Retail & eCommerce
Unify in-store and online selling with stock, orders, and customers managed from a single platform.
Who this page is for: Chains and brand owners selling through a mix of stores, their own site and marketplaces — roughly two to fifty stores, with cross-border sales starting or planned. A single shop does not need this page; a chain of hundreds needs a different class of system.
How the business actually flows
Seven steps every retail & ecommerce system has to chain together. Software that breaks the chain at any point just moves the spreadsheet one step downstream.
1. Assortment and SKU master
Items are created as variant matrices — style, colour, size, pack size — before anything else. This step decides how expensive the next five years are: a matrix created carelessly produces thousands of near-duplicate SKUs that nobody can count, reorder or retire.
2. Channel pricing and promotion
The same item carries different prices in store, on your own site and on a marketplace, and promotions run on different calendars per channel. Price therefore has to be a per-channel attribute, not a single field on the item.
3. Receiving into warehouse and store
Goods arrive against a purchase order, are inspected and split between warehouse stock and store stock. Warehouse-to-store transfers need both legs recorded, otherwise the units in transit exist in no location at all.
4. Selling: POS, own site, marketplace
Three intake channels with three different data shapes. The retail failure mode is not the sale — it is the day-end reconciliation when three channels disagree about what was sold.
5. Allocation and replenishment
Available stock is promised to channels under allocation rules so a marketplace cannot drain the units reserved for your own store. This is arithmetic, not a feature.
6. Returns and membership
A returned item is physically present but not sellable until inspected, and points or stored value earned on the original sale have to reverse consistently across channels.
7. Channel P&L and margin review
Commission, payment fees, fulfilment cost, advertising and return freight are deducted per order line — otherwise profitable and loss-making channels look identical on the revenue report.
Common challenges
- !Omnichannel stock inaccuracy
- !Slow order fulfillment
- !Limited customer insight
How a modern ERP helps
- ✓Real-time multi-channel inventory
- ✓Automated order routing
- ✓Unified customer profiles
Must-have functionality
If a product cannot do these, the rest of the demo is irrelevant.
- ✓One availability figure per SKU per location, with on-hand, allocated and in-transit kept as separate quantities
- ✓A store POS that keeps selling when the network drops, and reconciles cleanly afterwards
- ✓Membership, points and stored value shared between store and online — not two loyalty schemes
- ✓Order intake from the channels you actually use, with store pickup and local delivery where relevant
- ✓Channel-level contribution margin that deducts platform commission and freight, not just revenue
- ✓Returns handling that quarantines goods until inspection instead of reselling them immediately
Worth paying extra for
Optional only in the sense that the must-haves come first — in practice one of these is often the reason a project pays for itself.
- ◇Promotion engine covering full-reduction, gift-with-purchase and coupon stacking per channel
- ◇Store-level profit reporting for chains that hold each branch accountable
- ◇Same-city delivery and in-store pickup workflows
- ◇Assortment and seasonal markdown planning for fashion-led categories
Candidate products and why they are on the list
Every candidate is listed with the reason it is on the list and the limitation that could remove it. Fit is judged against the must-have list above, not against brand reputation, and it is a judgement about the shape of the requirement — not a statement that the features work.
Fit
Our judgement of whether the product covers the must-have list above — the shape of the requirement, not proof that any feature works. Evidenced means the capability question is also answered; potential — needs validation means the scenario matches but the deciding capability is still unproven, so the card is a possible candidate and never a recommendation.
Capability
Verified means we read it on the vendor’s own page on a stated date. Partly evidenced means some requirements are documented and the deciding one is not. Unverified means we found no evidence — which is not the same as unsupported.
Delivery
Confirmed would mean our sales side has confirmed we can contract for the product, the version and the implementation. Everything here is TBD, so no card is an offer.
