↳ Solutions · Not an OMS
DataGlass is not an OMS — what the difference is, and when to run both
Thailand has a mature order-management category. Shipnity, Zort, Page365, BigSeller, SellerPao, MyCloud and others have spent years on order sync, stock sync across channels, packing queues, shipping labels, COD reconciliation and chat-to-order. If what you need is for orders to be received, picked, packed and shipped correctly, one of them is the right answer and DataGlass is not.
Here is the test that separates the two systems, and it is not a criticism of anyone. On 4 August 2026 Shopee Thailand raised its Sale Transaction Fee: of 1,437 leaf categories, 1,382 rose by 2 percentage points before VAT.[1] Every OMS in the country kept shipping orders that morning exactly as it had the day before, correctly. Nothing in an OMS's job description is to notice that 1,382 categories just moved and re-derive which SKUs stopped being worth selling and which keywords stopped being worth bidding on. That is the whole difference. One system makes the order happen. The other decides what should happen next.
An OMS is a system of record for what already happened. A decision layer is an argument about what to do tomorrow. Confusing them is how a shop ends up with perfect operations and a shrinking margin.
Who answers which question
The clearest way to see the boundary is to list the questions a marketplace shop asks in a day and ask which system is supposed to answer each one.
| The question | Order management system | DataGlass |
|---|---|---|
| An order arrived — how does it get picked, packed and shipped? | Core job. Packing queues, label printing, carrier handoff. | Not our job. Shopee shipping documents can be archived and reprinted here, but there is no packing workflow. |
| How much stock is left, and has it been deducted across channels? | Core job. Real-time central stock deduction across marketplaces. | We do not deduct stock. We forecast when it runs out and how much to reorder. |
| Closing a sale over LINE or Facebook chat | Core job for several Thai vendors. | Nothing. We only see what the marketplace APIs expose. |
| Which SKU actually lost money after fees, vouchers and returns? | Sales reports, generally not profit reconstructed per SKU. | Core job. Every order rebuilt from settlement data down to the SKU. |
| What should tomorrow's ad budget be, per campaign? | Out of scope. | Core job on Shopee, with the downside bounded and the model inputs shown. |
| The fee schedule just changed — which SKUs flipped to loss-making? | Nothing changes. Orders keep shipping, correctly. | Recomputes the catalogue against the new schedule and flags what crossed. |
| Issuing tax invoices and feeding the accounting system | Built in, or integrated with Thai accounting software. | Not available. |
Read the table down the right-hand column and the shape is obvious: DataGlass is absent from every row about executing an order and present in every row about choosing what to do. That is deliberate. Building a fourth Thai OMS would be a worse use of the time than building the layer none of them contains.
What DataGlass does not do
Stated plainly, because a vague capability list is how software gets bought for the wrong reason and churns three months later.
- No order intake, packing workflow or real-time stock deduction. If two channels oversell the same unit, DataGlass will show you afterwards; it will not prevent it.
- No chat commerce. No LINE OA or Facebook inbox, no chat-to-order, no broadcast. A shop selling mainly through chat is largely invisible to us.
- No tax invoices, no accounting integration, no payroll. Those belong to Thai accounting software and to an ERP; see the ERP comparison below.
- No COD reconciliation against carriers. We reconcile marketplace fees and escrow-to-payout; the money in transit at the courier is not something we see.
- No public API and no published OpenAPI specification. Data comes out through the interface and through exports, not through an endpoint you can build against.
- No ad bidding on Lazada, and TikTok Shop bidding is not generally available. Ad budget automation is Shopee today; fee reconciliation and per-SKU profit cover all three marketplaces.
- No claim of a measured causal profit lift. The optimisation is evaluated within its own model against the alternatives it considered — that is a counterfactual, not a controlled experiment, and we say so on every page that touches it.
Running both, and the one place they collide
The two systems mostly do not touch. An OMS writes to orders, stock and shipments. DataGlass reads from the marketplace and writes to ad budgets, bids, prices and promotions. They are not competing for the same objects, which is why "use both" is not a diplomatic answer but the structurally correct one: the OMS stays the system of record, DataGlass becomes the decision layer beside it.
There is exactly one seam. Both can change a price. If the OMS pushes prices from a master product list and DataGlass proposes a price change on the same SKU, the last writer wins and nobody knows which one it was. Pick one place to set price and make the other read-only on that field. The same discipline applies more loosely to stock: DataGlass will tell you when a SKU runs out and what to reorder, but the reorder should be executed wherever your stock master lives, not in parallel with it.
When DataGlass is the wrong tool
- If today's problem is that orders are being packed wrong, shipments are late, or the same unit is oversold on two channels, that is an OMS problem and DataGlass does not solve it. Fix the operations first; a better decision on top of broken fulfilment is worth nothing.
- Below roughly 50 orders a month, neither category is urgent. The cost reconstruction that makes per-SKU profit meaningful needs enough orders to be stable, and the ad budget question barely exists at that volume.
- If most of your sales close over LINE or Facebook rather than on the marketplaces, DataGlass sees almost none of your business, and its picture of your margin will be wrong in a way that is hard to notice.
- If you need data flowing into your own systems programmatically, there is no public API to build against today. Exports exist; an endpoint does not.
- If the primary need is financial consolidation, multi-entity reporting or procurement, that is an ERP, not this and not an OMS.
Adjacent comparisons, if the tool you are weighing us against is something else: DataGlass vs an ERP, DataGlass vs Shopee Seller Center.
↳ Methodology
The description of the order-management category is drawn from how those vendors position themselves publicly, not from a feature-by-feature bake-off, and no claim is made about any individual vendor's capabilities. The category is named because sellers name it; the comparison is with the category, not with a competitor.
The DataGlass column describes what ships today. Where something is not available it is listed as not available rather than as roadmap, and the fee change used as the worked distinction comes from the DataGlass Marketplace Fee Dataset (v2026.08.25, verified 25 August 2026), which transcribes each platform's own dated fee documentation. Nothing on this page claims a measured profit lift.
Sources
- DataGlass Marketplace Fee Dataset (Thailand), v2026.08.25 — dated change log: Shopee raised the Sale Transaction Fee on 4 August 2026, with 1,382 of 1,437 leaf categories up 2 percentage points before VAT. CC BY 4.0. /research/marketplace-fees
- Shopee Thailand — Fee Policy and the Sale Transaction Fee schedule by category, effective 4 August 2026. help.shopee.co.th
- How DataGlass gets and uses marketplace data: Data ingestion for Shopee sellers, Multi-shop analytics across Lazada, Shopee and TikTok.
Keep the OMS. Add the layer it does not contain.
DataGlass connects directly to Shopee, Lazada and TikTok Shop — no change to how your orders are received or shipped — and returns per-SKU profit after fees plus the daily ad and pricing decisions that follow from it. First 14 days free.
Get started for free