Most shops importing from 1688 expand by adding whatever looks hot at the moment. Six months in, the warehouse holds twenty SKUs, half of them slow, cash is locked up, and the owner spends the day chasing orders instead of reading numbers. This is about the right sequence for scaling, so the catalog grows without breaking cash flow or making operations unmanageable.
Why you need a winning SKU before expanding
Expanding before you have a stable base is the biggest capital mistake in the early stage. It sounds obvious, but many shop owners start adding SKUs when the first one is only "okay," not actually winning.
A winning SKU in practical terms means: selling consistently for at least six to eight weeks in a row, with margin after all fees (platform fee, ads, returns, shrinkage) still positive at above fifteen percent, and a sell-through rate that does not require you to constantly adjust ads or price just to keep it moving. If the SKU sold for one week then slowed, or only earns margin when discounted, that is not a winner. That is a test still in progress.
The reason you need a stable base before expanding: when you add a new SKU, you are using cash from the current one to fund the import, and the new SKU has its own test window before it generates any revenue. If the original SKU is also unstable, you are squeezing cash flow from both sides at once.
Step one: understand your winner before replicating it
Before searching for new SKUs, spend real time understanding the one that is working. The questions are not "why does it sell" in a vague sense, but:
- Why do customers buy it? Lowest price on the market, or genuinely differentiated product? If the answer is price, the SKU is fragile because anyone can undercut. If it is a product feature or quality edge, that is something more durable.
- Which channel drives sales? Platform ads, creator content, or organic? High organic share means the product has pull on its own. High ad share means you are buying revenue.
- Why do customers return it? If the return rate is above five percent, that is a quality signal to fix before scaling.
- What is the real cash cycle? From the day cash leaves your account to the day revenue lands back, how many days does that actually take? That number determines how much working capital each SKU needs.
Understanding your winner deeply helps you choose expansion SKUs intelligently, because you know what you are good at and what logic the business runs on.
Step two: go deep before going wide
This is the biggest difference between shops that scale well and shops that sink into dead stock.
Going deep means adding variants or closely related versions of the winning SKU, from the same supplier or within the same narrow category. If you are selling a phone stand that is working well, add the car-mount version, add another color, add a size. Operations barely change because you already know the supplier, you already have the QC process, and the selling channel is already live. Risk is low, learning cost is low.
Going wide means jumping to a different category entirely: from phone accessories to kitchen goods, from sports equipment to stationery. Each new category comes with a new supplier set, different lead times, different packaging requirements, and its own demand characteristics. The learning cost is real, and the first import in a new category is almost always less efficient than the third or fourth.
The smart order: mine the depth until the winning SKU and close variants account for around seventy percent of stable revenue. Use the remaining thirty percent to test width, in controlled small batches.
Step three: the cash rule when adding new SKUs
Every new SKU means cash sitting in a warehouse before there is any revenue to show for it. Lead time from 1688 to Vietnam runs roughly eighteen to thirty days depending on route and product type, plus whatever time it takes to sell through the first batch. For a small test order, the full cycle can take four to six weeks. Add three SKUs at once and you multiply that by three.
A practical rule: only add a new SKU when you have spare capital after fully funding the SKUs already running. "Spare capital" here is not the balance in your account. It is what is left after subtracting upcoming import orders, this month's operating costs, and a buffer for surprises (typically at least fifteen to twenty percent of monthly operating cost).
If that number covers one new SKU and keeps the buffer intact, you add one SKU. Not two. The reason is that a new SKU needs attention in the early phase: tuning the listing, watching the sell-through, handling the first quality feedback. If you open two or three at once, no SKU gets enough attention and all of them perform at average.
Step four: do not let old SKUs eat young SKUs' capital
Once you have five or seven SKUs, the question is no longer "what to add" but "what to cut."
A SKU occupying warehouse space and capital without contributing proportional margin is draining resources from healthier parts of the portfolio. Signs to review:
- Sell-through rate falling for three consecutive weeks with no clear reason (not seasonal, not a platform glitch).
- Net margin after fees and ads dropping below eight percent with no sign of recovery.
- Stock sitting for over forty-five days from the most recent import batch.
When that happens, pick one of two paths: run down the inventory through controlled discounting or a bundle, then stop reordering, or find the real root (cost too high, supplier quality slipping, category saturated) and fix it meaningfully in the next cycle. There is no third option of holding on and hoping.
The capital you recover from a weak SKU is the right source of funds for the next genuine addition.
Step five: track the catalog as a catalog, not as isolated SKUs
Once you have multiple SKUs, the way you monitor has to change. Looking at each SKU in isolation does not show you which ones are dragging the whole portfolio, or where capital is being tied up inefficiently.
A simple weekly table for the full catalog should include: revenue, real landed cost, gross margin, current inventory measured in days of supply, and capital locked in each SKU. No complex software needed at this stage. A spreadsheet with five columns already shows the full picture.
When you look at locked capital by SKU, it becomes obvious fast which ones are consuming capital out of proportion to what they return. A SKU holding thirty percent of your locked capital while contributing only ten percent of margin is a flag to address immediately.
A stage-by-stage framework
There is no single right expansion speed for every shop, but a practical framework by stage:
- Stage one, one to two SKUs: focus entirely on finding a genuine winner. Do not expand until at least one SKU has run consistently for eight weeks with stable margin.
- Stage two, three to five SKUs: add variants and adjacent SKUs within the category. One SKU at a time, not several. Watch the cash cycle carefully at each addition.
- Stage three, six SKUs and up: this is where you genuinely need process, because you cannot keep it all in your head. ABC classification (group A carries most revenue, group C is the cut list) and systematic reorder points start paying for themselves here.
Each stage asks a different question. Stage one asks "is this SKU actually winning." Stage two asks "do I have the cash and the attention to add this one." Stage three asks "is capital being allocated sensibly across the whole portfolio."
Bottom line
Scaling from one SKU to a full catalog is not a race to add as many products as possible. It is a capital and attention allocation problem with a specific order: understand the winner deeply, go deep before going wide, control cash flow tightly at each addition, and cut weak SKUs before they drain the ones that are working. Shops that do this right tend to carry smaller catalogs that are significantly healthier than shops that keep adding without stopping to cut.