A Mulebuy spreadsheet is a curated database of products from Chinese e-commerce platforms (Taobao, 1688, Weidian) that have been cataloged with pricing, seller information, and quality assessments. This page explains how these spreadsheets are structured, what data they contain, and how to evaluate them critically.
The product data in a Mulebuy spreadsheet comes from multiple verification layers. Initial product discovery happens through platform searches on Taobao, 1688, and Weidian, filtered by minimum seller ratings and transaction history. Products are then cataloged with their original listing data.
For quality verification, a subset of products undergo actual test orders. This involves purchasing the item through an agent, documenting the QC (quality check) photos, measuring actual vs. advertised specifications, and noting any discrepancies. Not every product in the spreadsheet has been personally ordered—this would be financially impractical—but high-interest categories receive priority testing.
Price data is captured at the time of cataloging and rechecked periodically. However, Chinese platform prices fluctuate based on promotions, seller adjustments, and currency changes. Prices in any spreadsheet should be treated as approximate reference points, not guarantees.
Product availability, pricing, and seller status change frequently on Chinese platforms. A product listed as available in this spreadsheet may be out of stock, repriced, or removed by the time you view it. Seller ratings can also change. Always verify current status before ordering, especially for items over $50 or bulk purchases. This data represents a snapshot, not a live feed.
Understanding what each column represents helps you evaluate whether a spreadsheet meets your needs. Here's a breakdown of typical data fields and what they actually mean.
Usually translated from Chinese, often by machine translation. Names may not match Western product naming conventions. Generic descriptions like "High Quality Men's Watch" are common—focus on the images and specifications instead.
Prices are captured at listing time in Chinese Yuan (CNY) and converted to USD at that day's exchange rate. This does not include agent fees, shipping, or currency fluctuation since capture.
Indicates whether the product is from Taobao (consumer marketplace), 1688 (wholesale/factory), or Weidian (small seller platform). Each platform has different minimum order requirements and quality expectations.
Product categorization applied during cataloging. Categories may not perfectly match your mental model—one spreadsheet's "Fashion" might include items another would call "Accessories."
Direct link to the original listing or a product ID that can be searched on the source platform or through an agent. Links may break if sellers remove or modify their listings.
Product images sourced from the original listing. These are seller-provided photos, not independent verification photos. Expect some variance between listing images and actual products received.
Building a reliable product spreadsheet for Chinese e-commerce platforms is not a simple scraping exercise. It requires iterative refinement based on real ordering experience, and honest acknowledgment of what can and cannot be verified at scale. Here's a transparent breakdown of the methodology, including the failures and compromises that shaped the final approach.
Products enter the spreadsheet through a multi-stage funnel. The initial pool comes from category searches on Taobao, 1688, and Weidian, filtered by minimum seller ratings (4.6+ on a 5-point scale) and minimum transaction counts (typically 100+ sales for the specific listing). This eliminates the vast majority of listings immediately—most products on Chinese platforms come from new or low-volume sellers.
From this filtered pool, products are manually reviewed for listing quality: clear product photos, complete specifications, reasonable pricing relative to category averages, and coherent descriptions. Products with obvious red flags—watermarked images from other sellers, prices significantly below manufacturing costs, or nonsensical machine-translated descriptions—are excluded at this stage.
The criteria deliberately favor established sellers over new ones, even though this means missing some legitimate budget options. The tradeoff is lower risk for users who rely on the spreadsheet for initial product discovery.
Products are removed from the spreadsheet under specific conditions: seller rating drops below threshold, multiple user reports of quality issues, listing becomes unavailable, or price increases beyond 50% of original listing. Removal is not instant—a single complaint doesn't trigger removal, but patterns do.
Products are also removed if test orders reveal systematic problems. For example, if a watch listing advertises "316L stainless steel" but test purchases consistently show corrosion within weeks, that product gets flagged and eventually removed regardless of seller rating.
The early versions of this spreadsheet methodology had significant problems. The first iteration relied too heavily on automated scraping without manual review, resulting in listings that looked legitimate but came from sellers who simply copied successful competitors' photos. Several test orders from that period arrived as completely different products than pictured.
Another early mistake was trusting platform ratings without context. A seller might have 4.9 stars on 10,000 reviews for shipping speed and communication, but their actual product quality reviews (buried in the detailed breakdown) might be significantly lower. The current methodology weights product-specific ratings more heavily than overall seller ratings.
Perhaps the most costly lesson was underestimating seasonal variance. Products that tested well during summer production runs showed quality drops during Chinese New Year periods when factories use temporary workers. This isn't something spreadsheet data can fully capture, but it's documented in the notes where relevant.
There's an inherent tension between spreadsheet comprehensiveness and accuracy. A spreadsheet with 50,000 products cannot have the same verification depth as one with 500 products. The current approach prioritizes breadth for discovery purposes while concentrating verification efforts on high-interest categories and popular items.
Price accuracy is another compromise. Checking prices for thousands of products weekly is not feasible. Instead, prices are snapshot values that drift over time. Users should treat spreadsheet prices as ballpark figures, not quotes. The actual price might be 10-20% different by the time you order.
The spreadsheet also cannot capture intangible factors like seller responsiveness or dispute resolution quality. A seller might have great products but terrible communication when problems arise. This kind of information only emerges through actual ordering experience, which can't be systematized into spreadsheet columns.
