What ERP software really costs in Bangladesh

Search for the price of ERP software in Bangladesh and you will find figures that differ by two orders of magnitude, none of them sourced. That is not because the market is chaotic. It is because the pages answering the question blend three different kinds of evidence into one column of numbers.

Below, those three never share a column. It makes the page longer and it makes it usable.

Checked 19 August 2026. Every figure and quotation below was read from the vendor's own public pages on the date shown. Vendors change pricing, packaging and localisation without notice, so check the source link before you rely on anything here.

5 of those rows are a vendor publishing its own price. 3 are a vendor publishing none. That ratio is the single most important fact about buying software in this country, and it is the reason the rest of this page exists.

The most consequential pricing decision is not the headline number, it is whether the number multiplies by your headcount. Odoo publishes a per-user price, which makes it the one competitor whose cost at scale can be computed exactly rather than guessed.

The dollar figures are left in dollars deliberately. An exchange rate written into a page is wrong within weeks and wrong in a way nobody notices, so convert at the rate on the day you are deciding rather than at a rate we baked in months ago.

The general point survives the caveats: a per-user price is a tax on growth, and it is why so many businesses in this market end up sharing logins, which is a security problem created by a pricing model.

1. 01Per-user creep. A quote sized for the team you have today is a quote that grows with the team. Ask for the cost at your headcount in three years, in writing, before you sign.
2. 02Customisation. The gap between what the software does and what you do is filled by somebody, at a day rate. It is also the line that makes future upgrades expensive, because every customisation is a thing that has to survive the next version.
3. 03Data migration. Chart of accounts, parties, stock items, opening balances. Cheap if scoped, expensive if discovered. Barcodes usually do not exist yet and creating them is real work that gates a till rollout.
4. 04Annual maintenance. On a perpetual licence this is a recurring percentage that is easy to forget when comparing against a subscription. Over three years it usually closes most of the gap.
5. 05The person who becomes the system's carer. Somebody applies the updates, checks that the add-on still works, and fixes it at month end. This is the largest omitted cost in every quote we have seen, and on a self-hosted open-source stack it is the whole cost.

Software is bought monthly and lived with for years, so compare over three. Our own figures below are derived from the published rate card rather than typed, so they cannot drift from it.

Set that against the shape of the alternatives. A self-hosted open-source stack is a small hosting bill plus an implementation quotation plus a person. A per-user cloud product is a number that multiplies. A perpetual licence is a large first year, a smaller recurring maintenance percentage, and a server somebody owns. None of those three is inherently worse; they fail in different places.

Publishing these is the one structural advantage on this page, and it is not a tactic anyone can copy without changing how they sell.

• /4 published tiers, from BDT 1,990 a month, plus a custom package that carries no published number because it is scoped per customer.
• /Priced per workspace rather than per seat. Each tier carries a ceiling on users, concurrent tills, warehouses and storefront domains, and those ceilings are printed next to the prices rather than discovered later.
• /No separate setup fee on the published tiers.
• /Implementation for a larger scope is quoted, and that is also not a published number. We are not going to pretend otherwise on a page about vendors who do not publish numbers.

The full rate card with the ceilings is on the pricing page. The numbers on this page are derived from the same source, so the two cannot disagree.

1. 01Find the unit. Per user, per outlet, per till, per company, or per year. Everything else in the quote depends on it and it is often the least prominent line.
2. 02Ask what it costs at your year-three headcount and outlet count, in writing.
3. 03Separate one-time from recurring, then total the recurring over three years before you compare anything.
4. 04Ask what happens in year two: does maintenance start, does a discount expire, does support change tier.
5. 05Ask which of the things you demonstrated are in the base product and which are customisations, and get that list itemised.
6. 06Ask who applies updates and what happens if you skip one for a year.
7. 07For anything involving VAT, ask whether the statutory capability is in the product or in an add-on, and who supports the add-on.
8. 08Ask for one reference customer of your size and trade, and call them.

Every published figure above came from the vendor's own pricing page on the date shown. Third-party figures are attributed in the table itself and are never presented as a vendor's price.