Why Marketplaces Block AI Shopping Agents

Autonomous AI shopping agents are moving quickly from novelty to reality, with both financial and legal implications.

AI-first browsers such as Perplexity’s Comet and OpenAI’s Atlas can now search, compare, and initiate purchases with minimal human involvement.

That process, called agentic commerce, creates faster shopping for consumers and fewer clicks for merchants. It also challenges many ecommerce conventions, including the role marketplaces play in product discovery, transactions, and advertising.

Amazon and eBay have responded. Both are moving to restrict independent AI agents from completing purchases, citing security and user experience concerns. Yet in reality, the fight is almost certainly about control.

AI shopping agents threaten marketplaces such as eBay, Amazon, AliExpress, and many others.

Amazon vs. Perplexity

In November 2025, Amazon sued Perplexity, alleging that the Comet web browser masquerades as a human, accesses Amazon accounts, and places orders in violation of Amazon’s terms of service and computer fraud laws.

Third-party bots, according to Amazon, must operate openly and only with platform permission.

Perplexity countered that Comet acts on behalf of a human, with credentials stored locally for security, and suggested Amazon’s action was an attempt to protect its ad-driven business model and preserve control over shopping flows.

Essentially, Perplexity asks whether a platform can say no if a human authorizes an AI to shop.

eBay’s Ban

Just this month, eBay updated its user agreement to prohibit, without prior approval, “buy-for-me” agents and end-to-end LLM-driven checkout flows.

eBay positions the change as a safeguard against auction manipulation, fraud, and mistaken orders. The company, however, did leave room for “formally sanctioned” shopping agents, thus opening the door for partnerships that eBay can control.

Marketplace Concerns

Taken together, eBay’s update and Amazon’s lawsuit suggest that marketplaces seek to control agentic commerce relationships.

It makes sense. Marketplaces exist to aggregate and centralize shopping. It is the core service they provide and how they earn revenue. Hence agentic commerce is a threat.

Advertising. For the Amazon marketplace specifically and other marketplaces generally, advertising revenue is likely a chief concern.

According to its 2025 Q3 filing with the Securities and Exchange Commission, Amazon generated $47 billion in “advertising services” revenue in the first nine months of last year.

The company is much more than a product marketplace. It is a publisher, too, offering sponsored listings, recommendation units, and paid placements — all deeply embedded in search results and category pages.

Autonomous agents bypass the ads. Instead of scrolling through sponsored products and recommendations, the AI shopping agent skips to an item and initiates checkout.

First-party data. A related concern is shopper data.

Ecommerce marketplaces observe, track, and use shopper behavioral information. They monitor what shoppers search for, which products they view, and the items they abandon. Those signals feed ranking algorithms, recommendation systems, and personalization models.

That data disappears when an external AI agent performs comparisons and decision-making outside the marketplace, which sees only the final purchase.

Transactions. In its case against Perplexity, Amazon did not dispute that the AI agent completed the transaction via Amazon’s own checkout. Nonetheless, an AI-driven checkout creates at least two concerns.

First, the marketplace has no way to ensure that the transaction was proper. What if the AI agent made an error? What if the price is wrong? Could those errors lead to customer service problems or even increased return rates? Maybe.

Second, upselling becomes presumably impossible when the human shopper never sees it.

Compromise

Yet the developers of AI shopping agents disagree.

Agentic commerce startups argue that shoppers should be free to choose their preferred AI when they interact with services or websites. An AI agent, the argument goes, is more like a browser or an accessibility aid than a competitor.

Per the developers, marketplaces that allow only a few AI partners block human shoppers, stifle innovation, and foster monopolies.

The coming compromise will likely enable marketplaces to approve access within reasonable limits.

Thus AI agents, perhaps even Perplexity’s Comet, will eventually access marketplaces via official APIs, subject to rate limits, identity verification, and possibly commercial arrangements. Think affiliate programs for bots that pay for access.

For small-to-medium ecommerce businesses, the agent-marketplace relationship will likely be a primary route for getting products into Perplexity, ChatGPT, and similar platforms. It could be a key revenue channel.

