Your ₹9 Lakh ERP Handles One GSTIN. The Second Costs ₹5.4 Lakh Extra

TL;DR: Most ERP vendors quote per legal entity, not per user, so a second GSTIN doesn't add users, it adds a new customer. The real cost of multi-GSTIN rollouts hides across five layers (license, implementation, integration, support, training) and a second entity typically adds around 60% of the first entity's total cost. Once you cross a few GSTINs, a custom multi-entity ERP with shared masters and per-user licensing beats per-entity licensing on 3-year TCO.

Key Takeaways: - A single-GSTIN ERP quote is the floor, not the ceiling. The per-entity license model charges again for every new company you add. - Five hidden cost layers stack on top of the license: implementation, integration, support AMC, and training, each priced per entity. - Break-even for a custom multi-entity build sits between the second and fourth GSTIN, depending on user count and entity growth plan.

You approved the ₹9 lakh quote. Implementation kicked off, users got trained, and the books closed cleanly on entity one. Then finance added a new GSTIN for a regional subsidiary, and the vendor's "incremental cost" came back at ₹5.4 lakh.

Nobody flagged this during the original sale. The sales team quoted the single-entity scenario, which is the floor of what you'll ever pay, not the ceiling.

This is not a sales trick. The single-entity quote includes implementation services that do not scale linearly, such as configuration, data migration, master setup, and user acceptance testing. Entity two still needs its own chart of accounts, opening balances, user training, and UAT cycle. The vendor reuses almost nothing, so they charge a large share of what entity one cost.

The base quote was never meant to represent the cost of running a group company. It represented the cost of onboarding one of them.

The trap becomes worse when the ERP you bought is still running in production five years later. Lock-in is real. Switching costs compound because your master data, custom reports, and integrations all carry forward into the new entity.

By the time you realise entity three will cost another large sum, you have two options. Pay the per-entity tax, or rebuild. Both are expensive.

The ERP implementation cost in India conversation has to happen at the multi-entity level, not the single-entity level. The trap isn't a sales tactic. It's baked into how enterprise software is priced, and understanding that structure is the only way to escape it.

SAP, Oracle, Odoo, and ERPNext all structure their multi-company modules around the legal entity. Each company in your group becomes a separate licensed company record. It carries its own books, its own tax registrations, and its own access controls.

The "unlimited users" promise that closed your first deal means unlimited users within one company. Anything outside that scope is a new deal. The moment you create a second company record, you're not adding users. You're adding a tenant.

That tenant needs its own user pool, its own license, and its own implementation scope. Even on a cloud ERP, hardware costs do not disappear. Multi-entity tenants require separate database schemas, separate compliance environments, and separate backup policies.

Infrastructure spend inflates before software licensing even kicks in. Cloud billing models charge for what you allocate, not what you share.

A second GSTIN also means a second integration surface. Each entity needs its own GST API credentials, e-invoicing registration, and bank or payment gateway integrations. None of these are shared.

Your ERP software development partner will tell you that per-entity integration is heavy work. Every connector has to be re-tested against the new entity's tax profile rather than reused as-is.

The "unlimited everything" headline that closed your first deal was scoped to one legal entity. A custom ERP India build can be architected around shared masters and entity-level filters from day one, but that requires a different conversation, not just a different vendor.

The license model is the first cost layer and the most visible one. The cost you can't see in any quote is where the real damage happens.

When a second GSTIN arrives, the bill doesn't come as one line. It comes as five. - Layer 1 - License: Incremental per-entity fees scale with each new company record, even though the underlying code is identical. You are paying for the right to create another company record. - Layer 2 - Implementation: First-entity implementation services are a multiple of the licensing fee. Entity two still needs data migration, master setup, and UAT. Vendors price this work separately because it doesn't fully reuse entity one's setup. - Layer 3 - Integration: Each GSTIN needs its own GST API credentials, e-invoicing registration, and bank or payment integrations. These are per-entity and rarely included in the base quote. - Layer 4 - Support and AMC: Annual renewals scale with entity count, and they compound over time. Most vendors charge renewals per entity, not per installation, so AMC inflates as you grow. - Layer 5 - Training and change management: The second entity often has different users, different processes, and a full re-training cycle that vendors price separately.

A vendor will quote you Layer 1 confidently and quietly, then bill Layers 2-5 as "additional scope" once you're already committed. This is where ERP hidden cost India becomes a board-level conversation rather than an IT one.

We mapped the same pattern across years of Indian ERP cost transparency work. The line item you didn't ask about is the one that doubles the bill.

Layer 3, integration, is where multi-GSTIN ERP work gets dangerous for compliance. Each entity's data isolation must be auditable.

The hidden cost there is not money. It's the audit trail you fail to build until a reviewer asks for it.

The question shifts from "how much does a second GSTIN cost" to "when does a custom multi-entity build beat paying per-entity forever."

Break-even sits between the second and fourth GSTIN. Beyond that, a custom multi-entity ERP that prices per-user rather than per-company is cheaper on a 3-year TCO basis. Below it, the off-the-shelf multi-tenant option usually wins on upfront cost.

