How to add a new colour to a product you already carry

It is not an edge case. It is the most frequent decision in receiving, and it recurs every time a vendor refreshes a range. A shop carrying forty styles, where each style picks up two new colourways a season across two seasons, faces roughly one hundred and sixty of these decisions a year. Each one takes a few seconds and each one is either right or permanent.

What makes it dangerous is that the wrong action is the default action. A document arrives describing something your catalog does not recognise, and the path of least resistance is to create it. Nothing warns you, nothing looks broken, and the purchase order totals reconcile perfectly against the invoice. The damage only becomes visible months later in a report that no longer describes reality.

Before creating anything, ask: do I already carry this style? Not this colour, not this SKU, the style.

If yes, this is a new colourway on an existing variant group, and the correct action is to add children to that group. If no, and you have genuinely never stocked this style in any colour, it is a new product.

That second case is far rarer than most receiving processes assume. A vendor changing their style-number format between seasons, or shipping a colour code your catalog spells differently, both look like new products and are not. The question is about the style, and the style usually already exists.

The mechanics differ by system, the structure does not. In Lightspeed Retail you open the matrix and add the colour to its attribute values, which generates a child item for each size in the run. In Shopify you add the option value to the product, which creates the corresponding variants. In Square you add a variation. Whatever the wording, you are extending an existing parent rather than creating a sibling to it.

Two things to get right while you are there. The colour value must match the colourway the vendor actually sells, not a colour word you invented: if they ship Black/White, the value is Black/White, not Black. And the new children need the same treatment as the existing ones, each with its own barcode, cost and price, so counting and reordering work per size.

The decision is easy once you know the style exists. The difficulty is that vendor documents are written in the vendor's vocabulary, and almost nothing on the page matches your catalog literally.

Three fields carry the answer, in descending order of reliability. The manufacturer SKU, if the vendor prints one and you stored it at creation, is exact: the same string on their document and in your catalog, so a lookup either finds the style or genuinely does not have it. The style number is next, but it usually encodes more than the style, commonly season and colourway, so it needs decomposing before it will match anything. A number reading like HYPRFRK-24-BLK is a style root, a season, and a colour code travelling together, and only the first part identifies the style.

The product name is the least reliable and the most used, because it is the field people search. Vendors abbreviate differently from you, word order varies, and a name that includes the colour will never match a catalog record that keeps colour in an attribute.

The practical technique is to search the style root alone, not the full identifier. Strip the season and colour segments and search what remains. If that returns a product, you are adding a colourway. If it returns nothing, widen once to a distinctive word from the product name before concluding the style is new, because a single naming inconsistency should not be enough to create a duplicate.

Two failure signals worth recognising. If a manufacturer SKU lookup returns dozens of unrelated products, that field is holding a brand or category label rather than an identifier, and it cannot be trusted for this decision. And if the style root search returns two products that look like the same style, you have already accumulated a duplicate and are about to add a third: stop and reconcile the existing pair first, because adding to the wrong one compounds the problem rather than fixing it.

None of this requires tooling. It requires knowing that the question is about the style, and that the document is unlikely to state the style plainly.

Three cases where creating is correct. A style you have never carried, in any colour. A style whose attribute set is fundamentally different from the existing group, such as a version that varies on length where the original did not. And a product you deliberately want to report on separately, which is a merchandising decision rather than a data one and should be made knowingly.

Prior-season carryover sits on the boundary. The same style, same colour, new style number, arriving a year later: one record or two is a judgement call about whether you need to compare seasons. Either answer can be right. What is never right is having it resolved by accident because the style number did not match.

Stock splits first. A style with six units on hand reads as four and two across two records, which is enough to trigger a reorder you do not need or suppress one you do.

Then reporting fragments. Sell-through is calculated per record, so a style selling well in one colour and badly in another stops reading as one style with a colour problem and starts reading as two unrelated products, one healthy and one weak. The insight that would have changed next season's buy simply is not visible.

Open-to-buy inherits the same distortion, because committed spend is attributed to a style that the system no longer believes is one style. And counting gets slower in a way nobody attributes to the catalog: staff meet two records for one shelf position and have to work out which is which.

Cleanup afterwards is disproportionately expensive. Creating the duplicate took one careless import; removing it means reconciling two stock counts, two sales histories, and any open purchase orders pointing at either record. That asymmetry is the whole argument for getting the decision right at receiving.

Three things reduce the error rate without changing anyone's habits. Keep the manufacturer SKU populated, because it is the one identifier that appears identically on the vendor's document and in your catalog, and matching on it is exact rather than fuzzy. Record colourways as the vendor sells them so their codes have somewhere to land. And hold one naming convention, so a search for the style finds it regardless of who typed it in.

A fourth helps more than it looks: write the colour-code mappings down. Codes are stable within a vendor and arbitrary across them, so the knowledge that one brand's BWH means Black/White is worth recording once rather than rediscovering every season. A short per-vendor list turns the hardest part of the match into a lookup.

None of that removes the judgement. It just means the person receiving can see that the style already exists, which is the only information the decision actually requires. Most duplicates are not created by someone deciding wrongly; they are created by someone who had no way to see that there was a decision to make.