The free inventory system nobody maintains: five questions before you install one

Every comparison of free inventory software is written from the same seat: day one, clean demo database, no warehouse staff, no month-end close. That is not where these projects fail. They fail in month three, when nobody remembers who was supposed to apply the security patch.

I have watched small teams move from spreadsheets to self-hosted stock systems and back again. The pattern is boringly consistent. The install was never the problem. The questions below are the ones that decide whether the thing is still running a year later.

Question one: who owns the upgrade?

Open-source inventory software is free in licence, not in attention. Odoo Community is modular — you can start with inventory alone and bolt on sales, purchasing or accounting later — but it has to be hosted and configured by someone who knows what they are doing. ERPNext gives you raw material tracking, finished goods and HR in one place, and it comes with a learning curve to match.

So before you download anything, name the person who will handle version upgrades and security patches. If that person is "whoever has time", the honest answer is that you are really choosing a managed service instead.

Question two: what happens when you hit the free ceiling?

There are two very different kinds of "free", and they fail differently.

Freemium SaaS is rented access. You use it free up to a limit of orders, products or users, then you pay monthly. Nothing is wrong with that model, but you need to know where the line sits before your data lives behind it. Zoho Inventory begins charging once your order volume passes its monthly threshold; the free tier of Katana is deliberately tight for manufacturers who are actually growing.

Open-source is a different trade: no licence fee and no ceiling, but you carry the infrastructure. The decision is not "free versus paid". It is "subscription later" versus "server bill now".

Question three: can you get your data out?

This is the question almost nobody asks during a demo, and the one that hurts most later.

With a self-hosted open-source system you own the database. With a SaaS tier your records sit on someone else's servers, and if the terms change or the product is retired, your history goes with it. Before you commit, run one boring test: export your product list and your stock movements, then try to read the file in a spreadsheet. If the export is a screenshot in a PDF, you have your answer.

Question four: will the people holding the scanner use it?

A warehouse picker does not care about your module architecture. Two things decide adoption: whether the interface is in a language they read without effort, and whether the daily task is faster than the paper it replaced.

Vietnamese localisation exists in the major open-source projects, but "exists" and "is complete for your screens" are different claims — check the actual screens your staff will touch. Barcode scanning on a phone beats manual entry every time, and granular permissions matter too: the person receiving goods should not be able to edit prices or open the financial reports.

Question five: who fixes the invoice side?

In Vietnam this is where free projects quietly stall. Stock records eventually have to meet tax records. Your system needs to hand data to the local e-invoice providers in a form that satisfies VAT reporting, and that integration is rarely part of the free package.

Decide early whether you will bridge it with a spreadsheet export, a small custom script, or a local partner. Discovering this in the week of a tax filing is the classic way to abandon an otherwise good system.

A rollout order that survives contact

Do not migrate the whole catalogue on a Friday. The sequence that works:

1. Put the software on a modest server and get one login working — nothing else.
2. Import a slice of your stock, not all of it, from a CSV export of your current sheet.
3. Run the new system and the old sheet side by side until the numbers agree twice in a row.
4. Move the rest of the catalogue, then switch daily receiving and picking to the new tool.
5. Only once the routine is stable, connect the sales channels.

Step three is the one people skip. It is also the only step that tells you whether the free tool understood your stock at all.

Where "free" stops being free

Budget these before you start, because they arrive whether or not you planned for them:

• Hosting. A virtual server is the floor. Local hosting also cuts latency for staff working in Vietnam during peak hours.
• Domain and certificate. Small money, but it is yours to renew.
• Setup and configuration. Either your own time or a consultant's invoice.
• Maintenance. Security patches and version upgrades are ongoing work, not a one-off task.
• Training. A capable system like ERPNext pays back only after the team has learned its workflow.

What I would ignore in the first month

Feature matrices. Multi-location tracking, purchase orders, low-stock alerts, turnover reporting — every serious option claims all of them, and you will not know which ones you actually need until you have run a month of real receiving and picking.

Pick the tool whose maintenance story you can live with, import a slice of data, and let the routine tell you what is missing.

The full comparison — the tools, their licence models, the feature checklist, and the notes on Vietnamese e-invoices and local hosting — lives in the original article: Free and open-source inventory software in 2026.