Kitting and Assembly in Germany and Poland: How Contract Packing Speeds Up Orders
13.07.2026
EU Battery Directive Compliance at German Amazon Warehouses: What Changes at FC Intake
13.07.2026
FLEX. Logistics
We provide logistics services to online retailers in Europe: Amazon FBA prep, processing FBA removal orders, forwarding to Fulfillment Centers - both FBA and Vendor shipments.
A Xentral user in Cologne sells across Amazon.de, a Shopware storefront, and a wholesale B2B channel. Stock counts live in Xentral, but the warehouse floor updates a spreadsheet, and Amazon listings get refreshed by hand twice a day. On a busy Tuesday, an order comes in for the last four units of a bestseller at the same moment a return is being restocked and a pallet is being picked for dispatch. Nobody notices until Amazon shows negative available quantity and the order queue backs up behind it. This is what happens when Xentral ERP integration is treated as optional rather than as the backbone connecting inventory, order status, and marketplace visibility. The fix is not a better spreadsheet habit. It is a direct data feed between Xentral, the warehouse, and Amazon that updates stock, pushes orders, and pulls tracking back automatically, so the reader can decide whether manual sync is still workable at their current order volume or whether it is already costing more than it saves.
What a Xentral-to-3PL Feed Actually Connects
A working integration is three data flows moving in a loop, not one generic connection. First, inventory sync: the warehouse management system reports real stock levels back into Xentral at intervals short enough that Amazon, Shopware, and any marketplace connector see the same number. Second, order push: when an order lands in Xentral from any sales channel, it is transmitted to the warehouse as a pick instruction without a human retyping SKUs or quantities. Third, tracking pull-back: once the carrier scans a parcel, that tracking number and dispatch timestamp flow back into Xentral, which then reports it to Amazon or the storefront to close the order.
Each of these three flows can run through Xentral’s native API, or through a middleware layer such as ChannelEngine integration when multiple marketplaces are involved. For sellers also running Plentymarkets fulfillment or Billbee integration fulfillment alongside Xentral, the same three-flow logic applies, just with a different connector doing the translation between systems. The point is not which tool sits in the middle. The point is that stock, orders, and tracking move on one feed instead of three separate manual routines.
What the Operator Must Control
The seller, not the 3PL, usually owns the SKU master data in Xentral: product codes, FNSKU mappings, bundle logic, and channel-specific listing rules. If a SKU is renamed or a bundle is restructured in Xentral without notifying the warehouse system, the sync will either fail silently or push the wrong item against an order. The operator also owns the sync interval decision. A 15-minute inventory sync behaves very differently from a nightly batch when a bestseller sells through in an afternoon flash sale. Choosing that interval is a commercial decision, not a technical afterthought, and it should sit with whoever owns Amazon Buy Box performance, since stockouts and overselling both damage account health metrics that are hard to recover.
What Breaks Without This Control
When SKU mapping drifts or sync intervals are too wide, the visible symptom is overselling: Amazon accepts an order for stock that has already been picked against a different channel. The seller then has to cancel the order, which can trigger a defect rate hit on the account, or scramble to source replacement stock at a worse margin. On the other side, understocking the ERP view causes phantom stockouts, where Amazon suppresses a listing even though physical stock sits on a shelf. Both failures show up as the same root cause during a review: the ERP shipping integration was treated as a nice-to-have rather than a control point tied to revenue.
The Handoff Point Most Sellers Miss
The single most common failure is not the sync itself, it is the moment stock is physically moved without a corresponding system event. A pallet gets relocated during a warehouse reorganization, a return gets restocked into a different bin, or a removal order pulls units out for disposal, and none of these actions generate a transaction in Xentral. The system then reports a stock level that has no relationship to reality. Before scaling order volume, confirm that every physical movement in the warehouse, not just sales and receiving, has a matching event that flows back into the ERP.
Manual CSV Uploads Versus Automated Sync as Volume Scales
Many Xentral users start with manual CSV exports: pull a stock report, upload it to Amazon Seller Central, repeat daily. At ten or twenty orders a day, this is tedious but survivable, because a human can catch obvious errors before they become customer-facing problems. The economics change once order volume climbs past roughly fifty to a hundred orders a day across multiple channels, because the CSV process cannot compress into the time available between sales spikes and dispatch cut-offs.
The commercial case for automated sync is not about elegance, it is about the cost of the errors manual process cannot catch fast enough: a cancelled order because of oversold stock, a late dispatch because an order sat in an export queue for six hours, or a customer service ticket because tracking never made it back to the marketplace. Each of these has a direct cost in refunds, account health penalties, or agent time. Automated sync trades a fixed integration cost for a recurring reduction in these failure rates, and that trade becomes favorable well before most sellers realize it.
Signals It Is Time to Automate
Order volume that regularly exceeds what one person can manually reconcile in under an hour each morning is the clearest signal. A second signal is running more than two sales channels feeding the same physical stock pool, since manual reconciliation across channels compounds the chance of a mismatch. A third is any recent overselling incident that triggered an Amazon policy warning, since that is a lagging indicator that the current process has already failed at least once.
