Replacing 1C: what to migrate first

The question “what do we replace 1C with” is almost always framed wrongly. What has to be replaced is not a program but the set of habits, documents and reports that grew around it over the years. Below: the order that keeps the business running, and what can be left alone for a long time.

Why “rewrite everything, then switch” fails

The urge to build the whole new system and flip over on a single Monday is understandable: it feels cleaner. In practice it is the riskiest option. While the new system is written over six months, the old one keeps living and changing — new documents, new rules, new exceptions. By switch-over day the new one is already behind.

The second problem is that Monday. If something goes wrong, rolling back is close to impossible: over the weekend the new system has accumulated data the old one does not have. So migration happens in pieces, and every piece has to work alongside the old system rather than instead of it.

What to migrate first

The best first piece is whatever hurts daily and has the fewest ties to everything else. Usually that is customer-facing paperwork: invoices, acts, delivery notes. They are produced by hand or half by hand, they eat hours, and an error in them is visible immediately.

Why those. First, the result shows within a week rather than six months — which settles the question of trust in the whole undertaking. Second, they barely depend on the complicated accounting machinery: you need a customer, a contract, line items and a total. Third, if something is wrong, the old system has not gone anywhere and the document can be produced there.

The worst first piece is stock or production. That is where the ties are densest, the exceptions most numerous and the mistakes most expensive.

What can stay untouched for years

Statutory reporting. If the accountant files from 1C and it works, that is the last thing worth touching. It changes with legislation, costs a lot, and brings the business nothing new: a report is either filed or it is not.

A practical arrangement that lasts: operational accounting and everything customer-facing in the new system, statutory reporting left in 1C, and an export between them. That is not an unfinished job — it is a deliberate split by cost of change.

Where data goes missing

First place: reference data. The same counterparty often exists in three variants — with and without the legal form, with and without a full stop, with a typo in the name. Until those are merged, any report in the new system shows the wrong thing. It is dull work, and it has to happen at the start, not “later, when there is time”.

Second: opening balances. The moment they are fixed must be single and documented. Otherwise, six months on, nobody can explain where the difference came from.

Third: everything that officially is not in 1C — spreadsheets, arrangements held in a manager’s head, discounts that are “always like that for this customer”. This surfaces only by talking to the people who work daily, and no technical audit will reveal it.

How not to stop the business

A period where both systems run is unavoidable. The only question is whether it is controlled. The rule is simple: at any moment, for each document type, exactly ONE system is authoritative. Not “either will do” — that guarantees divergence.

During that period the exchange between them is better kept one-way. Two-way sync during migration is a source of conflicts that take months to untangle afterwards.

And last: the old system is not switched off right after the transition. It stays readable at least until the annual period is closed — simply because the question “how was it in March” is certain to come up.

Frequently asked

How long does a migration take

It depends on how many document types there are and how tangled the reference data is. The first useful piece — customer paperwork — usually shows within a few weeks. Migrating operational accounting in full takes months, and it is more honest to plan it in stages than to name a single date.

Could we move to an off-the-shelf ERP instead of a custom system

Yes, and in a fair number of cases that is the right answer: if your processes are typical, an off-the-shelf system is cheaper than development. A custom system earns its cost where the process is non-standard and that process is the advantage — for example when accounting has to live alongside equipment, terminals or a complex catalogue.

What happens to the history from previous years

Normally the balances as at the migration date are carried over, not the full history. The history stays readable in the old system. Migrating ten years of transactions is expensive and almost never necessary — but that should be a conscious decision, not something discovered afterwards.