What is the difference between an ERP, a WMS and a data platform?

They record the same stock at different grains. A warehouse management system (WMS) records where each unit is and every move it makes: dock, bay, pick, load. An ERP records what the stock is worth and which orders, receipts and invoices it belongs to. A data platform keeps copies of both, and of other systems, over time, so questions that span them can be answered. The difference that matters in practice is where two of them record the same event, and disagree. Many yards and mills have no separate WMS, and for them the same logic applies to the ERP and the mill or scale system.

What the usual answer says

Explanations of the three define them by function. The ERP runs finance, orders and purchasing. The WMS runs receiving, putaway, picking and shipping. The data platform stores and analyses data from many systems for reporting and AI. Those definitions are right, and some explainers note that the ERP holds inventory at a higher level than the warehouse system. They leave out why a business with several systems can still get a simple number wrong.

Three grains of one record

Vendor documentation shows the difference in grain. Microsoft's description of its warehouse management module, as an example of the category, lists "full control of location stocking limits and location volumetrics" and "full traceability of workers' material handling". That is a record of places and movements, at the level of the bay and the scan.

An ERP's inventory record is coarser in place and richer in money. It holds quantity by item and site, the value of each receipt, and the link to the order and the invoice. It is the record the accounts close on.

A data platform sits beside both. The same vendor describes its analytics platform as one that "supports end‑to‑end data workflows, including data ingestion, transformation, real‑time stream processing, analytics, and reporting." Its job is to keep history from several systems and join them. Apart from reference tables, mappings and adjustments that people add, it mostly holds copies of records made elsewhere.

Demand for cross-system reporting is high, even inside ERP projects. Panorama Consulting Group's 2026 ERP report, from 170 respondents, found business intelligence was the initiative most often deployed as part of an ERP project, with 55.3% saying they had deployed it to a large extent. The share reporting benefits from "removing silos" rose from 55.2% to 77.4% on the previous year, which the report treats as a lagging benefit of earlier ERP investment.

Where they disagree

Trouble starts with events that two systems both record. A transfer between branches is the clearest case. To the WMS it is a movement: units leave one site and arrive at another. To the ERP it depends on how the transfer was set up. It may be a stock movement at cost, or it may be booked like a sale.

From Quarri's own work with a lumber and millwork manufacturer: inter-branch transfers booked at zero revenue but full cost produced a negative margin that didn't exist.

In a case like that, a record of the movement would show nothing wrong: the units left one site and arrived at another. The error is in how the value was booked, and a data platform that copied only the ERP's view would copy it too. Regraded stock is another shared event: one system may record a grade change while the other still values the unit at its old grade.

How to use the difference

Match each question to the grain that answers it: location and movement from the WMS or mill system, value and margin from the ERP, change over time across both from the data platform.

Then, for every event recorded in more than one system, write down which system is the record for quantity and which for value, and how the other should agree with it. In our view that list is short, and it is where unexplained numbers are worth looking first.

When it doesn't apply

Small yards and mills without a WMS keep locations in the ERP or not at all, so the ERP is the only record and the question of disagreement does not arise. Some ERPs include warehouse modules, which reduces the number of systems without changing the grains. And a data platform is unnecessary where one system answers every question the business asks.

Quarri for operations teams is built for the people running the crews, the lines and the yard.

Quarri is an AI-native data platform for the timber supply chain. It connects buying, production, sales and inventory for forest management, sawmill, wood products and pulp, paper and packaging operators.