What makes the difference between successful shops and struggling shops when looking at the numbers
What I did
I conducted a comparative analysis of two shops on minne that handle transparent and clear-material accessories by obtaining publicly available data from the platform.
There are three perspectives I analyzed.
1. Performance comparison per listed item
I calculated the "sales efficiency" per item based on the total number of listings and total number of reviews.
2. Match rate of title keywords with actual demand
I tallied whether the keywords that buyers actually search for were used in the product titles for all items.
3. Comparison of category composition and navigation design
I compared the product category distribution within the shops and the mechanisms for encouraging repeat purchases.
The shops featured use only publicly available information and have been anonymized so they cannot be identified.
Finding 1: A shop with twice as many listings has 1/11th the reviews—a 23-fold difference in "efficiency per item"
First, let's line up the data for the two shops.
▼ Shop A (Successful)
Number of listings: 17, Total reviews: 1,093, Per item: 64.3, Price range: ¥500–¥2,400
▼ Shop B (Struggling)
Number of listings: 34, Total reviews: 96, Per item: 2.8, Price range: ¥1,500–¥5,500
Even though Shop B has twice as many products listed as Shop A, its total number of reviews is less than 1/11th. When calculated per item, the difference is 23-fold.
The idea that "if you increase listings, you will sell more" is not only failing to hold true, but the opposite phenomenon is occurring.
Why does this happen?
Looking at Shop A's lineup, it specializes almost entirely in earrings, with 15 pairs of earrings and 2 hair clips. Furthermore, 5 out of the 17 items are "resale" products. "Resale," which involves listing a product again after it has sold out, indicates that demand has already been confirmed. The decision to repeatedly create and list popular products like "Sea Fragments" and "Flower Waltz" is directly reflected in the performance per item.
In contrast, Shop B has a composition of about 13 brooches, about 18 pairs of earrings/pierced earrings, and 1 necklace out of 34 items. While the range of production is wide, there is no evidence of narrowing down to 'what sells best'.
If you re-evaluate your shop from the perspective of 'how much is sold per item' rather than the number of listings, your focus on where to put your efforts will change. If Shop B were to increase its performance per item to half that of Shop A (30 reviews/item), it would be calculated to reach 1,020 reviews even with 34 listings. The problem is not the number of listings.
Discovery 2: Even though they are both 'translucent accessories', the way they reach search results is 6 times different
Looking at the products of both shops, it is clear that they are made with a common taste of 'translucency, clear, and glass-like'. Despite the similar materials and production styles, there is a large difference in the number of reviews.
One of the reasons for this difference that emerges is the use of title keywords.
On minne, product titles function directly as search keywords. When a buyer searches for 'resin earrings' or 'allergy-friendly earrings', products that do not contain those keywords in the title will not appear as candidates.
Here are the results of aggregating the titles of all items from the two shops.
'Resin' inclusion rate: Shop A 53% (9/17 items), Shop B 9% (3/34 items)
'Allergy' related: Shop A 59% (10/17 items), Shop B 3% (1/34 items)
'Resale' indication present: Shop A 29% (5/17 items), Shop B 0 items
Shop B's product titles feature expressions like 'translucency', 'Nordic', 'moist', 'glossy', and 'delicate'. The intention to convey the world view of the works is clear. However, almost no buyers search for 'translucency'.
What is searched for are practical demand keywords such as 'resin earrings', 'allergy-friendly earrings', 'surgical stainless steel earrings', and 'resin pierced earrings'.
In contrast, Shop A always includes practical demand keywords in the latter half of their world-view titles, such as 'Sea Fragment Earrings/Pierced Earrings <Allergy-friendly metal parts selectable, blue, cube, resin>' or 'Flower Waltz Earrings/Pierced Earrings <Allergy-friendly, metal parts selectable, small and simple, resin, fan activity>'.
They are balancing world view and search keywords.
Out of Shop B's 34 listings, only 3 (9%) include 'resin' in the title. The remaining 31 are unlikely to appear as candidates in practical searches like 'resin earrings' or 'resin brooches'. If Shop B's works are made with clear materials and resin-based methods, just adding one word like 'resin' or 'clear resin' to the title would widen the gateway for search traffic.
'Translucency' is a word that conveys the charm of a product. 'Resin earrings' is a word entered by people looking for a product. These two are different things, and both can be included in the title.
Discovery 3: Category concentration vs. dispersion—whether it is conveyed in 1 second 'what kind of shop it is'
When you open the top page of a shop, whether you can instantly tell 'what this shop specializes in' affects subsequent browsing.
Of Shop A's 17 listings, 15 are earring/pierced earring types. If you look at the top page, you can immediately tell it is an 'earring specialty shop'. Furthermore, Shop A has developed multiple variations of the 'Sea Fragment' series (regular version, dangling version, Hawaiian blue version), and it is designed so that if you like one, it is easy to look at other colors and shapes. Browsing is generated by the series.
Shop B has a composition of 13 brooches and 18 earrings/pierced earrings. When a person who visits the shop for a brooch opens the top page, more than half are earrings/pierced earrings. A wavering perception is created: 'Is this a brooch shop or an earring shop?'
Brooches and earrings have different purchasing motivations. Brooches are chosen as accents for outfits or as gifts. Earrings are often bought in multiples for daily use. When these are mixed in the same shop, the sense of it being a 'specialty store' fades for both types of buyers, weakening the motivation to 'come back again'.
I don't know if Shop A consciously narrows its categories or if that happened as a result. However, the numbers show that 'narrowing down leads to higher performance per item'.
If Shop B wants to specialize in brooches, they have the option to lower the ratio of earring listings or move them to a separate shop. If they want to specialize in earrings, they could consider lowering the ratio of brooch listings. By shifting toward one or the other, it becomes easier for buyers to form the perception that 'if I come to this shop, I will find X'.
What AI can and cannot do
Analysis like this—retrieving publicly available platform data, aggregating keyword appearance rates, and calculating performance per item—can be executed in a short time using AI. Calculating the keyword match rate for 34 titles across all items would take a human 1 to 2 hours if done manually. AI can organize it in a few minutes. It can show 'where the weaknesses are' using numbers rather than intuition.
However, what AI is looking at is ultimately just 'numbers and strings'.
The quality of photos, the care taken in packaging, and the actual texture of the work do not appear in the data. Even whether the items in Shop B, which I compared this time, are actually made of resin cannot be fully confirmed from public information. I merely inferred it from expressions like 'transparency/clear'.
What numerical analysis shows is the priority of 'where to start for the most effective results'. Adding title keywords can be done today. Narrowing down categories can be done this week. Visualizing performance per item can be done right now. These are issues separate from the quality of the work and are improvements that any creator can equally undertake.
The state of 'not knowing why it isn't selling' is what most stops action. By identifying the cause with numbers, you can see 'what to do next'.
Why not look at your shop from the outside?
I am currently accepting free diagnoses for Creema and minne shops.
Analysis content:
Visualization of strengths and weaknesses in product lineup
Comparison with 3 to 5 competitor shops
Gaps in price range, keywords, and series composition
A three-stage action plan: 'improvements to make this week,' 'initiatives for this month,' and 'a 3-month direction'
Cost: Free (as it is currently in test operation)
If you are interested, please send your shop URL via the form below. I will deliver a report to you by email within a few days.
→ [Apply for a free diagnosis](https://tally.so/r/ja5byR)
As this is currently handled manually, it may take some time.
This article is intended for general information purposes only and does not provide any individual guarantees.