Commercial disclosure. ERP Guide Hub is an editorial resource. We do not present ourselves as an authorised partner, reseller or distributor of any product named on this page. We may receive commission or referral fees on products we cover, Chanjet products included. That does not buy placement, a better position in a list, a softer verdict or the removal of a limitation. Nothing here is an offer to supply any product named on this page. Any figure or feature claim still needs checking against the vendor's current documentation. Read our editorial policy for the full terms.
This page specifically. Every candidate above is a fit judgement against the must-have list for Retail & eCommerce, not proof that the features work. Delivery is still TBD for all of them, so no card here is an offer to contract.
Chanjet HaoYeCai — store POS and membership
Verified for store POS, membership / points / stored value and online-offline integration, sitting on the same dataset as purchasing, inventory and accounting — so a store sale posts to the ledger without a nightly export.
Watch out: Positioned by the vendor for small chains; capacity for hundreds of stores is not published. Offline-tolerant POS, cross-store transfer costing and franchise settlement are not evidenced — verify with your own store count before shortlisting.
Verified capability sheet →Chanjet HaoShengYi (好生意) — small-format retail and wholesale
Verified for multi-store and multi-warehouse management, batch and shelf-life control, scan-based receiving and stocktake, smart replenishment, and connections to domestic marketplaces (Taobao, Tmall, JD, Pinduoduo, Kuaishou, Douyin) with store pickup and local delivery.
Watch out: Aimed at micro and small businesses with no production or project modules. Marketplace connectors evidenced are domestic only — no Amazon or Shopee. We have not confirmed that we can supply this product; treat it as a fit assessment, not an offer.
International mid-market retail suites (NetSuite, Dynamics 365 class)
Deeper omnichannel order management and multi-entity consolidation for chains selling across several countries under one brand.
Watch out: China invoice (fapiao), tax and export-rebate work arrives as partner localisation, and licence plus implementation cost is an order of magnitude above domestic products. Confirm the localisation owner in writing.
Still unverified
These are commonly assumed but we found no official evidence. They are questions for the vendor, not selling points.
- Offline POS operation during a network outage, and how the queued transactions are uploaded afterwards
- Capacity at hundreds of stores: concurrent users, transaction throughput, peak-season load
- Warehouse-to-store and store-to-store transfers with in-transit costing
- Franchise versus directly operated store settlement
- Platform-commission and advertising allocation down to order-line contribution margin
- Native connectors to cross-border marketplaces (Amazon, Shopee, TikTok Shop)
Ask the vendor this
Put these in writing during the demo. An answer that arrives only in the follow-up email is not an answer.
- 1Does the POS keep selling offline, and how are those transactions reconciled?
- 2How many stores and concurrent tills has the largest live customer deployed?
- 3How are on-hand, allocated, in-transit and available-to-promise distinguished in the stock report?
- 4Can marketplace commission and freight be allocated to the order line?
- 5Are franchise and directly operated stores settled differently, and how?
- 6Which cross-border marketplaces have native connectors, as opposed to API work?
What is verified, what is partial, what is unknown
Each requirement is marked against what we could actually read on a vendor’s own page. “No evidence” means we could not verify it — not that the product cannot do it. Ask for a demonstration before you remove a product for that reason.
On a phone, swipe the table sideways to see the evidence column.
| Requirement | Status | Evidence |
|---|---|---|
| Store POS with membership, points and stored value | Verified | HaoYeCai retail capability — store POS, membership / points / stored value, online-offline integration. hyc.chanjet.com, read 2026-10-29. |
| Domestic marketplace order intake and store pickup | Verified | HaoShengYi — connections to Taobao, Tmall, JD, Pinduoduo, Kuaishou and Douyin, plus store pickup and same-city delivery. hsy.chanjet.com, read 2026-10-29. |
| Batch and shelf-life control with scan-based stocktake | Verified | HaoShengYi — multi-store / multi-warehouse, batch and shelf life, barcode receiving and stocktake, smart replenishment. hsy.chanjet.com, read 2026-10-29. |
| One availability figure across store and online | Partial | Purchasing, sales, inventory and cash run on one chain in HaoYeCai, which is the prerequisite. How availability is split between allocated and in-transit is not documented on the official pages. |
| Offline-tolerant POS with queued upload | No evidence | No official statement found on offline operation. Not the same as unsupported — it is unproven. |
| Cross-border marketplace connectors (Amazon, Shopee) | No evidence | Capability matrix records domestic channel integrations only; cross-border marketplaces are explicitly marked as having no evidence. |
| Contribution margin after platform commission and freight | No evidence | No official evidence of fee allocation to order line. Assume period-level reporting until proven otherwise. |
Where these projects actually go wrong
Four areas, the failure each one produces, and the control that prevents it. None of this is vendor-specific — it is what the projects have in common.