Affiliate relationships create obvious bias potential. Products from platforms or sellers with referral arrangements could theoretically receive preferential placement. To control for this, the spreadsheet uses standardized inclusion criteria that don't reference commercial relationships. A product either meets the quality and seller thresholds or it doesn't—affiliate status doesn't factor into inclusion.
However, it would be naive to claim complete objectivity. Personal product preferences inevitably influence which categories receive more attention during the curation process. Fashion products get more granular categorization than industrial supplies simply because there's more user interest and testing depth in those areas.
Several data points were tested for inclusion but ultimately dropped. Estimated shipping times proved too variable to be useful—the same seller might ship in 2 days or 2 weeks depending on stock and season. User review summaries were tried but Chinese reviews are often incentivized or fake, making aggregation misleading.
Real-time stock indicators were technically possible through API scraping but created maintenance overhead that wasn't justified by accuracy improvements. Sellers frequently show products as "in stock" that are actually made-to-order, making the data unreliable anyway. The current approach acknowledges stock uncertainty rather than providing false precision.
The value of a product spreadsheet isn't just the data it contains—it's the thinking behind what data is included, how it's verified, and what limitations are acknowledged. A spreadsheet built on automated scraping alone will contain garbage. A spreadsheet with no commercial relationships might be more "pure" but also less maintained. Understanding these tradeoffs helps you evaluate whether any given spreadsheet deserves your trust.
Don't browse the entire spreadsheet randomly. Use category filters to narrow to your interest area first. This prevents decision fatigue and helps you compare similar products within a category rather than across unrelated items.
Spreadsheet data may be days or weeks old. Before adding an item to your cart, click through to the original listing and verify: Is it still available? Has the price changed significantly? Is the seller still active and rated well?
If a product looks promising, search for the same item on other platforms or from other sellers. Price differences of 30-50% for identical products are common. The spreadsheet gives you a starting point, not the only option.
Product price is only part of the total. Add: agent service fee (typically 5-10% of product cost), domestic shipping to warehouse, warehouse storage if combining shipments, international shipping (often the largest cost), and potential import duties. A $10 product might cost $35-50 delivered.
When your items arrive at the agent's warehouse, always request detailed quality check photos before approving international shipment. This is your last chance to identify defects or wrong items. Returning items domestically within China is far cheaper than international returns.
A Mulebuy spreadsheet is one tool among several options for Chinese product sourcing. Understanding when it's useful—and when alternatives are better—helps you choose the right approach for your situation.
Best approach: Use spreadsheets for initial discovery and popular categories. Switch to manual searching for niche items or when you need guaranteed current availability.
Reality check: No spreadsheet is definitively "best." Different lists serve different communities and use cases. Try multiple sources rather than relying exclusively on any single one.
If you've read through this guide and want to explore the actual product data discussed above, the database is available for browsing. The listing includes products across multiple categories from Taobao, 1688, and Weidian, organized using the methodology described in previous sections.
View the tested product list — the database is searchable by category, price range, and platform source.
New products are added on a rolling basis, typically several times per week. Price verification happens less frequently—expect prices to be 2-4 weeks old on average. Major audits to remove dead links and problematic sellers happen monthly. The "Last Verified" date at the top indicates the most recent systematic review.
The spreadsheet includes a mix of unbranded products, generic alternatives, and items from Chinese brands. Products claiming to be from Western brands at significantly below retail prices are almost certainly not authentic. This spreadsheet does not specifically curate counterfeit goods, but platform reality is that brand claims should be viewed skeptically. If authenticity matters, buy from authorized retailers.
Dispute resolution depends on your agent and the seller's policies. Most agents can help you negotiate partial refunds or replacements for clear defects documented in QC photos. However, returning items to China is often more expensive than the item's value. For expensive purchases, always get detailed QC photos before approving international shipping.
Seller ratings are captured from the source platform at listing time. Chinese platform ratings tend to be inflated compared to Western standards—a 4.8 rating is good, 4.6-4.7 is acceptable, below 4.5 warrants significant caution. Ratings also capture overall seller performance, not necessarily the specific product you're buying. A seller might excel at some products and underperform on others.
Purchasing products from Chinese platforms through agents is legal in most countries. However, specific products may face import restrictions (certain electronics, food items, plants, etc.), and you're responsible for any applicable customs duties and taxes. Counterfeit branded goods face seizure risk at customs. Check your country's import regulations for specific categories.
If you order a product from the spreadsheet and encounter significant issues (wrong item shipped, seller fraud, major quality problems), you can report it through the contact information in the footer. Documented reports with photos help improve the database. However, normal variance (slightly different color, minor quality differences from photos) is expected with Chinese platform shopping.
Products are removed when: the original listing becomes unavailable, seller ratings drop below acceptable thresholds, multiple users report quality issues, or prices increase beyond reasonable bounds. Removal doesn't necessarily mean the product was "bad"—sellers often discontinue listings or close shops for various reasons.
Product requests are considered but not guaranteed. The spreadsheet focuses on products that meet specific quality and seller criteria. If you've found a product that you think deserves inclusion, you can submit it for review, but most requested products don't meet inclusion standards upon investigation.