Amazon Rules Product Discovery, for Now

The Amazon marketplace is the world’s most popular product search engine. Yet its dominance faces emerging challenges from AI and social commerce.

For more than 20 years, Amazon has made it easy for shoppers to discover products, compare options, read reviews, and buy.

A 2024 Jungle Scout survey (PDF) of 1,000 U.S. online shoppers found that 56% initiated product searches on the Amazon marketplace, compared to 42% on traditional search engines (such as Google), and 29% on Walmart.com.

Why Amazon?

Amazon’s Prime membership was a stroke of ecommerce genius. The service changes the way some consumers think about prices and shipping.

Products on Amazon’s marketplace are often more expensive than competitors’, and Prime costs $139 per year. But to many shoppers, there’s little reason to look elsewhere when shipping is free, fast, and reliable.

Selection

Moreover, Amazon’s product selection is massive and all-inclusive. Amazon itself sells more than 12 million products. Third-party sellers add upwards of 600 million, according to published reports. A shopper looking for an item will likely find it on Amazon.

Trust

Consumers trust Amazon. They assume products will arrive on time, with returns and refunds issued without hassle.

This trust is worth a lot. A 2025 Salsify report (PDF) found that 87% of shoppers have paid more for a product because they trust the brand. Those same consumers would likely search for products on a trusted marketplace.

Reviews

The volume of reviews on Amazon attracts shoppers.

Reviews serve as decision insurance. They reduce uncertainty and shorten the research cycle, especially for products where use cases matter. Instead of reading a handful of articles, comparing retailer sites, and searching Reddit threads, shoppers can pull social proof from thousands of real buyers without leaving Amazon.

That convenience changes behavior. The marketplace becomes a place for decision-making, not just to buy. So why not start a product search where other shoppers can guide you?

Mobile app

Amazon’s mobile app provides an advantage.

Searching for products in a mobile web browser is frustrating, even in 2026. Pages load slowly. Pop-ups appear. Cookie prompts get in the way. Shoppers must pinch and zoom, navigate cluttered menus, and jump between tabs.

Amazon’s app eliminates much of that friction for mobile consumers. The search box is always one tap away, filters are quick to apply, product pages are consistent, and the comparison process happens naturally through scrolling rather than clicking across multiple sites.

It’s a good experience, and shoppers use it.

Search iteration

“Search iteration” is the refinement of a query.

Consumers in the product discovery mode typically have specific needs. Amazon search can route shoppers toward products they are likely to buy.

Brand and mindshare

Amazon is ubiquitous beyond products. Prime Video, Audible, Kindle, Fire TV, Echo devices, and Amazon’s creator and influencer content indirectly contribute to search dominance and habit.

Boston Consulting Group, for example, asserts that such “mindshare” is highly correlated with purchase consideration.

Put another way, the folks who watch Prime Video are likely to search for products on Amazon.

AI and Social

Taken together, these factors serve as a playbook for the leading product search engine and offer both lessons and dilemmas for merchants. A shop can, for example, decide to include products on Amazon solely for discovery benefits.

Another consideration is whether Amazon maintains its lead in product search.

Some 56% of respondents on the Jungle Scout 2024 survey began product searches on Amazon. But that percentage is down from the 61% reported by Jungle Scout in 2022 (PDF).

Something is chipping away at product search and discovery. In 2026, that “something” is likely AI and social.

AI commerce is likely to shift where the first query occurs, thus eroding Amazon’s product-search dominance.

As shoppers ask for “the best” product option, generative AI platforms will increasingly assemble shortlists from multiple sources, reducing the need to start with Amazon. AI will pull discovery and comparison out of the marketplace interface, although Amazon can still win the transaction.

Social commerce on TikTok, Instagram, and YouTube will increasingly resemble search engines for lifestyle-driven categories. Shoppers, especially younger ones, often arrive at Amazon with a product already selected.

In those cases, Amazon becomes the fulfillment destination rather than the discovery engine, which changes the economics of product search and advertising on the platform.

Sell on Amazon without GTINs or UPCs

Amazon’s marketplace requires some form of a product identifier for nearly every category. Whether it’s baby toys or industrial bolts, sellers need a Universal Product Code (UPC) or, similarly, a Global Trade Item Number (GTIN).

