Half the variants sit at the bottom of the catalogue: when should you simplify the range?
Before simplifying a range, examine the role of a rarely purchased variant. Consider the completeness of the family, support for existing sets and the customer's path after the change.
Article language: English

In this article
Description: Before simplifying a range, examine the role of a rarely purchased variant. Consider the completeness of the family, support for existing sets and the customer's path after the change.
The last item in a sales ranking is a tempting candidate for removal. It would no longer take up space in the catalogue, on the shelf or in team discussions. Before it disappears, check whether someone buying the other products needs it to complete the whole task. A rarely ordered item can be a small part of a much larger choice.
Imagine a manufacturer of jewellery display trays for cabinets. It offers four widths in two finishes, eight combinations in all. Four are ordered occasionally and the others regularly. The narrow tray sells much less often than the larger ones, so the company is considering withdrawing it. Some shops, however, have a narrow side section in their cabinets. When furnishing an entire shop they need different trays in the same finish, not just the most popular size.
We do not yet know whether there is a reason to keep this item. We do have a reason not to remove it on the basis of one ranking. The decision concerns the variant's role and the work required to keep it in the range. Designing how customers choose it on the website comes later, once the company knows what it intends to keep selling.
A rare purchase may complete a larger list
Look at orders containing the narrow tray. Does the buyer add it to a larger set? Order it alone to extend an existing display? Or use it for a purpose the manufacturer did not consider in its description? Each answer changes the question about the product's future.
If a shop orders many wide trays and one narrow one, the latter accounts for a small share of units. It may nevertheless determine the choice of one supplier for the whole cabinet. Do not assume that relationship without checking. Ask the buyer what they would do without the item, and distinguish their answer from actual purchase history.
For an addition, the customer may need compatibility with a previously received set. A different finish or a similar but larger size may not be equivalent. A new shop may have more freedom to choose the entire display. A shared customer identifier does not reveal whether the purchase is a first set, an extension or a replacement for a worn tray.
Read low sales alongside the situation of use. An item may be needed during an infrequent refit while others are ordered more often. It may also remain in the catalogue without a justified role and generate mainly availability questions. Neither story follows automatically from its place at the bottom of a list.
In a 2001 study, Peter Boatwright and Joseph Nunes analysed a natural experiment by an online grocery retailer. The original abstract distinguishes reducing the number of items from reducing the attributes available in the range and describes different responses to their removal. It concerns consumer categories. It does not determine how many sizes a cabinet equipment manufacturer needs.
For the trays, ask which attribute disappears with the product. Can the narrow section still be fitted out? Or is only one of several similar shades lost? The number of removed items can be the same while the change in use is completely different. Counting catalogue entries alone will not show that.
Do not assign an exceptional mission to every variant either. If no identified need requires a particular width and the customer confirms another is useful, the company may have a basis for simplifying the family. It still needs a reason tied to the product and its users. “It sells least” describes a result, but does not yet explain the effect of a change.
A product family differs along more than one axis
Dimensions, equipment and finish can meet different needs. AUER Packaging's own page on Euro containers, for example, shows solid and perforated containers, versions with lids and a choice of colours. This is a specific manufacturer's range. It reminds us that two similar-looking items can differ in more than the appearance of a photograph.
For the trays, separate width from finish. Withdrawing the narrow version in one colour may preserve the use itself but break the family's consistency for a particular set. Removing every narrow version changes the uses available. The company needs to know which decision it is making before the team says that “only one variant is disappearing”.
A family map helps: what is the common base, what differs between items and which elements are bought together? This is material for a business decision. A product catalogue content model (in Polish) will later be needed to store and display the chosen structure correctly. It should not decide by itself whether a size remains in the range.
In a family of eight combinations, removing four could mean giving up one finish across all widths. It could instead mean keeping two widths in both finishes. In the first arrangement the customer can still choose a width but loses a particular appearance. In the second they keep a choice of finish but may not be able to complete the dimensions. The list becomes equally short, yet the uses of the remaining family change differently. Show the team both arrangements before approving the reduction, even if the original plan was simply to cross out the last rows.
Check names in documents too. The team may sell an older version under a new abbreviation or use several names for one finish. The ranking then splits one product into several entries. Before assessing the bottom of the catalogue, make sure you are comparing actual variants rather than the history of naming and order registration.
