ERP for an Indian MSME: Tally, Odoo, SAP Business One or custom?

Your accounts team lives in Tally. Sales sends orders on WhatsApp, the stores team keeps stock in a spreadsheet, and the owner gets the real numbers by phoning three people. At some point someone says the word ERP, and the conversation jumps straight to products.

Choosing the product first is the most common mistake we see. The better question is which process is breaking, and how badly.

Quick answer

• If your pain is mainly accounting volume, GST and reporting, stay on TallyPrime and build a small amount around it.
• If you manufacture, run several warehouses or need approvals across departments, look seriously at Odoo or SAP Business One.
• Choose Odoo if you want flexibility and are willing to keep customisation under control. Choose SAP Business One if you want a more fixed structure and a partner who supports it for years.
• Build custom software only for the workflow that makes you different, and connect it to a standard accounting core.

Start with what is actually breaking

Spend a week writing down where time and money leak. Be specific. "Inventory is a mess" does not help. "We cannot tell how much of item 4471 is in the Surat godown without calling Ramesh" does.

The usual suspects in an Indian MSME are:

• Stock figures that differ between the books, the godown and the sales team.
• Production planning done on a whiteboard, with raw material shortages discovered on the shop floor.
• Purchase orders raised without approval, or approved on WhatsApp with no record.
• Month end closing that takes two weeks because data arrives late from branches.
• Costing that nobody trusts, so pricing decisions are made on feel.

Rank these by cost to the business. The top two decide your route far more than any feature list.

When Tally is still the right answer

TallyPrime is fast, cheap to run, understood by every accountant in the country and handles GST, e-invoicing and e-way bills well. For a trading business with one or two locations and a few thousand invoices a month, replacing it is often a bad trade.

What usually works instead:

1. Keep Tally as the accounting book of record.
2. Add a simple order and stock application for sales and stores, on the web or a phone, that writes vouchers into Tally through its integration interface.
3. Put a reporting layer on top so the owner sees sales, receivables and stock without asking anyone.

The limits are real. Tally is not built for multi-step manufacturing, detailed approval workflows or many concurrent users editing operational data. If those are your top problems, add-ons become a pile of workarounds.

Odoo: flexible, open, and needs discipline

Odoo covers accounting, inventory, manufacturing, purchase, sales, CRM, HR and more, with an Indian localisation for GST. The Community edition is open source; the Enterprise edition adds features and is licensed per user.

Its strength is that you can shape it. That is also its risk. We have seen Odoo projects where every department asked for its own changes, the partner said yes to all of them, and two years later nobody could upgrade to the next version.

Odoo suits you if:

• You have someone internal who will own the system and say no to unnecessary changes.
• Your processes are reasonably standard, with a few genuine exceptions.
• You want to add modules over time as you grow.

Budget for implementation, data migration and training, which together usually cost more than the licences. For a 20 to 50 user rollout, a realistic timeline is three to six months.

SAP Business One: structured and partner led

SAP Business One is aimed at small and mid sized companies and is sold and implemented through partners. It has a stronger fixed structure than Odoo, good inventory and production features, and a recognisable name that some auditors, lenders and parent companies like to see.

The trade-offs: licences and annual maintenance cost more, you depend heavily on the quality of your partner, and changes outside the standard design are slower and more expensive. It fits businesses that are happy to adopt a standard way of working.

Ask any partner for two references of a similar size and industry, and call them yourself. Ask what went wrong.

When custom software makes sense

Custom software is the right choice for a narrow set of problems: a pricing engine for made to order products, a dealer scheme calculator, a job card flow on your shop floor that no package models properly. It is the wrong choice for general ledger, GST returns or payroll, where a package carries years of regulatory updates you would otherwise have to maintain.

The pattern that holds up is a standard core for accounting and statutory work, with custom applications at the edges that talk to it through APIs. If you build, make sure you own the source code and that the system is documented well enough for another team to pick it up.

How to decide in about three weeks

1. Week one: list the processes, rank the pain, and collect sample data, such as a month of invoices, the item master and the stock register.
2. Week two: run scripted demos. Give each vendor the same three real scenarios from your business and watch them run those.
3. Week three: compare total cost over five years, including licences, implementation, support, upgrades and your own staff time.

If two options look close, pick the one your team is more likely to use correctly every day. An average system used properly beats a strong one that people work around.