Best eCommerce Business Models for 2027: What Each One Earns After the New Fees and Rules

The best eCommerce business models for 2027 are the ones whose margin still holds after the costs already dated for next year, and which one fits depends on what you start with. An established brand with products people already buy is usually best served by multichannel selling: Amazon first, then Walmart, TikTok Shop or eBay, with a direct store where it has an audience. A founder with capital, a better product and a year of patience fits private label on Amazon. Somebody who needs revenue sooner fits brand-authorised reselling. A refurbisher with volume fits Amazon Renewed, a brand whose products are bought at work should add Amazon Business, and a brand without the team to run any of it should look at a revenue share partnership.

Below, the seven are compared on startup capital, margin, the 2027 changes, risk, time to first sale and fit. Fees are the platforms' own, read on 14 November 2026, or from our verified guides; anything only reported is labelled, and where a number depends on the product we say so. This explains how the models work; it is not financial or investment advice.

Amazon FBA fees, 15 January and 1 February. Amazon says fulfilment fees rise $0.19 a unit on average, "approximately 0.8%" of the average selling price. From 1 February two storage surcharges give way to one age-based rate, flat or lower for stock that sells within six months and higher for stock that sits. The detail is in Amazon's 2027 FBA fee changes.

Carrier rates, 21 December to 17 January. UPS raises daily rates an average net 5.9 percent from 21 December 2026 and FedEx its list rates 5.9 percent from 4 January 2027, both as reported by the trade press. USPS has filed a 5.9 percent average rise on its domestic package products for 17 January 2027, with an 11 percent fuel surcharge, pending regulatory review. What an online store pays from January works through the arithmetic.

De minimis, 1 July 2027. Duty-free entry for shipments worth $800 or less has been suspended since 29 August 2025, and on 1 July 2027 its repeal takes effect in law. Every parcel sent from abroad to a US customer needs a customs entry, and an electronic informal entry carries a $2.77 processing fee in fiscal 2027 before any duty. What the end of de minimis changes covers who it reaches.

Shopify script tags, 1 March 2027. Shopify stops running them in storefronts, so an app still relying on them stops working.

Private label is your brand on a product a manufacturer makes for you, with you as the only seller on the listing. The capital goes on the first order at the supplier's minimum, then freight, duty, photography, packaging, a trademark for Brand Registry, and the launch advertising that buys a new listing its first sales. The total depends on the product and its minimum order, so we give no figure; in our experience, first budgets built around the inventory miss by roughly half, as what private label actually takes explains.

The margin is the highest of the seven, because nobody sits between the factory and the customer. Amazon takes a referral fee by category: 15 percent in Home and Kitchen, in Toys and Games, and in Sports and Outdoors, 8 percent in Consumer Electronics, with a $0.30 minimum in most categories. The Professional plan is $39.99 a month. FBA adds a fulfilment fee by size and weight and monthly storage; FBM replaces both with your own postage, packing and labour.

In 2027 the pressure falls on small items at higher prices: small standard products priced $10 to $50 are reported to rise about $0.40 a unit, about twice the average. Storage costs the same or less if stock turns within six months, and the FBA New Selection offer is reported to rise to up to $600 per new product. FBM sellers take the carrier increases instead. A container from abroad was a formal entry already, so de minimis barely touches it; samples and top-ups sent as small parcels now each need an entry.

The first sale is months away, since sampling, manufacturing and freight each take weeks, and in the plans we build a single product usually breaks even between month six and month twelve. The risk sits in one account, one supplier and the stock. It suits a founder with a genuinely better product, capital to survive being wrong once and a twelve month horizon; the FBA business plan lists the numbers to run first.

A reseller buys an established brand's products at trade price and sells them on listings that already exist, so there is nothing to develop and demand is proven before the first order. The capital is the stock. The margin is thinner: the trade discount, less the same referral and fulfilment fees a private label seller pays. It reaches a first sale sooner than any model that makes its own product.

Authorisation matters more than fees. Amazon's guide to creating a product page says that for a brand enrolled in Brand Registry, "you need to be a Brand Representative or Reseller authorized by the brand to list it using our standard processes." Without a role, a reseller can still add an offer to an existing listing where Amazon's approval rules allow, but cannot create one. How brands authorise sellers in Brand Registry sets out both sides.

The 2027 fees bite harder here, because sellers sharing a listing compete on price, which makes a fee rise hard to pass on. Cheap items need a check: five categories are reported to lose the Low-Price FBA rates from 15 January. The risks are the Buy Box, a brand that changes its distribution, and slow stock. It suits an operator with supplier relationships and working capital who wants revenue sooner and accepts a thinner margin.

A direct store gives you the customer, the data and the whole price, with no referral fee. The platform is the cheap part: Shopify's plans are a monthly fee, lower when billed yearly, plus payment processing. The expensive part is traffic. With no marketplace search sending buyers, every visit is earned through search, social or paid advertising, and that cost decides whether the extra margin per order survives. Shopify against Amazon FBA compares the two on exactly that.

2027 reaches a store through shipping. If you charge a flat rate or ship free, the whole January carrier increase lands on you, and the order value at which free shipping pays for itself moves up. A store shipping from a warehouse abroad now pays an entry on every US parcel, and has to decide whether duty is collected at checkout or left to the buyer at the door.

A store can be live within days; the first sale waits on traffic. The risk is acquisition cost rising faster than the margin it buys. It suits a brand people already know, and every brand as a second channel once a marketplace has proven the product.