When choosing a variant in a shop (in Polish), a combination that does not exist differs from one temporarily unavailable. That distinction matters here too. Low sales of a tray that could not be ordered for part of the period are not the same result as low sales of a continuously available tray. Record the conditions under which orders were observed.

Do not confuse absence from the catalogue with absence of demand. If a variant can be ordered only through a dedicated account manager, a new customer may not know about it. The company should identify that condition before concluding that nobody needs it. It does not have to create demand for every historical idea either. First establish what was actually offered and to whom.
There is more to simplify than the product list
The company can keep the narrow tray but change how orders for it are accepted, limit finishes or make it an item requiring prior agreement. It can also stop selling it. The choice should reflect the real capacity to make and support it. Moving a product from the shelf to a contact form does not itself remove the work needed to deliver it.
Map that work against a specific order. Who prepares the material, what must be changed over, how is the correct version identified and who checks the result? A rare variant may share the same base as a common one or require separate preparation. Only a comparison of the workflow will show the operational gain from removing it.
It is not enough to say it saves warehouse space if the tray is made only to order anyway. Nor is it enough to call production simple if it needs separate material and checking that the short description does not reveal. The team should distinguish actual work from reasons that sound good in a meeting.
Individual versions bring back the question of one customer's influence on service development (in Polish). A manufacturer may deliberately support a special variant for particular buyers. It must know what it is maintaining and why. It should not present it as an ordinary standard while treating every order as a surprising exception.
If the variant remains only for additions to existing sets, define the support. Does it concern earlier customers, a particular series or identifiable sets? The article on parts and support for an older series (in Polish) develops a similar distinction between an active offer and continued support for existing products. Hiding the purchase button does not explain that boundary.
Once the decision is approved, Aeneia Design can help organise the catalogue and the customer's next path. It needs to know which variants remain, what has ended and what contact is useful for an earlier set. The design should show the company's actual options without promising that products can automatically substitute for one another.
After removing an item, watch what remains too
Assess the reduction over a period suited to purchasing cycles. No new orders for the withdrawn tray is an expected consequence of ending the offer. It does not answer what happened to orders for complete sets. You cannot judge the withdrawal's success solely by the absence of further units of the removed product.
In a study published in 2006, Sloot, Fok and Verhoef analysed a detergent assortment reduction at a Dutch retailer. They compared two stores affected by the change with two controls, distinguishing earlier buyers of removed items from other groups. Short- and longer-term results differed. The authors noted the limitation to one category and a small number of stores.
Read several different events in your own catalogue. A customer may buy another tray, postpone buying the whole display or order only part of a set. They may also choose another supplier. Do not attribute that last story to every missing order. If the customer's decision is unknown, keep it as unknown and try to understand the reason.
Before the change, return to a few identified buyers of the narrow version. Ask about their next task: a new location, a refit or an addition. A declared attachment to the product does not establish when they will next buy. The conversation can nevertheless reveal a dependency to consider when planning the transition.
Record the timing and scope of the change too. If a colour disappeared with the variant and the family name and ordering process also changed, a later sales difference does not describe the effect of width alone. A company can decide on several changes at once, but should honestly describe what it cannot separate between them.
Leave room in the assessment for a signal that would change the decision. It could be repeated cases of being unable to complete a display or previously unknown work in supporting the remaining variants. The condition for reopening the discussion should follow from the purpose of simplification. One person's dissatisfaction does not settle the whole issue, just as the first trouble-free order does not close it.
The customer should recognise what happened to their choice
Withdrawing a variant need not mean immediately removing all information about it. Someone may return with an old order number or a saved product page. They need confirmation of which item they meant and which options are available now. Silently substituting a wider tray obscures their intent.
If the company suggests an alternative, it should explain the material difference. In our example, greater width may make it impossible to use the tray in the cabinet's side section. A shared finish or a similar photograph is not enough. Product dimensions and fit (in Polish) help relate the proposal to the space for which the customer is buying equipment.
Give existing customers the same answer the catalogue shows. If an account manager can still agree production, the site should not categorically announce that the variant will never be available. If sales have ended, a form should not suggest an ordinary order. This is one company decision presented through several channels.
Finally compare two descriptions: “we are removing the least-sold size” and “we are keeping a family that serves the specified uses and discontinuing this method of production”. The second requires more investigation. It lets the company check whether the simplified catalogue still solves the customer's task. Sometimes the last item really will prove unnecessary. Sometimes it held together several much better-looking entries in the ranking.