The per-user cost of a custom build drops once you stop paying per-entity license surcharges. A custom build amortised across four or more entities usually beats multi-tenant TCO at the per-user level.

The licensing fee per additional entity drops to near zero once the architecture is in place, because the per-entity license surcharge disappears. Implementation effort still exists for new entities, but the cost is configuration, not full re-implementation. This is the structural reason per-entity licensing loses against per-user licensing as entity count grows.

The flip only works if the custom build uses a shared chart of accounts, shared master data, and entity-level filters. Not separate databases bolted together. This is the ERP development cost decision that separates a cost-effective custom build that pays for itself in year three from an over-scoped one that never does.

A well-built custom ERP with shared masters runs in production for years before a major refactor. That's a stronger TCO signal than any perpetual license renewal on a multi-tenant platform.

The other variable most CFOs miss: per-user licensing on a custom build means your cost scales with headcount, not with corporate structure. If you add a new GSTIN with the same users, your marginal cost is configuration hours, not a new tenant.

Most CFOs stall at the modelling step. The four rows below close that gap.

Four rows. Run it before you sign anything. - Row 1 - One-time cost per entity: Add license, implementation, integration, and training for each GSTIN you operate today. Then add each entity you plan to spin up in the next 36 months. This is your multi-entity capex. - Row 2 - Annual recurring cost per entity: AMC, support, cloud infra. Multiply by year and by entity count. Renewals compound year over year, and they scale with the number of entities you maintain. - Row 3 - The custom-build scenario: Single implementation cost, single per-user license, near-zero incremental cost per additional GSTIN beyond configuration hours. This is your counter-quote. - Row 4 - Sensitivity test: What happens to per-entity cost if you add two more GSTINs in year two? What if the vendor raises renewals each year at the rate their contract permits? Run both.

Most Odoo implementation partners will hand you a Row 1 and Row 2 quote without prompting. Almost none will build Row 3 for you. That is the gap.

The ERP cost India question is not "what does one entity cost." It is "what does entity five cost in year three, and do we know that number today."

A second entity at ₹5.4 lakh and a fifth entity at the same ₹5.4 lakh is a per-entity cost of ₹5.4 lakh forever. A second entity at ₹5.4 lakh and a fifth entity priced as configuration hours only is a per-entity cost that drops with scale. The worksheet is the only way to see which quote you actually have.

Once the numbers are on paper, the conversation with the vendor gets shorter, and the decision gets clearer. Most procurement teams never get here, because running Row 3 forces a comparison the vendor would rather not put in writing.

You stop treating each new GSTIN as a capital event. Each one becomes a configuration task. The difference: a ₹5.4 lakh line item versus a configuration-hours line item.

The CFO walks into a board meeting with a per-entity cost that is known, bounded, and defensible. No more negotiating from scratch each time.

M&A scenarios become modelable. If you acquire entity number six, you can price the integration before the deal closes. The board stops asking "how much will this cost" and starts asking "which option has the lower 3-year TCO."

For context on how this plays out across years and multiple vendors, see how Your ₹8 Lakh ERP Needs a Shadow Team to Actually Work once the per-entity fees start piling up.

The vendor conversation also changes. You stop accepting per-entity line items as fixed. You start asking for per-user, per-entity, or shared-master pricing, the same language your ERP system architect would use.

And you have the SAP vs Odoo comparison on paper before the sales call, not after.

This is the kind of cost modelling that enterprise-grade ERP decisions are made on. The right pricing model is upstream of the right vendor. Get it first, and the rest follows. The vendors who can't model that conversation transparently are the ones most likely to overcharge you on entity three.

How much does a second GSTIN cost on an existing ERP in India?

For most off-the-shelf ERPs, a second GSTIN adds a large share of the first entity's total cost. Documented cases show ₹5.4 lakh on top of a ₹9 lakh base. The exact number depends on whether your vendor prices per-entity or per-user, and whether implementation services are reusable across entities.

Is a custom multi-GSTIN ERP cheaper than SAP or Odoo for a group company?

Usually yes once you cross 3-4 GSTINs. A custom build with shared masters and entity-level filtering removes per-entity licensing. That's the largest driver of per-user cost growth as you add entities. Below 3 entities, the off-the-shelf option often wins on upfront cost.

What are the hidden costs of adding a new GSTIN to an existing ERP?

Beyond the license fee, expect to pay separately for GST API and e-invoicing integration per entity, additional implementation hours for master data and UAT, renewed user training, and inflated annual support contracts. Vendors rarely bundle these into the base quote.

Do Odoo, ERPNext, and SAP all charge per-entity for multi-GSTIN setups?

Yes. All three structure their multi-company modules around legal entities. Each GSTIN needs its own company record and license. Most also need a separate database schema. Per-user licensing exists, but the per-entity overhead remains in implementation, support, and integration.

How do I budget for a multi-GSTIN ERP rollout over 3 years?

Model three line items per entity: one-time implementation cost, annual recurring cost (renewals that scale with entity count), and integration cost. Then run a parallel custom-build scenario where the second and subsequent entities cost only configuration hours, not full implementation.

Run the worksheet above and bring the numbers to your next vendor call.