Yet some items have no such identifier and require an “exemption” from Amazon.

GTIN

GS1, a global standards organization, developed the GTIN to identify products. The identifier encompasses several worldwide formats, including the familiar UPC (GTIN-12) in the United States and Canada. In Europe, merchants use the European Article Number (EAN), technically GTIN-13.

The GTIN standards identify any trade item (product or service) priced, ordered, or invoiced. Hence GTINs help Amazon and the merchants selling in its marketplace keep track of inventory and manage orders.

Identifying Products

For resellers offering products from prominent consumer brands, listing a product on Amazon is as simple as looking up the GTIN or copying the UPC number from the item’s barcode.

There are, however, many products that don’t have an obvious barcode or identifier. Here are some examples.

• Print-on-demand products. T-shirts are perhaps the best example. Services such as Printful and Gooten are excellent sources of quality shirts with crisp printing, but they don’t have UPCs.

• Private label items. Some small brands, including in-house manufacturers, may not have a GTIN.

• Handmade merchandise. Artisans creating anything from leather belts to home-sown bibs may not have UPCs.

• Parts or accessories. Replacement parts, components, and accessories may not come with individual GTINs. An example is a generic phone case.

• Bundled or items in a multipack. Most major brands have a GTIN-14 or similar for multipack, but some smaller companies with bundled products do not, despite Amazon’s guidelines.

For those and other numberless products, Amazon marketplace sellers have two options: obtain a GTIN or get an exemption from Amazon.

GTIN Exemption

Amazon often awards exemptions for categories with private-label, handmade, or similar products.

The process is straightforward. Create a listing in Amazon Seller Central, select “I don’t have a product ID,” and click “Next.” An “Apply” button will appear for products requiring a GTIN exemption.

If the product category is restricted, the seller will need double approval: an authorization to sell and a separate GTIN exemption.

Brands with existing GTINs cannot apply for an exemption for other like-branded items.

Exemption approval typically requires excellent documentation, as follows.

• Clear photographs. Amazon wants detailed product and packaging photos, not illustrations or markups. These pictures should show the item and packaging from each side, but will not appear in product listings.

• No GTIN. There cannot be (i) any barcode or identifier on the item or package or (ii) a GTIN associated with the product. Amazon has an exhaustive list of registered GTINs and will check.

• Consistent branding. The brand name on the GTIN exemption form must exactly match the packaging and product. And the packaging must permanently show the brand name or logo.

• Proper category. Ensure the product category for the GTIN exemption is correct. An error will almost guarantee a rejection.

According to its support, Amazon usually approves or rejects an exemption request within two days. Merchants who have obtained an exception often recommend waiting an additional day afterward before building those product listings, allowing Amazon’s systems to reflect the status.

Amazon’s notification of denied GTIN exemptions may state why. For example, the product and packaging photographs might not demonstrate the absence of a GTIN or may belong to a brand requiring a UPC.

Sellers with denied exemptions can reapply after updating or clarifying the initial culprits.

Finally, if all else fails, a company can purchase a UPC for about $30, plus an annual fee.

The Pricing Strategy of Temu Sellers

Temu sellers show massive discounts to boost perceived savings and win customers. The strategy is working.

Launched in September of 2022, Temu has grown rapidly. By December 2023, Temu had captured 17% of the U.S. dollar store market — Five Below, Dollar General, more — according to Reuters.

Certainly Temu’s prices are low. A recent comparison from Omnisend showed shoppers can often save 40% on Temu versus similar products on Amazon (but not Amazon Haul).

However, seller success on Temu relies as much on discount psychology as the savings.

Temu’s success relies in part on the psychology of discounts.

Cheap Headphones

Consider a hypothetical contract manufacturer that sells on Temu. Among the many products the company produces for various brands are noise-canceling, over-the-ear headphones that retail for $99.

This manufacturer-turned-Temu seller has the materials, models, and expertise to build a nearly identical set of headphones. Stopping short of using the original brand’s patented and expensive noise-canceling technology, the company tweaks the earpieces.

