Global e-commerce revenue is on track to hit $6.88 trillion in 2026, roughly a fifth of all retail spending on the planet, with forecasts pushing past $7.38 trillion in 2027. A number that large makes opening a store sound like a guaranteed win. It is not. Somewhere between 60% and 80% of new e-commerce businesses fail inside their first year, and the causes are boringly consistent: the wrong product, an underfunded marketing budget, and unit economics that never worked on paper in the first place.
The stores that survive share a pattern. They picked the model that fit their capital, they proved demand before buying inventory, and they knew their true cost per sale from day one. This guide walks through the five legitimate ways to sell products online in 2026, what each one really costs, how long profitability takes, and who each one actually suits. One change since this guide first ran deserves top billing, because it quietly rewrote the math for anyone importing cheap goods: the $800 duty-free import threshold is gone. More on that below.
The five ways to sell products online in 2026
Dropshipping: low capital, low margins, brutal competition
You sell products you never touch. A customer orders from your store, you forward the order to a supplier, and the supplier ships directly to the buyer.
Startup cost: $200 to $2,000.
The appeal is obvious: no inventory risk. The reality is thin margins (15% to 30%), heavy competition, and a customer experience that lives or dies on a supplier you do not control. Industry data from Failory points to slow shipping and quality inconsistency as a leading reason dropshipping stores collapse inside their first few months. In 2026 there is a second problem stacked on top, one that did not exist when many dropshipping courses were written: imported low-cost goods now carry duties they used to dodge (covered in its own section below).
Treat dropshipping as a validation tool, not a destination. It teaches product research, ad management, and conversion work without risking thousands on stock. Learn the mechanics, prove a niche, then graduate to a model you can defend.
Full playbook: Dropshipping Reality Check: Why Most Fail and How to Be the Exception
Amazon FBA: massive reach, massive competition, real capital required
Amazon handles storage, shipping, and customer service. You handle product selection, listing quality, and advertising.
Startup cost: $5,000 to $15,000 or more, with inventory as the big line item.
Amazon puts your product in front of more than 310 million active customer accounts, but you are competing against roughly 1.65 million active third-party sellers plus Amazon itself. About 82% of active sellers use FBA, so its fulfillment network is effectively table stakes. The upside is real for those who find a niche with genuine demand and weak competition: more than 100,000 sellers now clear $1 million in annual sales, up from around 60,000 in 2021. The catch is the tax Amazon takes. Between referral fees, fulfillment fees, storage, and advertising, the average FBA seller pays 30% to 35% of revenue back to Amazon and nets a 15% to 20% margin after cost of goods.
Full playbook: Amazon FBA: The $5,000 Gamble, Real Costs, Margins, and Timeline
Print on demand: creative freedom, no inventory, a modest ceiling
Your designs get printed on products (shirts, mugs, phone cases, posters) only after a customer orders. No inventory, no upfront product cost.
Startup cost: $100 to $500.
The ceiling sits lower than the other models. Most print-on-demand sellers earn $500 to $3,000 a month. It remains a legitimate path for designers and creatives who want to monetize their art without touching logistics, and the global print-on-demand market, worth roughly $13.1 billion in 2026 and growing more than 23% a year, still has room. But AI image tools like Midjourney and DALL-E made design creation trivial for non-designers, which flooded the category and pushed differentiation onto niche and brand rather than the artwork itself. One 2026 wrinkle to plan around: Etsy now requires sellers to disclose generative AI in a listing and to relabel affected products from “Made by” to “Designed by,” so build your workflow around that rather than hoping to hide it.
Full playbook: Print on Demand Isn’t Dead, But Generic Designs Are
Etsy: a built-in audience, a heavy fee stack, niche dependent
Etsy’s roughly 87 million active buyers specifically hunt for handmade, vintage, and unusual products. If your goods fit that audience, you inherit demand without building your own traffic.
The catch is the fee stack. Etsy left its headline rates unchanged in 2026, but they still add up: a $0.20 listing fee per item, a 6.5% transaction fee on the full sale including shipping, payment processing at 3% plus $0.25, and an offsite advertising fee of 12% to 15% when a buyer arrives through Etsy’s ads. All in, fees commonly consume 15% to 20% of revenue, more on low-ticket items. With around 5.7 million active sellers chasing those buyers, treat Etsy as a channel, not a business, and move demand to your own store once you have proven it. Etsy’s January 2026 AI-disclosure enforcement, which removed roughly 12,000 listings and issued about 8,500 warnings in the first quarter alone, is a reminder of how quickly a platform can change the rules under you.
