How to choose ERP software in Bangladesh

Start with the problems, not the feature list

Most ERP evaluations begin with a vendor demo and a long list of modules. That is the wrong starting point. Before you compare products, write down the five operational problems that cost you the most today: stock counts that never match the register, purchase orders raised twice, invoices that wait for someone to return from the factory, month-end reports assembled by hand from Excel files, or an owner who cannot see the day's sales without phoning three people.

Each of these problems maps to a specific capability — a live stock ledger, an approval workflow, a shared customer master, an automated report. When you know which capabilities you need first, you can judge every ERP against the same short list instead of being impressed by features you will never switch on.

Involve the people who will use the system daily. A store-keeper, a merchandiser and an accountant see different problems from the managing director. Their input at this stage prevents the most common outcome of ERP projects in Bangladesh: a system that management bought and staff quietly work around with WhatsApp and paper.

• List the five most expensive operational problems and who owns each one
• Translate each problem into a capability the ERP must provide
• Note which processes are truly specific to your business and which are standard

Understand what ERP cost in Bangladesh actually includes

The headline price is rarely the real price. ERP cost in Bangladesh usually breaks into five parts: the software itself (a per-user subscription for packaged products, or a build price per module for custom systems), the implementation and configuration work, data migration from Excel and older tools, training for every role, and ongoing support. Hardware such as barcode scanners, label printers and a reliable internet connection at each site is a sixth item that is often forgotten.

Per-user pricing deserves special attention. It looks cheap for ten users and becomes a growth tax when you want every line supervisor and store-keeper on the system. Ask each vendor what the annual cost would be with two or three times your current user count. Ask, too, what happens when you need a change — is customisation billed by the hour, and does it survive the next upgrade?

Compare options on a three-year total that includes licences, implementation, changes and support. That is the number the owner will actually pay, and it often reverses the ranking that first-year prices suggest.

Local requirements a generic ERP may not handle

An ERP that works well in one market can be awkward in Bangladesh. Check how the product handles VAT and Mushak forms, whether it can produce the documents your buyers or auditors expect, and whether it copes with intermittent power and connectivity — offline capture on the floor with later sync is worth far more than a polished dashboard nobody can open during load-shedding.

Language and payments matter too. Interfaces and reports in Bengali reduce training time and errors for floor and store staff. Integration with bKash, Nagad, SSLCommerz and local banks removes manual reconciliation. Garments and textile companies should also confirm support for buyer-specific compliance documentation, bonded warehouse records and style/colour/size level inventory rather than a flat SKU list.

None of these requirements is exotic, but they separate products built for the market from those merely sold in it. Put them on your evaluation list explicitly and ask for a demonstration with your own examples, not the vendor's sample data.

Custom ERP versus packaged ERP: how to decide

Packaged ERP suits businesses whose processes are close to standard practice and who are willing to adapt to the software. It is faster to start, comes with vendor documentation and a support organisation, and the per-user subscription is predictable at small scale. The trade-off is that customisation is limited or expensive, and every workaround your team invents outside the system erodes the single source of truth.

Custom ERP suits businesses with processes that are a genuine competitive advantage, with unusual product structures, or with growth plans that make per-user pricing painful. You own the code and the data, the screens match how your floor actually operates, and modules are added when you need them. The trade-off is that you depend on the quality of your development partner and must invest more time in the design phase.

A useful test: list the ten workflows that matter most and mark how many a packaged product handles without workaround. If most fit, packaged is sensible. If half or more need customisation, the honest comparison is between a heavily customised package — which is expensive to maintain — and a system designed around your processes from the start.

• Choose packaged when processes are standard and the team accepts adapting to the tool
• Choose custom when workflows are specific, ownership matters, or headcount will grow substantially
• Avoid the middle ground of a package customised beyond recognition — it costs the most over time

Questions to ask every ERP vendor

Once you have a shortlist, ask each vendor the same questions and record the answers side by side. Who owns the data and can you export all of it at any time? Where is the system hosted and who controls the accounts? What does a change request cost and how long does it take? Can you see the system running with your product structure and your VAT scenario, not the demo company's?

Ask about implementation experience with businesses like yours in Bangladesh — the same industry, similar size, similar number of sites. Ask what training looks like for a store-keeper who is more comfortable in Bengali than in English. Ask how support works during a month-end closing at nine in the evening.

Be cautious of any answer that includes a specific percentage improvement or a guaranteed return. Serious partners speak about outcomes qualitatively — fewer re-entries, same-day reporting, one stock figure everyone trusts — and put effort into a scoping exercise before quoting.

Plan the ERP implementation in phases

A big-bang go-live across every department is the riskiest way to implement an ERP. A phased plan starts with the module that solves the most expensive problem — usually inventory or sales — proves it with real data, trains the roles involved and only then moves to the next module. Each phase produces value early and teaches both sides how to work together.

Data migration deserves its own timeline. Cleaning years of Excel registers, standardising item names, de-duplicating customers and reconciling opening balances takes longer than most owners expect and is the foundation of every report the ERP will ever produce. Assign one person on your side to own master data.

Finally, plan for the weeks after go-live. Hypercare — daily check-ins, quick fixes and a visible feedback loop — is what turns a delivered system into an adopted one. An ERP implementation in Bangladesh succeeds or fails on adoption, and adoption is a management decision as much as a technical one.