The “nearly identical” headphones cost $5 to manufacture. This company lists the headphones on Temu at a regular price of $99, with a 75% discount. That $24.75 price could produce more profit than it gets from building the client’s high-end version.

This made-up scenario is extreme but instructive.

Temu Prices

Remember that Temu is a marketplace with thousands of sellers. Temu does not buy and sell inventory. It connects manufacturers and brands to shoppers.

Thus the prices are not Temu’s but those of sellers. Yet Temu’s success relies on low-cost and, perhaps, low-quality alternatives to popular and brand-name products.

For example, the Omnisend comparison found that Temu listed an item that “matched” a product in Amazon’s camera and photo category about 96% of the time. But none — 0.0% — of those Temu items were the same brand or known quality as on Amazon. Instead, the items were unbranded and generic and “closely resemble well-known brands, often with blurred logos or modified packaging,” per Omnisend.

For identical items and brands, Temu’s prices are roughly the same as Amazon’s.

Temu Discounts

Sellers on Amazon and Temu differ significantly on the frequency and amount of discounts.

“While Amazon offers discounts on around 47% of its products, 65% of Temu’s listings are marked down. In addition, some items on Temu see discounts as high as 98%, compared to Amazon’s highest discount being 67%,” according to Omnisend.

Collectively, Temu’s deep discounts make products feel significantly cheaper. This is the psychology part.

For example, a February 2018 study in the academic journal “Fashion and Textiles” found that deep discounts can enhance shoppers’ perception of savings, making them feel like they are getting a better deal.

The study noted that discounts elicit an emotional response impacting shoppers’ perception of savings, quality, and value.

These findings are more or less common sense to seasoned marketers and precisely what is happening with Temu.

Product Perception

Yet deep and frequent discounting can be dangerous. Substantial markdowns may lead to doubts about product quality or authenticity, and rightly so.

Imagine a machine with three dials:

• Quality,
• Savings,
• Value.

Ecommerce sellers can manipulate the dials. Changing one has the opposite effect on another. The proper mix leads to an ecommerce sale.

Many Temu sellers have learned that turning up the perceived savings dial overcomes the associated decline in perceived quality. Temu customers presumably realize they are buying what could be a subpar item, but the perceived savings make it worthwhile.

To be sure, Temu’s success does not rely solely on psychology. “Many of Temu’s products ship directly from manufacturers to consumers so that product prices can remain low by bypassing distributors and other third-party handling fees,” explained Greg Zakowicz, senior ecommerce expert at Omnisend, in an email to Practical Ecommerce. “They also use the least expensive shipping method to reduce prices further.”

Nonetheless, the quality-savings-value exercise is a lesson for all ecommerce sellers. Every merchant has access to the purchase intent machine with its various dials, knobs, and switches. Good marketing is tuning the mix for success.

Born to Run with Amazon

Will enterprise brands boost inventory positions with Amazon in 2025? While a recent survey of Amazon sellers doesn’t necessarily answer that question, it does point to a now seven-year-old inventory program meant to help brands grow.

Some 41% of U.S.-based enterprise brands and retailers planned to use an Amazon-sponsored program — such as Vine or Born to Run— to help drive sales in 2025, according to the recently released “State of the Amazon Seller 2025” report from Jungle Scout.

In January 2025, Jungle Scout queried nearly 1,500 Amazon vendors, marketplace sellers, and folks just getting started with the platform. About 75% of  respondents came from the United States, and 47% worked at a “large brand or retailer.” The report reflects the sentiment of surveyed sellers but not necessarily the overall marketplace.

Nonetheless, more than four in 10 respondents from enterprise-level businesses ticked the “Born to Run/ Vine / Amazon Programs” box for planned growth channels in 2025. Vine is Amazon’s program that invites trusted reviewers to share their candid product opinions. The unspecified other “Amazon programs” tells us little, but “Born to Run” is interesting.

What Is Born to Run?

Amazon’s Born to Run program is an exclusive, invitation-only initiative to help Amazon vendors accelerate sales of new or existing products.

Started in 2018, Born to Run lets the vendor — a company selling directly to Amazon, not a marketplace seller — request purchase orders from Amazon and specify the anticipated unit sales in 10 weeks of a given product. If approved, Amazon purchases the requested quantity.

