LED Mask or Panel? A Distributor’s Assortment and SKU Rationalization Guide
A channel assortment should cover distinct customer needs, not maximise the number of products on the shelf. Every SKU carries a fixed cost that never appears on the price list: forecast uncertainty, working capital, listing work, spares, training and a documentation set of its own. The question is not “should we carry both a mask and a panel” but “does this SKU serve a need the others cannot, and can this channel support it”.
Table of Contents
- 1. An Assortment Covers Needs, It Doesn’t Maximise Count
- 2. What Justifies a Separate SKU
- 3. The Full Cost of Complexity
- 4. The Rationalization Scorecard and Its Triggers
- 5. Partner With Rainbow Technology for Custom LED Devices
- 6. References & External Links
1. An Assortment Covers Needs, It Doesn’t Maximise Count
Assortment decisions fail in two directions, and the second one is more common than the first. A range that is too narrow leaves whole customer groups unserved — the buyer who needs a panel because they want to treat a shoulder, not a face. A range that is too broad looks responsive and quietly spreads the same support capacity across twice the products.
The second failure is harder to see because nothing about it is visibly broken. The catalogue looks fuller. Every line has a listing, a picture and a person who can answer questions about it — except that the person now answers those questions for six variants instead of three, and the documentation set has six versions that must all stay current.
A useful test before any addition: name the customer who cannot be served by what you already carry, and say what they would buy instead. If the answer is a preference rather than a need, the SKU is a merchandising decision being made with an inventory budget.
2. What Justifies a Separate SKU
Six additions come up repeatedly in this category. Each has a version of the argument that holds and a version that is really about how the range looks on a page — and the two are easy to confuse in a meeting where nobody has to carry the stock.
| Proposed SKU | What Would Justify It | What Would Not |
|---|---|---|
| A panel alongside a mask | A segment the mask genuinely cannot serve — a larger treatment area, a standing routine, a clinic setting. | “Customers like having options”, with no evidence of a separate buying context. |
| A second model at a different price | A documented difference a customer can perceive or verify — output, modes, materials, warranty length. | A different colour, a different box, and the same bill of materials. |
| A size variant | Fit data showing the primary size excludes a material share of your audience. | A hunch that small, medium and large looks more professional in a range. |
| A travel or compact version | Evidence that the portability constraint stops purchases you would otherwise make. | A smaller carton around the same device. |
| A market-specific configuration | A labelling, plug or regulatory requirement that genuinely differs by jurisdiction. | A translation that belongs on the same SKU’s multi-language label. |
| A bundle or kit | A basket pattern in your own sales data, where the same items are repeatedly bought together. | A supplier’s promotional idea, imported without testing. |
Where two products do serve the same need, the merchandising problem becomes explicit: you have to explain what differentiates them, or you have created two answers to one question and left the customer to choose arbitrarily. That is usually where returns begin.
3. The Full Cost of Complexity
Unit cost and margin are the two numbers everyone brings to the meeting. Seven others arrive afterwards, and they are all consequences of the SKU count rather than of any individual product.
- Forecast uncertainty, multiplied. Each variant needs its own demand estimate, and each estimate is wrong in its own direction. More lines mean more ways to be overstocked and understocked at once.
- Inventory exposure. Working capital sits in the slowest variant, not the average one — which is where the average is least informative.
- Listing and content work, per variant per channel. Images, copy, specifications and attributes are rarely reusable across variants, and every marketplace has its own upload.
- Spare parts and warranty handling. Straps, inserts, stands and cables wear out, and a variant that shares no parts with its neighbour doubles what must be stocked or refused.
- Training and support load. Staff who must explain six products explain each one less well, and the difference shows up as avoidable contacts.
- Returns profile. Overlapping variants invite the wrong purchase, and the wrong purchase comes back.
- A documentation set of its own. Specifications, instructions, warning language and market documents belong to a model, not to a product family. A distributor cannot substitute a visually similar model’s paperwork, and assuming otherwise is the fastest way to hold stock you are not allowed to sell[1].
That last point deserves the emphasis it rarely gets. Before listing anything, confirm that the supplied specifications, instructions, warning text and claims are appropriate for the channel and the jurisdiction you are selling into — and that they name the model in the box[2]. Documentation is a gate, not a line item that can be caught up on later.
On the numbers themselves: use your own sell-through and return history, and where no history exists, label the assumption and test it in a limited launch rather than extrapolating from a category average. We deliberately do not print inventory or margin figures here, because a number that is not tied to your channel’s data is worse than no number at all.
4. The Rationalization Scorecard and Its Triggers
Six criteria, scored for every line you carry and every line you are asked to add. The score matters less than the discipline of applying one set of criteria to both directions.
- Distinctiveness. The named customer need this line serves that no other line serves.
- Evidence readiness. Whether the model-specific documentation, warnings and claims are complete and current — a yes/no in practice, whatever the score says.
- Expected demand. From your data where it exists, from a stated and testable assumption where it does not.
- Gross margin. After freight, packaging and the support load this specific line creates.
- Inventory risk. Capital tied up per unit, shelf life where relevant, and how easily the line can be reordered or returned.
- Service burden. The number and difficulty of the questions this line generates compared with its neighbours.
Then set the triggers that force a review, rather than waiting for an annual range meeting to notice. Four are enough in most channels: sell-through below plan across two cycles, returns above the channel’s norm, two lines whose customer profiles overlap in your own data, and repeated documentation gaps on the same model. The last one is a supplier conversation, not a range decision, but it belongs on the same list because it costs the same shelf space.
A disciplined range is not a smaller range for its own sake. It is one where every line can be explained in a sentence, supported by one team, and defended with documents that actually exist — and in this category, that combination tends to be both more profitable and considerably easier to sell.
Planning a range across masks and panels? Visit our
Custom Light Therapy Manufacturing Page.
5. Partner With Rainbow Technology for Custom LED Devices
A supplier who pushes a wide range is describing their catalogue, not your channel. Rainbow Technology runs ODM and OEM programmes from our Shenzhen facility and would rather help you keep the range defensible:
- A document set per model, not per family — specifications, instructions and warning text that name the exact configuration, so a listing cannot accidentally rely on a neighbour’s paperwork.
- A clear statement of what is shared and what is not between variants: which parts, which test reports, which labels. That single conversation usually removes one SKU from the plan.
- Spare-part and warranty terms in writing, so the service burden of a line can be costed before it is listed rather than after the first return.
- Honest guidance on where a narrow range is the better commercial answer, including when two proposed variants would compete for the same customer.
- Sample and small-batch production from 500 units, with a stated capacity of 100,000 units per month; both figures depend on model, configuration, packaging and schedule, and are confirmed in a formal quotation.
- Compliance documentation matched to the exact model for your target market. Where a model holds a 510(k) record we provide the K-number so you can check its scope in the FDA’s public database; clearance is not the same as FDA approval, and the scope differs from model to model.
6. References & External Links
[1] International Organization for Standardization.
ISO 13485:2016 Medical devices — Quality management systems — Requirements for regulatory purposes. The document-control and traceability requirements behind the rule that a model’s records belong to that model; incorporated by reference in the FDA Quality Management System Regulation that took effect on 2 February 2026.
[2] International Organization for Standardization.
ISO 20417:2026 Medical devices — Information to be supplied by the manufacturer. Second edition, published March 2026, replacing ISO 20417:2021; it sets the general requirements for identification, labelling, packaging and accompanying documentation, including instructions for use.
