Two-Tier ERP Explained: When Subsidiaries Should Run a Different System Than HQ
Rolling the corporate ERP out to every subsidiary can cost more than the subsidiaries earn. Here's how to decide which entities belong on a lighter system, and which integrations break if you get it wrong.
Choosing between one global template and a lighter system per country
If your group runs SAP S/4HANA, Oracle Fusion Cloud ERP or Dynamics 365 Finance at headquarters and owns a dozen smaller entities abroad, someone will eventually ask whether every subsidiary really needs the same system. The usual triggers: an acquisition in a new country, a rollout quote in the high six or seven figures for a 25-person sales office, or local finance teams keeping their real books in spreadsheets because the corporate template doesn't fit.
Two-tier ERP is the alternative. The parent keeps its tier-one system for group finance, consolidation and large operations. Smaller or structurally different subsidiaries run a lighter cloud ERP, commonly Oracle NetSuite, Microsoft Dynamics 365 Business Central, SAP S/4HANA Cloud Public Edition or Acumatica. They pass financials, intercompany transactions and sometimes orders up to tier one through defined integrations.
The technology isn't the hard part. You're deciding whether a subsidiary's processes differ enough, and its scale is small enough, that forcing the corporate template on it costs more than it returns. And if you split, you must decide how much integration discipline you'll fund so the group still closes on time.
Five criteria that decide which tier a subsidiary belongs in
Size alone doesn't decide it. A 30-person entity that manufactures components for the parent's plants is far more tightly coupled to HQ than a 200-person distributor selling its own product range. Score each entity against these criteria before you talk to any vendor.
| Criterion | Points toward one corporate instance | Points toward a subsidiary ERP |
|---|---|---|
| Operational coupling | Shares plants, inventory or planning with HQ; heavy intercompany supply | Sells, services or distributes largely on its own |
| Process fit | Same business model as the parent (same order-to-cash, same costing) | Different model: services, project work, retail, or a simple trading entity |
| Scale | Roughly 150-250+ users, or revenue large enough to absorb tier-one run costs | Under about 100 users with modest transaction volumes |
| Local compliance | Tier-one vendor has a mature localization for that country | Lighter ERP or a local partner handles statutory reporting and e-invoicing better |
| Ownership horizon | Core, long-term entity | Recent acquisition, joint venture or likely divestment |
Our position: a subsidiary that is operationally independent, under about 100 users and running a simpler business model belongs on tier two. An entity that shares production or inventory planning with HQ almost always belongs on the corporate instance, whatever its size. Splitting a tightly coupled supply chain across two systems creates more reconciliation work than license savings can justify.
Weigh the ownership horizon too. If an entity might be sold within a few years, a standalone subsidiary ERP makes the carve-out far cheaper, because you aren't untangling it from a shared instance.
Intercompany, consolidation and master data: where integrations break
Two-tier projects rarely fail during selection. They tend to fail at the second year-end close, when integrations designed on a whiteboard meet real transaction volumes. The trouble usually lands in three places.
Intercompany integration
When the subsidiary buys from the parent, the parent's sales order has to become the subsidiary's purchase order at the same price, currency and transfer-pricing basis, and the goods receipt has to match the parent's shipment. If that runs on emailed spreadsheets or a nightly file drop, mismatches pile up: timing differences, FX rate differences, partial shipments booked on one side only. Insist on automated document flow (order, invoice, credit note) with shared reference numbers, plus a clear rule for which side's FX rate wins. Net and settle monthly, not just at quarter end.
Multi-entity ERP consolidation
The subsidiary's trial balance has to map to the group chart of accounts. Groups typically choose one of two routes: a mapped chart in the subsidiary ERP that rolls up automatically, or a separate consolidation layer such as SAP Group Reporting, Oracle Financial Consolidation and Close, OneStream or a mid-market close tool. The common failure is local finance creating accounts nobody maps, which surfaces as an unexplained balance during the close. Put group-level account creation behind an approval workflow.
Master data
Item codes, customer and vendor records, units of measure and cost centers drift apart within months unless one system is the clear source of truth for each object. Decide per object. Usually HQ owns items and group customers, the subsidiary owns local vendors, and sync runs one way. Two-way sync of the same object is a red flag in any integration design.
When local tax and data rules decide the tier for you
Statutory requirements often force the decision before any process analysis. Real-time or clearance-model e-invoicing applies in a growing list of countries. Mexico's CFDI, Brazil's NF-e, India's GST e-invoicing, Italy's SDI, Saudi Arabia's ZATCA e-invoicing, Malaysia's MyInvois and Poland's KSeF all require invoices in specific formats, often validated by the tax authority before they're legally valid. Several other European countries are moving the same way, though timelines change, so check the current schedule locally.
For each country, check three things:
- Does the candidate ERP have a vendor-maintained localization, or does it rely on a partner add-on? Partner add-ons can work well, but confirm who updates them when rules change.
