How to Build a SKU System for Clothing Resale Inventory That Scales Past 500 Items

A good SKU system for clothing resale inventory uses four short blocks: category, source, intake date, and a sequence number. Keep size and color out of the SKU and in separate fields. That keeps the code readable at 500 items and still sortable at 5,000.

Here is the usual failure. You start with SKUs like BLUE-LEVIS-JKT-M-GW-1104. By item 300, half the codes are inconsistent, the label printer truncates the long ones, and you spend 40 seconds per item hunting for a bin. This guide gives you a format that avoids that, plus a way to test it before it breaks.

What a SKU needs to do (and what it should not)

A SKU is a lookup key. It should answer three questions fast: what is this, where did it come from, and when did it arrive. Everything else belongs in a field you can filter on.

• Do encode: category, source, intake date, unique sequence.
• Do not encode: size, color, brand, price, or condition. These change, get corrected, or are already item specifics on eBay.

Why the split matters: if you mislabel a size and it lives in the SKU, you either reprint the label or live with a lie. If the size lives in a field, you edit one cell. Grades work the same way. A tier like Very Good or Good can change after you re-inspect Fabric Condition or Cosmetic Appearance, so it should never be baked into a printed code.

The four-block SKU format

Use this pattern: CAT-SRC-YYMM-0000.

Full example: DN-GW-2611-0042. It is 14 characters, fits on a small label, and you can read it aloud across a packing table.

Pick your category codes once

Limit yourself to 8–12 categories. Common set: DN (denim), KN (knits), OW (outerwear), TP (tops), BT (bottoms), DR (dresses), SH (shoes), AC (accessories). If you need more than 12, you are encoding brand or style, which belongs elsewhere.

Source codes track ROI

A source code lets you pull cost and sell-through by origin later. Use two letters per source: GW for Goodwill outlet, ES for estate sales, TS for a named thrift store, and so on. Keep a one-page key and do not add a code until you have bought from that source twice.

Date and sequence

Use year and month, not the full date. YYMM sorts correctly as text and tells you at a glance how old a piece is, which feeds your aged-inventory checks. The sequence resets by whatever batch rule you choose, but pick one: either per month overall, or per category per month. Do not mix.

A naming rule for variants like size and color

Clothing rarely has true variants. Each garment is one unit, so each gets its own SKU. The trap is multiples: you buy six identical new-with-tags tees, same size and color. Here is the rule.

1. One-of-one garment: one SKU, one sequence number.
2. Identical multiples (NWT or NWOT, same style, size, color): one parent SKU plus a suffix, such as TP-TS-2611-0050-A, -B, -C.
3. Different size of the same style: new parent SKU. Do not use the suffix for size.

Size and color live in your item fields, written the way eBay item specifics expect them: "Size: M", "Color: Navy". Keeping one clean value per field means filtering for every Medium in outerwear takes seconds, no decoding required.

A naming convention for the listing title is a separate job. Titles are for buyers and search. SKUs are for you. Putting your SKU in the eBay Custom Label field ties the two together without cluttering the title.

Label format that survives packing and returns

The label is where most systems die. Use this layout on a 1 x 2.625 inch or similar small label:

• Line 1: SKU in large type (human-readable).
• Line 2: barcode (Code 128 handles letters and digits).
• Line 3: bin location, such as A3-07.

Rules that save time:

• Attach labels to the care label or a hang tag, not directly to fabric. Adhesive on knits and delicate fabrics can leave residue and hurt Cosmetic Appearance.
• Print a spare label for the poly mailer or bin sheet.
• Avoid the characters O/0 and I/1 confusion by using digits only in the sequence block and letters only in the category and source blocks.

Returns are where this pays off. When a package comes back, scanning one code pulls the original cost, grade, and sale. That matters for reconciliation, and it shortens the case when a buyer claims an item was not as described. An item you graded as Excellent or Very Good should have its Structural Integrity and Odor & Cleanliness notes attached to the same SKU record, so you can compare what you shipped to what came back.

Build the system in 7 steps

1. List your categories and cut them to 12 or fewer, each with a 2–3 letter code.
2. List your active sources and assign two-letter codes; write the key on one page.
3. Choose the date format (YYMM) and the sequence rule (per month, or per category per month).
4. Set up fields for size, color, brand, grade tier, and bin location, separate from the SKU.
5. Define the multiples rule: parent SKU plus an A, B, C suffix for identical units only.
6. Choose a label layout with SKU text, Code 128 barcode, and bin location; print a test sheet of 20.
7. Run the throughput check below on 25 real items before you label the backlog.

Screenshot placeholder: [FlipDesk intake screen showing category, source, and auto-generated SKU with bin field]

The throughput check: will it stay fast?

A SKU system is only good if intake stays quick. Test it with a timed batch.

1. Pull 25 items from a mixed pile.
2. Time each from "in hand" to "labeled and binned".
3. Divide total time by 25.

Example math, as an illustration: if the batch takes 25 minutes, that is 60 seconds per item. At 100 items a week, labeling costs 100 minutes. At 500 items a month, about 8.3 hours. If you can get to 40 seconds per item, you recover roughly 2.8 hours a month at that volume (500 x 20 seconds).

Then check three failure signals:

• Lookup time: finding an item by SKU should take under 30 seconds. If not, your bin scheme or sort order is the issue.
• Error rate: count mislabeled or duplicate SKUs in the 25. More than one means your format has a confusing block.
• Decode time: hand someone the code and ask what it means. If they need the key for more than the source block, shorten it.

Repeat the check at 250, 500, and 1,000 items. Sequence digits give you 9,999 items per scheme, which is plenty for most solo sellers. If you run past that in a month, add a fifth digit rather than a new block.

Multi-platform sellers: one SKU everywhere

If you crosslist to Poshmark, Mercari, or Whatnot, use the same SKU on every channel. Put it in eBay's Custom Label, and in the description or private notes where the other platforms allow. When a sale lands, you search one code, pull the item, and end the other listings. That is how you avoid double selling without a second system. Never create a per-platform SKU. It doubles your reconciliation work for no gain.

Try it on one batch

You do not need to relabel your whole stockroom. Take your next 25 intake items, apply the four-block format, and time it. If you grade as you go, record the tier (NWT, NWOT, Excellent, Very Good, Good, Fair, or Poor) in a field against the SKU rather than in the code. FlipDesk generates SKUs at intake, prints the label, and ties cost, grade, and bin location to one record. Try one batch and see what your per-item time looks like.