The most common mistake in integration projects is the order of steps. The order flow is enthusiastically opened, mapping is left for later, and problems begin on day one.
An order for an unmapped product falls into your system without knowing which item it belongs to. Stock does not decrease, no reservation is created, and the same product is sold again on another channel.
In this article, we explained why mapping should be done first, how it should be set up, and how to manage unmapped products.
Table of Contents
What is mapping?
Mapping is linking your stock card to a listing on a sales channel. Once this link is established, both sides point to the same product.
Without this link, the system cannot know which product an order coming from the channel belongs to. The order is recorded, but it has no counterpart on the stock side.
Similarly, stock notifications cannot be made either. The system cannot know what quantity to send to the channel; the listing is not updated.
In other words, mapping is the load-bearing structure of the integration. All other functions are built upon this link.
We discussed its place in the overall setup in the marketplace integration setup article.
Why should it be done first?
The importance of the order comes from the fact that the mistake is irreversible. Every day that passes without mapping produces a pile of records that need to be corrected.
The first result is stock deviation. Products that are sold but not deducted from stock continue to appear as available in the system and are sold again.
The second outcome is cancellation and point loss. Orders cancelled due to overselling lower marketplace performance and take a long time to recover.
The third outcome is retroactive correction work. Finding and correcting wrongly placed orders one by one takes much longer than doing the matching from scratch.
Therefore, the rule is clear: the order flow is not opened before the matching is completed. The wait may take a few days, but in return, weeks of correction work are prevented.
Selection of the matching key
Matching is established through a common identifier found on both sides. The selection of this key directly determines the automatic matching rate.
The most reliable key is the barcode. A barcode produced to international standards uniquely identifies the same product on both sides.
The second option is the stock code. If the seller code field is filled in the marketplace, matching can be done through this field.
Matching via product name is a last resort and is not reliable. Minor spelling differences break the matching or link the wrong product.
Having an organized barcode structure makes this step easier; barcode and multi-barcode management see the article.
Matching at the variant level
In products with variants, matching should be done at the variant level, not the model level. This distinction is often skipped and comes at a high cost.
When matching is done at the model level, the stock of all sizes is aggregated and reported. Even when one size runs out, the product continues to appear on sale.
The result is that the customer who ordered that size faces a cancellation. This is both a loss of sales and a loss of reputation.
In the correct structure, each variant has its own barcode and its own matching. Stock reporting is also done on a variant basis.
We discussed how to set up the variant structure in the variant product stock tracking article.
Managing unmatched products
Automatic matching is never one hundred percent. A separate work list must be kept for the remaining products.
Reasons for non-matching are generally in three groups: missing barcodes, use of different barcodes, and products not on the channel.
Missing barcodes are resolved on your end; the card is completed. In the case of different barcode usage, the multi-barcode definition comes into play.
A decision needs to be made for products not on the channel: will they be listed or not? This is a commercial decision and must be clearly marked on the matching list.
The order flow should not be opened until the list is cleared. This discipline is the most important rule of the setup.
Keeping the mapping up to date
Mapping is not a one-time task. It requires maintenance as new products are added and old ones are closed.
Making sure channel mapping is also done when a new product is launched should become a step in the product launch process. Mappings done later are often forgotten.
Mappings of discontinued products should also be removed. Otherwise, products that are no longer sold will continue to be listed on the channel.
When a barcode changes due to packaging changes, the mapping must be checked; the link established with the old barcode may break.
A monthly check is sufficient. The number of unmapped products should remain close to zero in this check; integration error monitoring see the article.
Checklist
We recommend a five-item check before opening the order flow.
One. Is the unmapped product list empty? If not, every row must have a decision.
Two. Is mapping established at the variant level for products with variants?
Three. Are stock quantities up to date? Was a count done before setup? Warehouse count see the article.
Four. Is the warehouse it will be fed from and the reservation share defined?
Five. Have channel prices been checked? Once these checks are complete, the order flow can be safely opened; definitions are kept in the e-commerce module .
Executing the matching process
Product matching is the most underestimated and time-consuming step of integration projects.
You should start by cleaning your own product data. A dirty source makes matching difficult right from the start.
Next, determine which field will serve as the key. Barcodes are the most reliable option, with stock codes coming in second.
Auto-matched records are separated from unmatched ones. Review is conducted exclusively on the second group.
Products with variants must be handled separately; color and size breakdowns are the group that generates errors most frequently.
Once the work is complete, verification should be done via sampling, and randomly selected records must be checked one by one.
Points to consider
Incorrect matching is the most expensive integration error because it leads to wrong product shipments and returns.
Therefore, synchronization should not be turned on before matching is complete. An integration launched in a hurry creates days of corrections.
A rule is also needed for new products; unmatched products must not be sent to the platform.
Storing matching records saves time during future platform migrations.
In cases of product merging or splitting, the matching must be reviewed again.
Once this discipline is established, the vast majority of integration-related errors disappear.
Frequently asked questions
How long does the matching take?
If your barcode structure is well-organized, the majority is completed automatically. The remaining manual list varies from a few hours to a few days depending on the number of products.
What happens if incorrect matching is done?
The stock of the wrong product is reported and the wrong product is shipped. Therefore, a sample check after matching is recommended.
Can the same product be linked to multiple listings?
On different channels, yes. If there are duplicate listings on the same channel, that must be fixed first.
Can matching be done with supplier codes?
It is possible if multiple codes are defined; supplier cards and multiple codes check out the article.
Product matching is the boring but decisive step of integration. When done right, it remains invisible; when skipped, it creates weeks of correction work.
The rule in a single sentence is this: the order flow does not open until the mapping is complete. This single discipline prevents the vast majority of setup issues right from the start.
It is also important not to see mapping as a one-time task. When added as a step to the new product creation process, the backlog never accumulates.
By consulting with the EQLEM team you can plan your product mapping work.