- Does it produce the statutory financial statements, local chart of accounts and audit files the country requires, such as SAF-T where it applies?
- Where is the data hosted? Data-protection rules differ widely: GDPR transfer requirements in Europe, China's PIPL and data-export rules, localization expectations in parts of Southeast Asia and the Middle East. Some countries effectively require local hosting or a local partner.
Vendor availability varies too. Business Central and NetSuite are sold in many countries, but localization depth and partner density differ by market, and a vendor that's strong in Western Europe may have one thin reseller in East Africa. Use this section as a checklist for your advisors, and confirm tax and data-residency obligations with a licensed local tax adviser and counsel in each jurisdiction.
What each model costs over five years
License price is rarely the deciding number. What matters is the per-entity rollout cost plus the integration you must build and maintain. The ranges below are typical 2026 market figures for a subsidiary of 20-80 users. They vary widely by contract, region and partner, so confirm them with vendors.
| Cost line | Extend corporate tier-one ERP | Lighter subsidiary ERP |
|---|---|---|
| Subscription per user per month | $150-$350+ (often bundled into the group contract) | $70-$200, plus platform or base fees for some vendors |
| Implementation per entity | $250,000-$1,500,000+ | $40,000-$250,000 |
| Integration build (intercompany, consolidation, master data) | Minimal if on the same instance | $30,000-$150,000 initially, often reusable across entities |
| Ongoing integration and support | Shared central support team | $10,000-$40,000 per year, plus a local partner retainer |
| Typical go-live time | 6-18 months | 3-6 months |
Here's a hypothetical example. Say a manufacturer on S/4HANA owns six sales-and-service subsidiaries with 25-60 users each. Extending the corporate template at roughly $400,000 per entity comes to about $2.4 million before run costs. A two-tier approach might cost $120,000 per entity to implement plus a $120,000 integration framework built once: about $840,000 in total, then perhaps $25,000 per entity per year for integration upkeep and partner support. Over five years the two-tier model still comes out well ahead here, but only because the framework is reused. Give each subsidiary its own custom integration and most of the gap closes.
That's why the integration framework, not the subsidiary license, is the cost line to scrutinize.
Running the decision in 60 days
Most groups can reach a defensible answer in about two months if they work in this order:
- Score every entity against the five criteria above. Expect a clear split: tightly coupled entities on tier one, independent ones as tier-two candidates.
- Fix the group finance model first. Agree the group chart of accounts, intercompany policy and master data ownership before evaluating subsidiary systems. Otherwise every vendor demo will look fine.
- Check compliance per country. Shortlist only vendors with proven localization and a partner in each target country.
- Design the integration once. Specify intercompany document flow, consolidation mapping and master data sync as a reusable pattern, and ask vendors to quote against it.
- Pilot one entity through a full month-end and one quarter-end close before committing the rest.
The decision rule: put an entity on a lighter subsidiary ERP when it is operationally independent, its process model differs from HQ, and you can fund a standard integration that every tier-two entity will share. If any of those three conditions fails, extend the corporate system.
Quick answers
Is two-tier ERP only for large enterprises?
No. Mid-sized groups with even three or four international subsidiaries use it, often with NetSuite or Business Central at the parent and subsidiaries on separate instances or companies. What matters is whether the entities differ in process and compliance needs, not the parent's size.
Should the subsidiary ERP come from the same vendor as HQ?
It helps but isn't required. Same-vendor pairings, such as S/4HANA with S/4HANA Cloud Public Edition or Dynamics 365 Finance with Business Central, often come with prebuilt intercompany and consolidation connectors that cut integration effort. A different vendor can still be the better choice if it handles local compliance or the subsidiary's business model much better.
Do we need a separate consolidation tool?
Not always. With a handful of entities and a well-mapped chart of accounts, the tier-one ERP's consolidation features may be enough. With many entities, several currencies, partial ownership or complex eliminations, a dedicated consolidation layer usually pays for itself through a faster, more auditable close.
How long does a subsidiary ERP rollout typically take?
For a 20-80 user entity on a lighter cloud ERP, three to six months is typical once the group integration pattern exists. The first entity takes longer because you're building that pattern. Later rollouts can often run in parallel.
What is the biggest red flag in a two-tier ERP proposal?
Intercompany integration described as manual or file-based, or master data synced in both directions without a clear owner. Both tend to produce reconciliation backlogs that surface at year-end. Ask the vendor or partner to show a working intercompany order-to-invoice flow, not slides.
Can we move a subsidiary onto the corporate ERP later?
Yes. If an entity grows or becomes more tightly coupled to HQ operations, migrating it to tier one is a normal step. Keeping its chart of accounts and master data aligned with group standards from day one makes that migration much cheaper.
Written by the Fleetmesh Tech editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.