Here is a scenario. Imagine a brand called “Amazing Gizmos” that sells, well, gizmos. The Amazon purchasing team orders 400 units for the next 10 weeks. But Amazing Gizmos is about to launch a campaign on streaming television, and its marketing team expects to sell 1,000 units on Amazon.

With the Born to Run program, Amazing Gizmos can ask Amazon to increase its order to 1,000 units. Amazon agrees, and — assuming all goes well — Amazing Gizmos stays in stock and sells 978 units.

Hypothetically, Amazing Gizmos sold 578 extra gizmos (978 minus 400) because Amazon had ample inventory. What’s more, on its next regular order, Amazon’s purchasing team doubles down on the product without needing a second Born to Run request.

Similarly, a new vendor could use the Born to Run program to boost Amazon’s initial order. For example, Amazon says it will buy 100 of the new item, the vendor requests 300, and Amazon steps up. Just like that, the product launch could drive significantly more sales.

The key benefit in each case is that Amazon won’t run out on the vendor’s products.

Unsold Units

As long as the additional Born to Run items sell, an Amazon vendor should enjoy growth. If the extra inventory doesn’t sell, Amazon has two options.

• Return the items. Under the program terms, Amazon can return unsold units. The vendor refunds 100% of the product cost plus a 10% shipping and handling fee.

• Keep the items. Amazon can also retain the unsold units. The vendor will pay Amazon a “retention fee” equal to 25% of the cost of the unsold items — more or less a discount to Amazon to keep the items.

In either case, the “penalty” for being overzealous on project sales can be severe — 10% of the cost for returned units and 25% if kept.

Invitation Only

Sellers cannot apply for the Born to Run program. Amazon selects them based, at least in part, on a few requirements.

• Approved vendor. The seller must be enrolled in the Amazon Vendor Central program.
• Approved product. Only approved products with an Amazon Standard Identification Number (ASIN) are eligible.
• Product particulars. An item must sell for at least $5 and not be bulky, heavy, or classified as dangerous.

Finally, participating in an Amazon marketing program, such as advertising, could increase the likelihood of acceptance.

Survey

Based on Jungle Scout’s survey, sellers (including enterprises) are attracted to Born to Run and similar Amazon programs. The survey doesn’t reveal actual participation, but it’s a reminder that incentives like Born to Run exist and may work for some brands.

Temu’s U.S. Seller Program Is a DTC Opportunity

Direct-to-consumer brands eager to find customers have an opportunity with Temu, China’s rapidly growing discount marketplace.

In February 2024, Temu launched a U.S. Seller Program, effectively opening the platform to American businesses. The program gives U.S. brands access to an estimated 185 million domestic and international shoppers each month — and growing.

In 2023, Temu became Apple’s most downloaded free application and dominated the iOS and Android app stores in 2024.

The Temu U.S. home page focuses on discount items.

U.S. Seller Program

Temu, like its sister site, Pinduoduo, operates primarily on a consignment model.

Chinese and East Asian manufacturers fill Temu’s warehouse with goods and create product listings for the marketplace. When sold, the item is shipped directly from the Temu facility in a familiar, bright orange bag using clever air freight strategies to keep costs low.

The company recently changed tactics, allowing American merchants to list products and optionally employ Temu’s warehouse and fulfillment system.

The program is free for small sellers, but a business requires a subscription ranging from 2% to 5% of the selling price. At the time of writing, the marketplace also charged a payment processing fee of 2.9% + $0.30 per transaction, and sellers paid all shipping costs. Collectively, the fees make Temu similar to other marketplaces.

DTC Opportunity

Temu’s low prices may not fit traditional retailers, but DTC brands could have an opportunity.

A DTC product is unique. Similar products may exist in Temu, but none are identical. Plus, American-made products might have a competitive advantage owing to perceived value and quality.

All told, I see five potential benefits for DTC brands selling on Temu.

Brand building

DTC brands on Temu can introduce shoppers to the company and build relationships.

