Cuộc điều tra phân loại 100 store Shopify là dropshipping hay brand như thế nào?
Mỗi store được chạy qua engine classify-store của AliShopping Tools, trả về một trong năm enum — A_BRAND, B_RETAILER, C_DROPSHIPPER, D_POD, hoặc E_HYBRID — dựa trên bốn lớp tín hiệu: app stack (200+ app Shopify phát hiện qua DOM và asset URL), pattern catalog sản phẩm (độ rộng SKU và tính nhất quán ảnh), tín hiệu theme và copy (trang founder, chính sách hoàn hàng, ngôn ngữ vận chuyển), và tín hiệu thanh toán/tin cậy. 47 trong 100 store trả về C_DROPSHIPPER.
2
Bốn tín hiệu chính nào đánh dấu store Shopify là dropshipper?
Bốn tín hiệu chủ yếu trong 47 dropshipper được phân loại:
1dấu vân tay fulfillment app — 41/47 có DSers, AutoDS, CJDropshipping, Spocket, hoặc Zendrop được phát hiện;
2catalog rộng nhưng thiếu liên kết category — 28/47 trải ba category không liên quan như đồ thú cưng + dụng cụ làm đẹp + thiết bị bếp;
3ảnh sản phẩm kiểu AliExpress — 33/47 có ảnh nền trắng từ supplier AliExpress;
4ngôn ngữ vận chuyển 14–28 ngày, không có dịch vụ hoàn trả quốc tế.
3
Tôi có thể kiểm tra classify-store trên bất kỳ store Shopify nào miễn phí không?
Có.
Engine classify-store có sẵn trong Chrome extension AliShopping Tools (miễn phí, không cần tài khoản).
Cài từ Chrome Web Store, vào bất kỳ store Shopify — panel thông tin store của extension hiển thị phân loại (A_BRAND đến E_HYBRID), số app phát hiện được và category traffic.
Cuộc điều tra dùng đúng engine free-tier này.
4
Bao nhiêu phần trăm store trong quảng cáo sản phẩm TikTok là dropshipping bí mật?
47% trong mẫu 100 store trả về C_DROPSHIPPER.
Cộng thêm 18% E_HYBRID và 7% D_POD, tổng 65% store đang vận hành sourcing on-demand dưới một hình thức nào đó.
Chỉ 16% được phân loại là A_BRAND brand sản phẩm chính hãng.
5
Classifier phát hiện 6 dropshipper che giấu fulfillment app ra sao?
Sáu store chặn dấu vết app trên storefront vẫn được phân loại C_DROPSHIPPER dựa trên ba tín hiệu còn lại: catalog rộng trải nhiều category không liên quan, ảnh sản phẩm kiểu AliExpress (nền trắng, chữ tiếng Anh overlay độ phân giải thấp), và ngôn ngữ vận chuyển tiết lộ cửa sổ giao hàng 14–28 ngày không có logistics hoàn trả quốc tế.
47 Shopify Stores Secretly Dropshipping — How I Spotted Every One
Quick answer: Out of 100 Shopify stores sampled from TikTok product ads, 47 returned C_DROPSHIPPER from the AliShopping Tools classify-store engine — including 12 that presented themselves as premium brands. The classifier uses five store-type enums (A_BRAND, B_RETAILER, C_DROPSHIPPER, D_POD, E_HYBRID) based on app stack, catalog patterns, theme signals, and shipping language.
I scrolled TikTok for an hour in mid-April and screenshotted every product ad with a Shopify-hosted landing page. By the end I had 100 stores. None of them came with a label. Most looked premium-ish — clean photography, lifestyle copy, founder pages, a few testimonials. The kind of pages that read "small brand" if you do not know what to look for.
I ran every one of them through the AliShopping Tools classify-store engine. Forty-seven came back as dropshippers. Twelve of those were pretending to be premium brands hard enough that I would not have caught them with the naked eye. This post is the spreadsheet, the patterns, and the screenshots.