Variant explosion turns item setup into a months-long exercise, and every later report inherits the mess.
Control: Fix the variant taxonomy and a naming convention before loading data; cap the matrix and agree who may add attributes after go-live.
Marketplace APIs change without notice, and each channel's order and return semantics differ.
Control: Assign an owner for channel integrations, and require a monitoring alert when an order sync stops — silent failure is the normal failure.
Domestic cloud products are hosted in mainland China, so overseas store or warehouse staff work over a cross-border link.
Control: Measure latency from the actual overseas site before committing, and plan a local buffer process for peak trading days.
Overseas sales tax and VAT registration (EU IOSS, US state tax, UK VAT) is a filing obligation, not a reporting feature, and is outside what a China-standards product covers.
Control: Keep a local tax agent or a dedicated compliance platform, and define which system is authoritative for each jurisdiction.
Selection checklist
Work through this before you accept a demo. It takes an afternoon and removes most of the avoidable risk.
- 1Count SKUs after variant explosion, not before. If style-colour-size takes you past a few thousand, master-data governance is your first project, not the software.
- 2Decide how many channels you will genuinely run in year one — every additional channel multiplies reconciliation work.
- 3Ask to see one day-end reconciliation across store, own site and a marketplace, with your own product mix.
- 4Test the return path: does the item go back to sellable stock immediately, or after inspection? Who decides?
- 5Agree how platform commission and freight are recorded — as cost of the order or as period overhead. The choice determines whether channel margin is trustworthy.
- 6Confirm what happens when the internet drops in a store, and how the queue is cleared.
- 7Pilot one store and one online channel for a month, with parallel reporting against the old process.
What makes Retail & eCommerce ERP different
The single stock number problem
Every retailer says they want 'one view of inventory'. Almost none of them have it, because most systems store a single quantity when the honest answer requires four: what is physically on hand, what is already allocated to orders, what is available to promise to this channel, and what is in transit from a supplier or between stores. Collapse those into one number and the inevitable result is overselling — the inventory page says seven units, seven listed channels each promise the same seven units to different customers.
The fix is arithmetic rather than a feature. Demand allocation rules decide how much safety buffer each marketplace keeps, so a marketplace cannot drain stock reserved for your own store. Transit must be visible as a distinct quantity so it can be promised but not picked. Store-to-store transfers need both legs recorded, because stock that left warehouse A but has not arrived at warehouse B exists nowhere in a single-quantity system. Our inventory management guide breaks down exactly how those four quantities should be calculated and reconciled.
Returns deserve separate mention. A returned item is physically present, but not sellable until inspected, and the timing of that status change silently determines availability accuracy. Retailers who reintroduce returned stock immediately sell goods they do not have; those who quarantine it for three days wonder why their sell-through looks weak.
Unit economics after fees, freight and returns
Revenue reporting is easy. Contribution margin reporting is where retail ERPs are usually found wanting, because the costs that decide whether a given order made money are deducted outside the product cost — marketplace commission, payment processing, fulfilment fees, advertising attribution, return shipping and the write-off on goods that come back unsaleable.
A workable system allocates those costs down to the order line rather than leaving them as period overhead. Only then can you answer whether a promoted SKU sold through your own store was profitable while the same SKU sold through a marketplace at the same price was not. Without that granularity, promotions are decided on revenue and discovered to be loss-making at year end.