The introduction happens when a Temu buyer finds the brand’s products. The relationship starts with order fulfillment. DTC sellers could include in the packaging a physical product catalog, a coupon for a free item, or a note describing the brand’s story.

Items requiring a warranty registration offer the opportunity to collect the buyer’s email address and phone.

Revenue

Any established sales channel is a revenue opportunity. Temu has a massive user base, and those shoppers, discount-oriented as they may be, are the opportunity.

Temu’s media agency told me the company does not share estimated or average seller revenue. DTC shops should test, optimize, and iterate on the platform.

Marketing

DTC brands listing products on Temu can participate in platform-wide promotions and flash sales, driving traffic to listings and thus more interactions

Chinese expansion

DTC brands can flip the script and offer products to Chinese buyers via inventory stored in Hong Kong or other Temu locations.

Product development

The absence of a Temu listing fee facilitates the testing of new items. DTC brands can create short runs of prototype products, offer them on Temu, and learn what appeals to shoppers.

Marketplaces Generally

Selling on Temu should be part of a general marketplace strategy for DTC brands.

A brand selling on one marketplace can consider others. Hence listing on Amazon, Temu, Walmart, Esty, and eBay could all be part of an overall marketplace approach, such as:

• Set marketplace-specific objectives. Define revenue targets, customer acquisition rates, or brand awareness metrics for each marketplace.

• Establish marketplace audiences. Use analytics to learn customer demographics and purchasing behavior on a per-marketplace basis. A Temu shopper will likely differ from one on Amazon or Walmart.

• Align products with the marketplace. A brand might have several versions of similar items. Perhaps the top quality goods are on Amazon and entry-level items on Temu. Returns and seconds could sell on eBay.

• Optimize product listings. Common optimization tactics across all marketplaces include quality images, keyword-rich descriptions, and competitive prices. But keep in mind platform-specific practices, conventions, and rules.

How Amazon Haul Impacts Ecommerce SMBs

Last month Amazon launched Haul, a low-price marketplace competitor to China-based Temu and Shein. Both have made headway in the U.S., shipping extremely low-cost items directly from China.

Reuters reported in December 2023 that Temu had captured 17% of the U.S. market among discount retailers, which include Dollar Tree, Dollar General, and Five Below.

Temu and Shein are the first and second most downloaded shopping apps on the Apple App Store, per Yahoo Finance in April 2024.

Amazon Haul

Amazon needed more low-cost options to stave off this rapidly growing Chinese competition for, essentially, the dollar store market.

At the time of writing, Haul was available only in the Amazon app, and its listings were separate from the general Amazon marketplace.

Prices for Haul items are $20 or less, with many under $10 and some as low as $1. Shipping is free for orders of $25 or $3.99 otherwise.

Haul brings dollar-store items to Amazon via its app. Click image to enlarge.

Seller Impact

Informed Amazon sellers have warily anticipated Haul for months.

Amazon presumably believes Haul will attract shoppers, which seems to be the case. On Black Friday 2024, Amazon offered 50% off for everything on Haul to help kickstart interest. Marketplace Pulse reported many of the nearly 3,000 sellers participating in Haul sold out of key items during the promotion.

“If you’re a merchant selling low-ticket items, Haul is just another means to help price-conscious consumers find your products more easily,” wrote Katie Moro, global director of managed service at Productsup, a provider of ecommerce product content, in an email.

“Shoppers don’t have to set a filter to narrow their search to products in your price range. They can scroll through the Haul storefront on the Amazon app freely, knowing the products meet their budget requirements,” Moro continued, “Haul provides a huge benefit to your business in this way, as it helps compete with merchants on Temu or Shein.”

Thus Haul might be a marketplace booster.

It could also impact customer acquisition tactics, private label brands, and arbitrage sellers.

Customer acquisition

Some brands will likely use Haul to promote loss leaders and acquire customers. Loss leader strategies are similar to advertising. A brand buys an ad or sells a product at or below cost to get sales.

“It’s a similar strategy to the viral Lululemon belt bag. Compared to the usual price tag of its athletic wear, Lululemon made the bag significantly cheaper,” wrote Moro.