The sample is small and the stores are anonymised below — I am not in the business of dunking on individual operators. The purpose is to show what the classifier sees and what shoppers and competitor researchers can replicate in 30 seconds per store.
1. The setup — methodology and the classifier output
I picked the sample by scrolling the TikTok For You feed without filtering, screenshotting any product ad that linked to a Shopify storefront, and stopping when I hit 100. The sample skews toward whatever TikTok's algorithm fed me on April 14-15, 2026 — primarily home, beauty, fashion, and gadgets. It is not statistically representative of all Shopify stores; it is representative of what TikTok product ads were pushing in mid-April 2026.
Each store ran through the classify-store engine in AliShopping Tools, which returns one of five enums:
B_RETAILER — multi-brand retailer (boutique, curated). Stocks branded SKUs from third parties, has return / exchange policies that imply real warehousing.
C_DROPSHIPPER — fulfillment apps detected (DSers, AutoDS, CJDropshipping, Spocket, Zendrop), broad / unrelated SKU range, AliExpress-style product photography, 14-28 day shipping windows, no return logistics for international orders.
E_HYBRID — store mixes own-product with dropshipped. Has fulfillment apps but also evidence of owned inventory, often a brand pivoting into dropship for catalogue extension.
The classifier weighs four input layers: installed app stack (200+ Shopify apps detected via the storefront DOM and asset URLs), product catalogue patterns (SKU breadth, product image consistency, supplier signature), theme and copy signals (founder page authenticity, return policy specifics, shipping language), and payment and trust signals (currency switcher behaviour, payment processor variety, badge legitimacy).
The store info panel shows category, app stack count, and classification at the top of every Shopify storefront.
2. The 100-store breakdown
Here is the distribution from the 100-store sample.
Classification
Count
% of sample
C_DROPSHIPPER
47
47%
E_HYBRID
18
18%
A_BRAND
16
16%
B_RETAILER
12
12%
D_POD
7
7%
Total
100
100%
Forty-seven dropshippers, eighteen hybrids (also running fulfillment apps), seven print-on-demand. Sixty-five of the hundred stores are running on-demand sourcing of one form or another. Sixteen are owned-product brands. Twelve are genuine multi-brand retailers.
If you assumed a TikTok product ad was a real brand, you were right 16% of the time on this sample.
[Source: AStools classify-store engine snapshot, 2026-04-15. Sample is non-random — selected from TikTok For You feed product ads over a single browsing session.]
3. The 47 dropshippers — what flagged them
The classifier's job is to read the storefront the way an experienced researcher would, but in 2 seconds rather than 20 minutes. Across the 47 C_DROPSHIPPER classifications, four signals dominated.
Signal 1: The fulfillment app fingerprint
Forty-one of the 47 had at least one of DSers, AutoDS, CJDropshipping, Spocket, or Zendrop detected on the storefront. The detection runs against the page's loaded scripts, asset URLs, and Shopify object tags. None of these apps cloak themselves — they leave detectable signatures because they need to function inside the storefront.
The Apps panel surfaces every detected Shopify app on the storefront. DSers / AutoDS / CJ are hard signals for C_DROPSHIPPER classification.
The remaining six C_DROPSHIPPER stores had hidden their fulfillment apps (some operators block app traces from rendering on the public storefront). The classifier still caught them on the other three signals.
Signal 2: Catalogue breadth without category coherence
A real brand stays in its lane. A brand selling backpacks does not also sell kitchen knives. A real retailer stays in its lane too — a curated boutique sells a range of related lifestyle products, not a random scatter.
Twenty-eight of the 47 had catalogues spanning three or more unrelated categories: pet products + beauty tools + kitchen gadgets, or fitness gear + home décor + electronics. This is the signature of a dropshipper running multi-product testing rather than a brand running a focused range.
Signal 3: AliExpress-style product photography
Photos shot on a white background, often with the product slightly tilted, lifestyle prop in the foreground, English text overlaid on the image at low resolution. The classifier looks at image asset metadata, image dimensions, image-host CDN signatures, and visual consistency across the catalogue.
