How to choose ERP software in Bangladesh: a 15-point checklist

Choosing ERP software is a decision you will live with for five to ten years, and the demo is designed to make you like the vendor rather than to test the product. This checklist gives you fifteen things to check for a business in Bangladesh, in the order that tends to matter, with a way to score vendors against each other. Use it with any vendor, including us.

Before the checklist

Write down your three worst problems in one sentence each, for example: “Stock in the Chattogram godown never matches the sheet”, “Month-end VAT takes a week”, “I cannot see daily sales without calling”. Give those sentences to every vendor before the demo and insist that the demo starts with them. A vendor who cannot show your three problems being solved on their standard product is going to solve them with customisation, which costs more than the licence.

The 15-point checklist

Compliance and local fit

1. Native Mushak forms. The system should print a Mushak 6.3 from a real sale with sequential numbering, build the 6.1 purchase register automatically and produce a return summary for the 9.1, without a partner add-on. Ask to see a 6.3 for a registered buyer with a BIN. See accounting and Mushak VAT.

2. MFS and gateway payments. bKash, Nagad, Rocket and Upay as payment methods at the counter with transaction references, SSLCommerz for online collection, and wallet accounts that reconcile like bank accounts. Be sceptical of the phrase “bKash integration” until you see what it means.

3. Bangla where your staff need it. Staff work faster in the language they think in. Check whether the interface is available in Bangla today or only on the roadmap, whether Bangla text works everywhere as data (item names, customer addresses, receipts, invoices), whether SMS templates send Bangla Unicode correctly to Grameenphone, Robi and Banglalink numbers, and whether training and support are given in Bangla.

4. Bangladesh payroll. Attendance from ZKTeco or similar devices, shifts and overtime, festival bonus, salary advances, and income tax slabs that are configurable per financial year following the NBR circular. See HR and payroll.

Operations

5. The modules you need now, switchable later. Buy for the next two years, not the next ten. The system should let you switch modules on and off per business so you are not paying for manufacturing you do not do.

6. Multi-outlet and multi-godown. Stock by location, transfers with approval, outlet-level access so one branch does not see another’s margins, and consolidated reports for the owner. See POS software for the outlet side.

7. Offline resilience. Load-shedding and internet cuts are facts. Ask exactly what happens at the counter when the internet drops for ten minutes, how quickly the system recovers, and whether the vendor offers on-premise deployment where connectivity is bad. Be wary of anyone who says it never happens.

8. Works on a phone. Most owners run the business from an Android phone. Check whether the system is a responsive web app, a native app or both, and try approving a purchase from a phone during the demo.

Vendor and data

9. Local support in your hours. Support in Bangla, by phone or WhatsApp, during Bangladeshi business hours, from people who know the product. Ask what the response time is for a counter that cannot bill.

10. Data export and ownership. You should be able to export every table, customers, items, transactions, ledgers, to Excel or CSV yourself, at any time, and the contract should say the data is yours. Test the export during the trial. See reports and business intelligence.

11. Customisation path. Every business needs something specific eventually. Ask who does customisation, how it is priced, whether it survives upgrades, and whether you can see an example.

12. References and a demo on your data. Ask for two customers you can call. A new vendor may have few; that is not disqualifying, but they should say so plainly and compensate with a longer trial or a smaller first phase. Then insist on a demo using your own item list and a real month of transactions.

13. Security, roles and audit logs. Role-based permissions down to the screen, an activity log on every module showing who changed what and when, and a clear answer on backups and where data is hosted.

Money and contract

14. Total cost of ownership. Licence, implementation, training, support, hardware, SMS credits and customisation for three years, in one table. The 2026 price guide explains each line, and our pricing page shows how we approach it.

15. Contract terms. Notice period, cap on annual price increases, what happens to your data when you leave, uptime commitment for cloud, and the scope of implementation in writing with acceptance criteria.

Scoring vendors

Weight the points by your situation. A retail chain weights offline resilience and POS; a distributor weights multi-godown and party ledger; a service company weights payroll and reporting. The weights below are a starting point.

Point	Weight (1 to 3)	Vendor A	Vendor B	Vendor C
1. Native Mushak forms	3
2. MFS and gateway payments	2
3. Bangla where staff need it	2
4. Bangladesh payroll	1 to 3
5. Modules switchable	2
6. Multi-outlet and godown	1 to 3
7. Offline resilience	1 to 3
8. Works on a phone	2
9. Local support	3
10. Data export	3
11. Customisation path	2
12. References and demo on your data	2
13. Security and audit logs	2
14. Total cost of ownership	3
15. Contract terms	2

Score each vendor 0 to 3 per point, multiply by the weight, and total. The number is less important than the conversation it forces: the point where you argue about a weight is the point that matters to your business.

Red flags

• Every question is answered with “yes, we can do that” and nothing is shown.
• The demo runs only on the vendor’s sample data.
• Mushak forms are “coming next quarter”.
• Pricing changes each time you ask.
• Data export is possible “on request”.
• The contract has no exit clause, or has automatic renewal with a long notice period.
• Support is a ticket system with no phone number and nobody who speaks Bangla.

Local or international

Local ERPs (Troyee, PrismERP, Biznify, SORS ERP and others) tend to have native Mushak forms, Bangla support and local service, with smaller teams and shorter track records. International products (Odoo, ERPNext, SAP Business One, Zoho Books) have polished interfaces and large ecosystems, with Bangladesh compliance delivered by partners and pricing often in foreign currency. Neither is right by default; the checklist above is how you decide. Our comparison pages on Troyee, Odoo and SAP Business One set out the differences one product at a time.

Where SORS ERP fits

Against this checklist, SORS ERP scores well on native Mushak forms, MFS payment methods with SSLCommerz, Bangla text as data throughout with Bangla SMS and onboarding, training and support in Bangla, Bangladesh payroll with configurable tax slabs, per-business module switches, multi-outlet stock, role-based access with audit logs, Excel export and both cloud and on-premise deployment. Where we are weaker: SORS Technology was founded in 2024 so our reference list is short, the interface is English only today with a Bangla interface on the roadmap, there is no native mobile app (the web app installs as a PWA on Android), manufacturing is on the roadmap, and there is no direct bKash merchant API. We would rather you know that before the demo than after the contract.

Bring your three sentences and your item list, and request a demo that starts with them.

Written by the team that builds SORS ERP, the ERP described on this site. We implement it ourselves in Bangladesh, so the guidance here comes from setups we run rather than from general research. More about SORS Technology.

Frequently asked questions

Should I choose a local or an international ERP?

Local products usually have native Mushak forms, Bangla support and local service, with smaller teams and shorter track records. International products have polished interfaces and large partner networks, with Bangladesh compliance delivered by partners and prices often in foreign currency. Score both against the same checklist and let the totals decide.

How many vendors should I shortlist?

Three is enough to compare and few enough to give each a proper demo on your own data. More than that and the demos blur together. Drop any vendor that will not demo on your item list.

What should an ERP demo include?

Your three worst problems solved on the standard product, a Mushak 6.3 printed for a registered buyer, a sale with a split cash and bKash payment, a stock transfer between two locations, an Excel export of your data and a look at the audit log. Ask which features shown are live today and which are on the roadmap.

How long should an ERP contract be?

One year for cloud subscriptions, with a cap on annual price increases and a clear exit clause including data return. Avoid multi-year lock-in until the system has run through at least one full financial year, including Eid season and a VAT audit cycle.