Should I lower my price if a product is not selling?

Usually not first. In a 232-listing sample, six of the same octopus phone holder sat inside a ¥1.17 price band and sold between 63 and five million units, so price was not the variable. Check price only after you know buyers can find the listing at all. Cutting first gives away margin to fix a problem that may not exist.

Why does the same product sell 87 units in one listing and five million in another?

Because the deciding variables are not on the page. Price, silhouette and title keywords were near identical across eight listings of the same holder. What actually differed, listing age, impressions received, and whether anyone pays for traffic, is not exposed to buyers on the platform.

How do I know if my price is actually too high?

Compare it with the percentile band for your own product scene, not with your costs and not with one competitor. Above the 75th percentile you are genuinely priced high and a cut is justified. Between the 25th and 75th, price is not what is blocking the sale, and a cut only removes your margin.

What should I change first if price is not the problem?

The word that names the product. If buyers search for a phone holder and your title calls the item a decorative ornament, you are routed to gift shoppers and never enter the comparison at all. Change that one term, ship it alone, then check search impressions after seven days.

Should You Lower the Price If a Product Isn't Selling?

I ran a ¥13.40 octopus phone holder against a 20-listing benchmark and told the seller to cut the price by 26%. Four hours later, with 232 listings collected, the median price for that scene turned out to be ¥13.40 — exactly what the product already charged. Here is the correction, the eight listings that made it obvious, and where a buyer-side benchmark stops being able to answer the question at all.

Every week someone posts the same question: "views but no sales — is my price too high?"

I just went through 232 phone-holder listings on Pinduoduo, China's most price-competitive marketplace, to answer it with numbers instead of an opinion. Of those, 168 report a single-listing sales figure.

The short answer: price is usually not the thing stopping the sale. Eight listings of the same octopus-shaped holder sit within ¥1.17 of each other and sell between 63 and 5,000,000 units.

I got this wrong four hours earlier

This morning I judged one of those listings against a 20-product sample: a ¥13.40 octopus desk ornament. In that sample, cute desk ornaments sold between ¥3.90 and ¥11.90, median ¥9.90. So ¥13.40 sat above the band, and I told the seller to cut the price to ¥9.90.

Four hours later the sample was 232 listings, and the cute ornament scene alone had gone from 9 products to 40.

The new median price: ¥13.40.

The product sits exactly on the median. My instruction was to cut 26% off a price that was already typical. It was backwards, and it was confident, which is worse.

Twenty listings gave me a price band so narrow that I believed it — too few points, and they lined up neatly.

Eight listings of the same product

Price	Units sold
¥13.40	63
¥13.40	87 (the one I judged)
¥12.73	1,289
¥13.40	3,115
¥13.40	10,000
¥5.40	14,000
¥4.89	155,000
¥13.40	5,000,000

Six of these sit inside a ¥12.73–13.90 band — a spread of just over one yuan — and their sales run from 63 to five million. That is roughly a 79,000× gap.

The other two dropped to ¥5.40 and ¥4.89 and reached 14,000 and 155,000 units — an order of magnitude below the same-priced leader. Two samples prove nothing, but they suggest the volume a price cut buys may be smaller than assumed.

Why nobody is buying it

In the cute ornament scene, median sales are 8,733 units. The 10th percentile is 63.

At 87 units this listing sits near the bottom decile. That much the benchmark states precisely.

What it cannot state is why. Price is ruled out. Silhouette is identical. Title formula is nearly identical — the listing at 63 units also names the product as a phone holder. The variables that separate them — how long the listing has been live, how many impressions it gets, whether anyone is paying for traffic — are not visible from the buyer side.

That is a hard ceiling on what a buyer-side scrape can ever tell you. More samples do not fix it.

What to change instead

Leave the price alone. Change the noun in the title.

Buyers in this category search for a phone holder (手机支架), not an ornament (摆件). The listing I judged calls itself a "creative desktop ornament" — which routes it to gift shoppers instead of people whose phone keeps sliding off a desk.

Every listing in this cluster above 10,000 units names it as a phone holder. The one that says ornament has 87 units.

But this is necessary, not sufficient. The listing at 63 units also says phone holder. A title fix moves a product from unfindable to findable; it does not move 87 units to 10,000.

When you should lower the price

Compare your price against the percentile band for your own scene:

Above P75 (cute ornaments: above ¥18.38) — yes, cut. Go to the middle of the band in one move instead of testing downward.
Inside P25–P75 (¥8.82–18.38) — price is not your problem. Do not cut.
Below P25 (under ¥8.82) — never cut. Add volume, packaging, or a bundled item.

The octopus sits at ¥13.40, lower-middle of the band. Do not cut.

What to measure after you change it

Change one thing. Wait seven days. Then look at the search-impression count in your seller backend.

Impressions moved, sales did not → the problem is on the page.
Impressions did not move → the problem was never on the page.

Do not change the price and the title together. Change both and you will never know which one did anything.

Limits of this data

232 listings, one day, one category, one platform; 168 report single-listing sales. Only price and sales are available — no click-through, no conversion rate, no listing age. Every figure is the platform's own display value, independently unverified. None of this has been tested after the change. It is a benchmark comparison, not a result.