, in a week of assembly nobody has ever put on a budget line.
See how Light runs multi-entity groups, or book a demo.