“By attracting first-time buyers to its brand and exposing them to the quality of its products, Lululemon could establish customer loyalty — what starts as a small belt bag purchase evolves into a pair of leggings, and so on. Similarly, you can leverage the high visibility of Haul to expand your audience reach with a few cheaper items and then continue to build the customer relationship on the regular Amazon storefront with the rest of your product catalog,” according to Moro.

Private label rival

Many private label sellers on Amazon source products from China. Nothing stops those Chinese manufacturers from selling directly on Haul — a common practice for items on Temu and Shein.

If Haul products started appearing in the regular Amazon marketplace, sales of private-label goods could suffer.

Arbitrage competitor

Ecommerce drop-shippers that rely on AliExpress-style arbitrage could also face competition from Haul.

These sellers typically use Dsers and similar apps to identify products on the AliExpress marketplace and resell them at a profit via Shopify and marketplaces.

This sort of retail arbitrage works because relatively few consumers know about AliExpress or how to access it. Amazon Haul makes direct-from-China retail mainstream; every AliExpress seller could end up on Haul.

Cheap Ecommerce

Only time will tell whether any of these scenarios play out, and many other factors could impact direct-from-China retailing —new U.S. tariffs, for example.

What we do know is the popularity of cheap ecommerce items, very cheap.

Amazon’s Direct-from-China Plan Criticized

We asked industry pros to comment on Amazon’s plan to create a new section for Chinese sellers to ship directly to U.S. customers. The move is an apparent effort to recoup consumers who have turned to Temu and Schein, producers of inexpensive household goods and apparel, respectively.

China-based Shein sells inexpensive “fast fashion” apparel.

Bad Idea

It’s a bad idea, according to Phil Masiello, CEO of CrunchGrowth Revenue Acceleration Agency and a longtime Amazon seller and founder of multiple ecommerce companies.

Sellers and brands have been fighting with Amazon against cheap fakes from China for years. “It’s going to anger the brands on there,” Masiello said in a video interview, adding, “Amazon should go higher. They should go into exclusives versus trying to be a Temu.”

The Chinese competition is selling “junk to the uneducated. They buy it once. They’re not long-term Temu customers,” Masiello said. Temu’s popular, he says, but the business model is not sustainable.

“Amazon has one thing that any brand would love, which is retention,” Masiello added. “Everybody has the Amazon app on their phone. It’s the first place we look for something.”

Masiello believes the move will cost Amazon, where quality sellers face increasing fees — about 50% of sales go to Amazon.

Masiello’s not alone in his opinion that Amazon’s making a mistake.

Inviting Competition

“Amazon made a deal with the devil by letting this crap in from overseas,” stated Rick Wilson, chief executive officer of Miva, an ecommerce platform. “They invited the competition.”

Higher-end products will be insulated from the new storefront, but “it ultimately depends on the item.”

“Amazon continues to aggressively pursue overseas manufacturers and make it easier for them to become consumer brands themselves,” James Thomson, managing partner at Equity Value Advisors and a former Amazon executive, said. “In many categories, U.S.-based brands on Amazon are sourcing stuff overseas and now competing against their manufacturers. Amazon’s enabling them.”

“As Amazon goes after lower cost options, it’s harder for small brands in the United States to do well,” Thomson said.

Still, Thomson doesn’t see much of a problem for better sellers at higher price points.

“Lots of stuff on Temu and Schein is a spontaneous purchase,” Thomson said. “I don’t go to these types of sites thinking here’s what I need to refill my supplies at home.”

Thrasio, the ecommerce aggregator that recently emerged from bankruptcy, is focusing on quality and loyalty to avoid competing with Temu and Schein products.

Commoditized items, like kitchen utensils, will become even cheaper for consumers with direct-from-China offerings.

By focusing on quality, Thrasio hopes to avoid competing against inexpensive alternatives.

“There’s just no way to compete on some of those commodity products,” Stephanie Fox, Thrasio’s new chief executive officer, said in an interview. “Your margins are going to be 5%, which will never support a scaled business. Could solo entrepreneurs potentially compete against those products? Sure, but they won’t make a ton of money doing it.”