Dimensional reporting matters here for the same reason: being able to slice contribution by SKU, channel, campaign and customer cohort without rebuilding the chart of accounts is what keeps the analysis alive after launch. When the finance and operations sides disagree about which channel is profitable, the cause is almost always that one side is looking at revenue and the other at cost.
Cross-border retail: three checks that are frequently missed
Settlement currency differs from listing currency. A Chinese merchant selling on a US marketplace receives USD, pays suppliers in RMB, and pays platform fees in a third currency, so realised and unrealised exchange movements are real margin events rather than accounting formalities. The system must post those differences automatically at settlement and revalue open balances at each period end.
Inventory regimes cannot be mixed. Bonded warehouse stock, overseas warehouse stock and domestic general-trade stock sit under different customs statuses with different tax treatment. Treating them as interchangeable pool quantities produces declarations that do not reconcile with physical counts, which is precisely the kind of discrepancy that triggers inspection rather than a query.
Export rebate depends on documents existing. Cross-border e-commerce modes each have their own declaration and documentation requirements, and where a valid export declaration is absent, the associated VAT rebate generally cannot be claimed. This is a documentation problem long before it is a tax problem — the shipping workflow has to produce the right paperwork at the right time. Our guide to export tax rebate ERP software for China sets out the matching logic, and our buying guide for Chinese export companies covers how settlement and rebate connect to the wider order cycle.
In-depth Retail & eCommerce ERP guides
Practical selection and deployment guidance for retail & ecommerce teams operating across China and overseas.
ERP for Manufacturing: Modules, Features & Selection Tips
What a manufacturing ERP must do — MRP, BOM, shop-floor, quality — across make-to-stock, make-to-order, and engineer-to-order, plus how to shortlist vendors.
Read guide →IndustryERP for Retail & Omnichannel Commerce: Unifying Store, Web, and Warehouse
How retail ERP handles POS, e-commerce, inventory, and merchandising across channels, the integrations that make or break it, and which platforms fit.
Read guide →IndustryERP for Wholesale Distribution: Inventory, Pricing & Supply Chain
Distribution ERP essentials — lot/serial tracking, tiered pricing, dropship, and demand planning — for distributors competing on service level and margin.
Read guide →Compare the shortlist
Head-to-head pages on the same fixed dimensions, with each price carrying its source and the date it was read.
NetSuite vs Acumatica
How these two cloud-native platforms differ in pricing model, customization depth, and mid-market positioning.
Read comparison →Odoo vs ERPNext
Two popular open-source ERPs compared on modularity, community, hosting flexibility, and total cost.
Read comparison →Other industries
The must-have list changes with the operating model, so the shortlist does too.
Frequently asked questions
What is omnichannel retail ERP?
It keeps inventory, pricing, and customer data consistent across physical stores, web, and marketplaces, so every channel shows the same truthful stock number instead of conflicting spreadsheets.
Why does inventory accuracy matter most?
Every omnichannel feature fails if the base stock record is wrong. Treat perpetual cycle counting, clean SKU master data, and disciplined receiving as prerequisites — the software only amplifies the discipline already on the floor.
Do I need ERP or just a POS plus an e-commerce plugin?
POS plus plugins work for very small shops. Once you sell across channels and need real margin and fulfillment control, ERP unifies the silos that otherwise disagree at day's end.
What integrations should I check first?
Your e-commerce platform (Shopify/Magento), marketplaces (Amazon/eBay), and in-store POS. Confirm native connectors or an open API before signing, because your channel mix will change faster than your contract.
How does retail ERP handle promotions and loyalty?
It centralizes promotion logic and a single loyalty balance across channels, so a return in-store credits online points immediately and the brand feels like one company.
Ready to unify your global operations on one ERP?
Book a free scoping conversation with our ERP team. We help Chinese companies going global build a unified system across HQ, overseas factories, and international sales — finance, procurement, inventory, production, and operations in sync.
Book a Free Consultation