Is NetSuite overkill for a $20M-$30M GMV Shopify brand, or does that revenue band already justify the switch from QuickBooks?

Why the revenue band is the wrong question

A $28M single-brand DTC company selling 400 SKUs from one 3PL through one storefront in one country has a bookkeeping problem that QuickBooks Online Advanced solves. A $22M company with a US entity, a Canadian entity, a wholesale division on net terms, a retail door, and inventory that moves between two owned warehouses has a structural problem that no general ledger solves, at any price.

Same revenue band. Opposite answers. Revenue is a proxy for complexity, and at $20M to $30M it is a bad one, because that is precisely the band where complexity and revenue decouple.

The four triggers that actually decide it

1. Entity count. More than one legal entity that has to be consolidated, with intercompany transactions between them, is the single strongest ERP signal. In NetSuite this is not an add-on setting; it is OneWorld and the Multi-Subsidiary Customer feature, and NetSuite documents that once you enable that feature, disabling it requires reversing or deleting every transaction tied to a secondary subsidiary (per Oracle NetSuite documentation, September 2026).
2. Inventory complexity. Landed cost allocation, lot and serial tracking, multi-location transfer orders, assemblies and work orders, bin management. QuickBooks Online does not do these; QuickBooks Enterprise does some of them. This is where most of the genuine $20M to $30M ERP cases actually live, and it is usually invisible on a P&L.
3. Revenue recognition beyond the point of sale. Subscriptions, bundles delivered over time, deferred revenue, multi-element arrangements. NetSuite's Advanced Revenue Management is the module for this, and it comes with its own one-way door: "When you enable the ARM (Essentials) feature, it can't be disabled" (per Oracle NetSuite documentation, September 2026).
4. Channel count with real settlement differences. One Shopify store is a solved problem. Shopify plus Amazon FBA plus wholesale EDI plus retail POS, each settling on different terms with different fee structures, is four reconciliations that a general ledger was not designed to carry.

If none of the four is true, NetSuite is overkill, and it stays overkill at $50M. If two or more are true, the case is real at $15M, and waiting for a revenue milestone is just choosing to do it under more pressure later.

What the money actually looks like, and what nobody publishes

We will not quote a NetSuite total cost of ownership, because there is no published list price to quote and every third-party figure we found this session came from a reseller or a review site with an interest in the number. Ask for a quote, and ask specifically for the module list, the user count and the one-time implementation fee as three separate lines, because the software subscription is routinely the smallest of the three.

The number that actually decides it, and how to get it

Not GMV. Hours per month spent reconciling, plus the number of decisions your team defers because the data is not trustworthy.

The method we give clients, which costs nothing and takes a fortnight:

1. Count the manual touches in one close. Every spreadsheet, every journal someone types, every inventory number nobody trusts, every hour spent explaining a variance. Write them down as they happen, not from memory.
2. Separate the two categories. Touches caused by process (nobody built the report) versus touches caused by structure (the system cannot represent the thing). Process problems do not need an ERP. Structural problems do not go away without one.
3. Count the structural ones only. That is your real case.
4. Price the alternative honestly. Brightpearl by Sage is live on the Shopify App Store at 4.6 stars across 53 reviews, free to install, positioned as "a Retail Operating System for retailers & wholesalers designed for scale" (verified September 2026). Mid-market retail operations platforms are a real middle tier, and skipping straight from QuickBooks to NetSuite skips it.

If step 3 produces a short list, you have a reporting project. If it produces a long one, you have an ERP decision and the revenue number is irrelevant to it.

When NetSuite is genuinely right at $20M to $30M

• Two or more legal entities with intercompany activity. Consolidation in spreadsheets is a control weakness before it is an inefficiency.
• Manufacturing or assembly. Work orders and bills of materials are not a QuickBooks concept.
• Serialized or lot-tracked inventory with a recall obligation. Regulatory traceability is a structural requirement, not a nice-to-have.
• An imminent transaction. Diligence is the one context where "our numbers come out of a general ledger with an audit trail" is worth real money on its own.

When NOT to do it

• Because a board member asked what ERP you are on. That is a status question wearing a systems question's clothes.
• Because close takes too long. Measure first. A slow close is usually an unowned process, and an ERP inherits unowned processes rather than fixing them.
• In the same year as a replatform. Moving the storefront and the system of record in one year means that when a number is wrong, nobody can say which project caused it.
• Before you have a controller who has run one. The implementation is configured by people, and the most expensive NetSuite instances we hear about are the ones configured by an implementation partner against requirements nobody internal could defend.