“Competing in those low-margin commodity products, which is exactly what Amazon is focusing on, it’s not going to be worth it,” Fox said.

Mark Daoust, the founder of Quiet Light, an ecommerce brokerage, compared the move to the launch of Amazon Basics.

“I saw a business that was killed by Amazon Basics,” Daoust said, citing one client who sold lower-priced office chairs. “Most sellers want to build a brand, a quality product, and focus on a uniqueness that no one can imitate. The client made economical chairs that were accessible to a lot of people. That was the whole value proposition.”

It wasn’t necessarily the best chair on the market, but it was economical and worked very well — until it didn’t. “Amazon Basics destroyed that company.”

Prime Day 2024: What We Don’t Know

Here’s what we don’t know about Prime Day 2024. Amazon says the event was global but famously provides no revenue details. We’re left with estimates by Adobe and others of U.S. purchases.

Thus we don’t know the worldwide volume or the impact on Amazon’s income since most of the purchases are presumably with third-party marketplace sellers from which Amazon earns a commission — $1 billion of purchases at a 20% commission would be $200 million of earnings to Amazon.

Less clear, too, is the bottom-line effect of Prime Day discounts on third-party merchants who tell Practical Ecommerce that Amazon takes roughly 50% of every transaction with sales commissions, FBA fees, and advertising costs.

Certainly Prime Day is good for consumers and Amazon, which carries little inventory risk and earns fees and commissions no matter the selling price. Amazon is mostly a service provider, after all, and most of that is from its cloud computing division, not marketplace activity.

Nonetheless, here’s our recap of Prime Day 2024.

Prime Day 2024 Recap

Adobe Analytics is popular among large U.S. online retailers. It has long been a credible source for Prime Day data. For this year’s event, Adobe says it tracked 1 trillion U.S. web visits, 100 million SKUs, and 18 product categories. All told, per Adobe, U.S. consumers spent a record $14.2 billion during July 16 -17, up 11.8% from a year earlier. (Amazon stated only that the 2024 event was “the biggest ever.”)

Other Adobe findings:

• 49.2% of purchases were on mobile devices, up 18.6% from 2023.

• Sales of back-to-school goods (backpacks, lunchboxes, supplies) increased by 216% from last year, presumably owing to the event occurring a week later than in 2023.

• Electronics sales increased 61% due to “product refresh cycles” wherein consumers upgrade their devices.

• Housewares, furniture, and appliance sales were strong, with increases in kitchen appliances (up 76%) and cookware (up 26%) leading the category.

• Categories with the top discounts from list prices were electronics (23% off), apparel (20% off), home goods/furniture (16% off), television (16% off), and toys (15% off).

Other Metrics

Numerator is a U.S.-based data and analysis firm. For its Prime Day 2024 report, Numerator tracked 93,513 U.S. orders, 35,588 households, 188,000 items purchased, and 7,311 verified buyer surveys. The report included:

• Shoppers’ ages: Under 34 (14%), 35-44 (19%), 45-54 (18%), 55-64 ( 22%), 65+ (27%).

• Household spending: Under $100 (34%), $100-$200 (43%), $200+ (23%)

• The top-selling item was the Amazon TV Fire Stick, followed by Premier Protein Shakes, Liquid I.V. Packets, Glad Trash Bags, and COSRX Snail Mucin Serum.

Consumers want discounts, according to Pacvue, an ecommerce advertising platform.

“Shoppers are still feeling the effects of inflation and are hungry for deals,” Melissa Burdick, president of Pacvue, told Practical Ecommerce.

For once, Amazon didn’t dominate the site with its own products. “Interestingly, it’s one of the first years that we have not seen Amazon devices dominate the Prime Day home page, with Apple products owning the main landing page,” she said.

Advertisers allocated more money this year due to increased competition, according to Burdick. Compared to 2023, this year’s Prime Day saw a 30% higher total managed spend starting from the pre-event phase, with a 47.5% increase on the event’s first day. On the second day, average spend increased by 24% compared to the same day last year.

“In the lead-up period to Prime Day, brands increased their spend by 15% from last year,” Burdick said. “We’ve also seen brands reallocate ad budgets from August to support their Prime Day efforts.”