Thirty-three of the 47 had photo assets that traced back to AliExpress supplier listings (verified by reverse-search on a sample). Real brands almost always shoot custom photography. Dropshippers usually do not.
Signal 4: Shipping language and return policy
"Allow 14-28 days for delivery." "We ship from international warehouses." No domestic return address. No replacement guarantee. These are dropshipping tells. Real brands shipping out of a US, UK, or EU warehouse will quote 2-5 day delivery, will list a return address, and will offer a replacement policy with concrete language.
Thirty-six of the 47 had shipping language consistent with AliExpress sourcing. The rest had shipping language scrubbed but failed on the other three signals.
A store typically needs 2 of the 4 signals to score C_DROPSHIPPER with confidence. Most of the 47 had 3 or 4.
4. The 12 fake premium brands — the most interesting subset
The most interesting finding was not the 47 dropshippers. It was the subset within the 47 that had built a premium-brand façade around a dropship operation. Twelve stores out of the 100 had:
DSers + Loox; reviews show product arrived in AliExpress packaging
These are not illegal operations. Mark-up is allowed; positioning is allowed. The friction is between the brand promise (chemistry-led, designed-in-California, curated, made-for) and the supply chain (AliExpress generic SKU, 3-week ship from China). A buyer paying $140 for a "chemist-founded" skincare product is making a different purchase than a buyer paying $140 for a re-badged AliExpress generic. Most do not know which they are buying.
The classify-store engine does not pass moral judgement. It just outputs the enum. What you do with it is your call.
5. What this means — for shoppers, for competitors
There are two audiences who care about this output, and they care about it for different reasons.
If you are a shopper considering a purchase from a Shopify store you found on TikTok or Instagram: open AliShopping Tools, click the storefront, read the classification. C_DROPSHIPPER does not mean "do not buy" — there are perfectly good products being sold by dropshippers at fair markup. It means "calibrate expectations." Expect 2-4 week shipping. Expect international fulfillment. Expect limited return options. The product is probably available cheaper on AliExpress directly if you are willing to wait.
If you are a competitor researcher — a dropshipper sizing up which stores are running the same product, or a brand wanting to know whether a competitor is real or a re-badged supplier — the classification is the starting point of your research, not the conclusion. C_DROPSHIPPER stores share supplier risk, can be supplier-traced, and follow predictable scaling patterns. A_BRAND competitors are different beasts — owned IP, owned inventory, harder to undercut on price. The strategic response is different for each.
For the more granular question of detecting dropshipping inside a hybrid store, our detect dropshipping store guide goes deeper on the signal layering, and the Shopify store research guide is the pillar that ties classification, app detection, and live-sales velocity into one workflow. And for two well-known case studies, the Gymshark classification post and the Allbirds classification post walk through how the classifier reads a real branded operation versus a re-badged one.
6. Run this yourself in 30 seconds per store
The whole point of running the classifier is that it takes seconds. The 100-store analysis above took me about 90 minutes total — most of which was scrolling TikTok and screenshotting URLs. The actual classification step on each store was 2-4 seconds.
Workflow:
Open any Shopify store in Chrome with AliShopping Tools installed.
The store-info panel surfaces the classification badge automatically.
Open the Apps panel to confirm fulfillment-app fingerprint.
If you want to deepen the analysis, the Products panel shows catalogue breadth and the Live Sales panel shows live-order velocity.
That is the audit. No account, no signup, no per-store fee.
If you want to repeat my methodology with a fresh sample: open TikTok, screenshot 100 product ads with Shopify landing pages, classify each one, and post the breakdown. I would be surprised if the dropshipper percentage moved more than 5 points from sample to sample. The shape of the TikTok Shopify ad ecosystem in 2026 is what it is.
— Daniel
Disclosure: This article is published by AliShopping Tools. The 100-store sample was collected on 2026-04-15 from the TikTok For You feed, classified using the AStools classify-store engine on the same date. Stores are anonymised; the classification methodology is documented above. Email feedback via contact page.
All trademarks referenced are the property of their respective owners. This guide is for educational and competitor-research purposes; it is not a buyer recommendation.