What Stays Manual Even With Integration
Automation does not remove the need for a human owner to reconcile exceptions: a failed API call, a SKU that does not map cleanly between Xentral and the warehouse system, or a bundle product that needs manual kitting instructions. Shopware fulfillment 3PL setups in particular still need someone checking that promotional bundles and multi-packs are represented correctly on both sides. The goal of integration is reducing routine manual work, not eliminating the need for a named person who watches the exception queue daily.
Who Owns Which Part of the Feed
In a typical setup, the seller owns product data and channel listing rules inside Xentral. The 3PL or warehouse system owns physical stock accuracy and pick/pack execution. The integration layer, whether native API or middleware, owns the translation and timing between the two. When something breaks, the fastest diagnostic question is which of these three owners last touched the affected SKU, because that narrows the failure to a data entry error, a physical handling error, or a sync timing error almost immediately.
Where Integrations Quietly Fail Even When They Look Connected
A common blind spot is treating the connection as binary, either it is integrated or it is not, when in practice partial integration is the norm and the riskiest state. A seller might have order push working perfectly but tracking pull-back running on a delayed batch job, which means Amazon shows orders as unshipped long after the parcel left the building, quietly degrading delivery performance metrics. Another quiet failure is currency or unit mismatch between systems: Xentral tracks stock in individual units while a 3PL system tracks in case packs, and a rounding assumption buried in the integration config slowly drifts the numbers apart over weeks.
Multi-marketplace sellers running ChannelEngine integration alongside Xentral face a related risk: each marketplace connector may have its own sync cadence, so Amazon.de might update every ten minutes while a secondary marketplace updates hourly, creating brief windows where the same unit is technically available for sale on two channels at once. None of these are failures of intent, they are failures of nobody owning the end-to-end timing map across every connected system. The practical mistake is assuming that because data is flowing, it is flowing correctly and at the same speed everywhere it needs to.
Check Before Go-Live
- SKU and FNSKU mapping matches exactly between Xentral and the warehouse system
- Sync interval is set based on actual order velocity, not a default setting
- Bundle and multi-pack products have explicit kitting rules on both sides
- Currency and unit-of-measure definitions match across every connected system
Check After Go-Live
- Tracking pull-back timing matches actual dispatch scans, not a delayed batch
- A named person reviews the exception queue daily, not just on error alerts
- Stock counts reconcile weekly between Xentral and physical warehouse counts
- Removal orders and returns generate a stock event, not just a note in a spreadsheet
Sequencing the Integration Without Stopping Order Flow
The practical rollout order matters more than the technical architecture. Start with inventory sync alone and run it in parallel with the existing manual process for at least one full order cycle, comparing the automated feed against the manual export daily to catch mapping errors before they touch a live order. Once inventory sync is trusted, turn on order push for a single channel first, usually the highest-volume one, rather than activating every channel simultaneously, because isolating one channel makes it far easier to trace the source of any mismatch.
Tracking pull-back is usually the last piece to activate, since it depends on both of the earlier flows working reliably and on the carrier scan data being consistently available from the warehouse. Throughout this sequence, keep the manual fallback process documented and ready, because a partial integration that fails silently is worse than a manual process everyone knows is imperfect. Sellers moving from Plentymarkets fulfillment or a Billbee-based setup to a Xentral-centered feed should expect this same three-stage sequence, adjusted for whichever connector sits in the middle.
A Field Example Worth Copying
One Xentral seller running Amazon.de plus a Shopware storefront staged their rollout over six weeks: two weeks of parallel inventory sync validation, two weeks of single-channel order push on Amazon only, then two weeks adding Shopware order push and full tracking pull-back. The staged approach caught a bundle-mapping error in week three that would otherwise have caused mis-picks during a promotional spike. Slower rollout, fewer live incidents.
Inventory Sync
Warehouse stock levels feed into Xentral on a set interval, giving every connected channel the same available quantity.
Order Push
Orders from any channel route to the warehouse as pick instructions without manual re-entry of SKUs or quantities.
Tracking Pull-Back
Carrier scan data returns to Xentral and reports to the marketplace, closing the order loop without a manual update.
Deciding Whether Manual Sync Still Works at Your Volume
The decision point is not whether integration is technically possible, most Xentral setups support it. The decision point is whether the current manual process is still cheaper than the cost of its own failures: cancelled orders, account health warnings, and the hours spent reconciling stock by hand each week. If order volume has crossed the point where one person cannot confidently reconcile channels every morning, or if a recent overselling incident already triggered a marketplace warning, that is the signal to move from CSV uploads toward a proper Xentral ERP integration with the warehouse and Amazon feed. Sellers evaluating fulfillment tools and services alongside this decision should map out which of the three flows, inventory, order push, or tracking, is currently the weakest link before choosing a connector.
If overselling incidents, delayed tracking updates, or manual reconciliation hours are becoming a recurring pattern in your Xentral setup, FLEX. Logistik works with German and DACH sellers to connect ERP systems, warehouse operations, and Amazon feeds into one coordinated flow. Get in touch to walk through which handoff in your current process needs fixing first, whether that is inventory sync timing, order push mapping, or tracking pull-back reliability.