Multichannel is how an established catalogue grows and stops depending on one platform; it is not a model for a first product. The added capital is content, stock and people for each channel, and each marketplace charges differently.

Walmart Marketplace has "zero setup, monthly, or hidden fees", in its own words, and a referral fee by category on the total sale price: 15 percent in Home, Kitchen, Decor and Garden, Toys and Games, and Tools and Home Improvement.

eBay charges a final value fee of 13.6 percent of the total sale in most categories, up to $7,500 an item, plus $0.30 or $0.40 an order, on its published US fees in September 2026.

TikTok Shop's published US table lists a 6 percent referral fee for most categories. Seller publications reported a rise to 8 percent from 4 August 2026, which we could not confirm on TikTok's own pages, so read the rate on your settlement report. Creator commission comes on top, and creators are the channel's main way to sell.

In 2027, orders from other channels fulfilled out of FBA stock through Multi-Channel Fulfillment face standard-speed fees reported up about 3.5 percent on average, and self-fulfilled orders pay the carriers' new rates. The risk is operational: stock out of step between channels, and figures that disagree between reports, which multichannel analytics has to settle. It suits a brand already selling well on one marketplace, with supply to spare.

Amazon Renewed sells professionally refurbished products, a model for an existing refurbisher rather than a first business. Entry has a floor: itemised invoices showing at least $50,000 of qualifying refurbished purchases in the last 90 days, or 180 days in Home, Tools, Kitchen and Lawn and Garden. Amazon says approval decisions are typically made within 10 days. Who qualifies for Amazon Renewed covers the application.

The margin is tied to someone else's price: Renewed products must be listed at least 5 percent below the same product new, and every sale carries Amazon's 90 day Renewed Guarantee, whose claims the seller pays. 2027 helps. Amazon's Renewed page says a reduced referral fee applies automatically for one year from 16 October 2026, and the 2027 fee summary is reported to set it at 10 percent through 18 October 2027. Model the business at both rates, because a price cut made on the lower fee has to hold when it ends.

The risks are test buys, guarantee claims and that end date. It suits a refurbisher with volume and a grading process. A business that wants the channel without building the operation can hand enrolment, suppliers, compliance and listings to our Amazon Renewed service.

Amazon Business is the same marketplace sold to businesses and other organisations, and it needs no new account, only the Professional selling account a consumer seller already has. Amazon says Amazon Business drives $35 billion in annualised sales and serves 8 million organisations worldwide.

The margin structure changes in two ways. A business price or quantity discount sits beside the consumer price, and since 1 October 2025 Amazon has offered reduced referral fees on bulk business orders when that discount is 3 percent or more. And seller-fulfilled business orders are held to a 90 percent Business Hour Delivery Rate from 30 September 2026. The 2027 FBA and carrier changes apply as they do to consumer orders.

For a product already listed, the first business sale can come quickly; the work is pricing, documents and reliability, and the risk is low. It suits brands whose products are bought at work: office supplies, cleaning, tools, packaging and safety. What a consumer brand has to add for business buyers is the checklist.

The last model keeps the brand and the product with you and hands the running of the channels to a partner. You build no channel team, and the capital is the stock the partner will sell. A partner is paid by retainer, by a percentage of advertising spend, or by a share of revenue, usually the sales above an agreed baseline.

Under a revenue share, part of the growth goes to the partner, and the 2027 fee and carrier rises still land on your margin, so the agreement needs a margin floor written into it. How a marketplace growth partnership is paid explains why the baseline is the whole contract.

A brand already selling keeps selling from day one; growth from catalogue and advertising work builds over months, and in our view twelve months is a more honest term than three. The risks are a baseline set wrong, a partner chasing sales over profit, and exit terms nobody wrote down. It suits an established brand with clean numbers and no team to run several marketplaces, the case a retail growth partnership is built for: we open the channels and run the selling while the stock stays yours until it sells.

Under most of these models sits one question: does FBA still pay? For a small, light, fast-moving product with a healthy margin, the 2027 changes do not alter the answer: it usually still does. For heavy, slow or seasonal products, the honest test is a year of storage rather than one order, and the age-based rate makes stock that sits dearer than before.

So the answer is per product, not per account. Whether Amazon FBA is worth it is worked through route by route in our FBA guide; read it again with January's fees beside it. On Amazon, xRack, the analytics platform we built for the brands we manage, shows Profit and Loss beside Sales Performance, so a fee change appears on the products it touches. Walmart, TikTok Shop, eBay and Shopify are coming soon, and it is included with our account management rather than sold separately.

What can you afford to lose? Private label risks the most capital before a first sale; Amazon Business and a new channel for an existing catalogue risk the least.

When do you need revenue? Within a quarter rules out a new private label product.

Do you own a brand? Without one, reselling and Renewed are the routes. With one, private label, a direct store and multichannel open up.

Do you have an audience? If people already know the brand, a direct store earns traffic cheaply. If not, marketplace search is cheaper than buying it.

Who will run it? Every channel is a job. Without the team, a partnership costs a share of growth instead of salaries.

Then run each product's numbers at 2027's fees and carrier rates, not 2026's, before you commit.

When a brand asks us which model to run, we start from its products rather than the model: the margin each one leaves at January's fees, where its stock comes from after 1 July, and which channels its buyers use. The answer is often more than one model at once. Running Amazon, Walmart, TikTok Shop, eBay and Shopify as one operation is what our marketplace management does.