Most finance software can be replaced in a weekend. Change expense tools, change AP tools, change the reporting layer, and the business barely notices.
The general ledger is the exception. It holds the record everything else refers to, and swapping it means a migration: moving history, reconciling it year by year, rebuilding a chart of accounts, retraining everyone. Companies that have done it describe it as painful as a general rule, and they are right.
Which makes the general ledger the one component worth over-evaluating, because the cost of getting it wrong is not the licence fee. It is being stuck with the consequences until the next migration.
ledgers at Tillo, 7 instances of QuickBooks and 1 of Xero, on a single system
of the journal population an auditor extracts through the API, not a sample
countries supported out of the box, from US GAAP to UK MTD VAT
1. Can it hold more than 1 entity, properly
Not "supports multi-entity," which every vendor claims. The question is whether entity is a dimension of a transaction or a separate database.
If it is a separate database, the group view gets assembled every month, forever, and adding a market is a project. If entity sits on the transaction alongside department and country, the group view is the default and a single entity is a filter. Tillo runs 4 trading entities in 23 countries this way and closes in 5 days.
2. Is it immutable
An entry, once posted, should never be edited in place. A correction is a new entry referencing the original, so the record only grows.
This sounds like an accounting nicety and it decides what an audit costs. A trial balance pulled today reconciles to the entries as they were posted rather than as somebody later remembered them, and 3 years later the evidence is identical. Customers at $500M ARR have completed audits on that basis, with the auditor extracting the full journal population through the API instead of sampling it.
3. Does it hold the original currency
Most systems convert on entry and keep the original amount in a memo field, if at all. Everything after that is correction: revaluation entries, translation adjustments, a cumulative translation account only 1 person understands.
A ledger that records the transaction in the currency it happened in, keeps the rate that applied, and derives every other view is a different proposition. Tillo revalues 23 currencies as part of the ledger rather than as a month-end project.
4. Can it report under more than 1 standard
A group under IFRS with a US subsidiary on US GAAP needs both answers from the same transactions. The traditional resolution is a second set of books or a manual adjustment pack, and both create a permanent reconciliation.
Entry-level
Built for one entity
Fast to start, and every subsequent entity, currency and standard becomes something bolted alongside rather than held inside.
Enterprise
Built for a larger company
Holds the complexity and asks for a consultant every time the business changes shape. The architecture is right and the operating model is not.
What to look for
Multi-everything, admin-changeable
Entities, currencies and standards as dimensions of one record, with the finance team able to change structure without raising a ticket.
5. Is the API real
The test that separates marketing from architecture, and it takes 1 question: can an auditor extract 100% of the journal population through it, and can a developer write to it.
Read access matters for audit and reporting. Write access is what lets a finance team build the thing the vendor does not do. Dreamdata's controller built a layer on top of Light's API that reads its data, applies its own judgement over the company's chart of accounts, and now handles the bulk of the bookkeeping. All Gravy connected its own CPQ tool directly, so a contract closing in HubSpot generates the invoice.
Neither of those was a feature. Both were possible because the ledger was open.
6. Who can change it
The property that determines total cost of ownership, and the hardest to see in a demo.
When the business needs a new dimension, a changed approval threshold or a different revenue treatment, does an admin do it, does a partner bill for it, or does it go on a roadmap? A ledger that requires a consultant for every change is a ledger that stops changing, and the finance function then shapes itself around the software rather than the business.
What the answers add up to
Companies at this size are usually choosing between an entry-level ledger built for a single entity, with point solutions bolted alongside as complexity arrives, and an enterprise system sized for a much larger company.
The first is cheap until the second entity, then quietly expensive forever. The second is architecturally right and operationally heavy. Tillo, All Gravy, Oper Credits, Alva Labs, Paebbl, Omnea, KeyShot and Officeguru all took the same third route: one ledger built from the start to hold multiple entities, currencies and standards, changeable by the finance team.
"Group entity management has saved every member of the team hours and hours."
Harriet Stewart, Head of Business Systems, Tillo
The general ledger is the only decision on the shortlist that constrains every other one. Everything layered on top can be replaced in a weekend. Choose accordingly.
See what Light replaces, read the multi-entity evaluation, or book a demo.