Full playbook: Etsy’s Hidden Fee Trap, And the Profit Strategy That Works Anyway
Shopify and your own store: full control, full responsibility
Your brand, your store, your customer relationships, and your marketing problem to solve.
Shopify hosts around 6.8 million active stores, and in 2026 its merchants crossed a milestone: they processed more than $100 billion in GMV in a single quarter for the first time, then posted $115.6 billion in the second quarter, up 32% year over year. Plans run $39 to $399 a month, but the software is not the hard part. Driving traffic and converting it is. Shopify sellers who win invest heavily in paid ads (Meta, Google, TikTok), email, and content. Once you fold in ad spend and inventory, the real startup cost lands between $3,000 and $10,000 or more. The payoff is ownership: the highest margins of any model here and a customer list no platform can take away.
Full playbook: Your Shopify Store Will Fail Without This: The First-Sale Strategy
What each model actually costs
| Factor | Dropshipping | Amazon FBA | Print on Demand | Etsy | Shopify (Own Store) |
|---|---|---|---|---|---|
| Startup cost | $200 to $2K | $5K to $15K | $100 to $500 | $100 to $1K | $3K to $10K+ |
| Monthly fees | $29 to $79 | $39.99 | $0 to $29 | $0 to $15 | $39 to $399 |
| Per-sale fees | Payment processing only | 15% referral + FBA fees | Base cost per item | 6.5% + processing | Payment processing only |
| Typical margins | 15 to 30% | 15 to 20% net | 20 to 40% | 15 to 20% net of fees | 40 to 70% |
| Time to first sale | 2 to 4 weeks | 4 to 8 weeks | 1 to 3 weeks | 1 to 4 weeks | 4 to 12 weeks |
| Income ceiling | $2K to $10K/mo | $5K to $100K+/mo | $500 to $5K/mo | $1K to $15K/mo | $5K to $100K+/mo |
| Inventory risk | None | High | None | Low to Medium | Medium to High |
The table shows the tradeoff nobody escapes: the models with the highest income ceilings also carry the highest startup costs and the most inventory risk. There is no version of this where you skip the tradeoff.
The tariff change that rewrote the low-cost import model
For years, the entire cheap-goods playbook rested on one quiet rule: the $800 de minimis threshold. Any shipment valued under $800 entered the United States duty-free, with no formal customs processing. That single exemption is what made $8 gadgets shipped one at a time from overseas suppliers pencil out. It is gone. The exemption was suspended in 2025 and reaffirmed as eliminated in a February 2026 executive order, so now every commercial import requires a full customs entry and duty payment regardless of value.
The effect on thin-margin models is immediate. Broad import duties now run 15% and higher across high-volume categories like electronics and apparel, with some China-origin goods facing far steeper rates. A $15 product that used to clear the border for free can now carry several dollars in duties and processing fees before it reaches your customer. On a dropshipping margin of 15% to 30%, that is often the entire profit.
Running operations at a large telecom for two decades taught me one thing that applies directly here: a cost you do not control is a cost that will eventually move against you, and usually at the worst possible time. Import duty is now exactly that kind of cost for anyone sourcing overseas. The sellers adapting fastest are doing two things. First, they are pricing duties into unit economics before spending a dollar on ads, not discovering them at checkout. Second, they are testing domestic suppliers, where the goods are already inside the country, which trades a slightly higher product cost for a border cost that no longer swings underneath them. If your model depends on cheap imported goods, run the new landed-cost math first. It changes which models are even viable.
Match the model to your capital
$200 to $1,000, testing the water: Start with dropshipping or print on demand for the low risk and fast feedback, and use it to validate a niche before you commit money to inventory. Just price in the new import duties before you decide anything is profitable.
$1,000 to $5,000 with a creative skill: Start on Etsy paired with print on demand or handmade goods. Prove demand, then expand to your own store. For products that pair well with physical goods, see 27 digital product ideas that actually sell in 2026.
$5,000 to $15,000, building a product brand: Amazon FBA for access to existing demand, or Shopify with private-label products for brand ownership. Both need inventory investment and both offer the highest ceilings.
$15,000 and up, serious about e-commerce: Shopify with private label, paid advertising, and a real brand build. This is the road to $10K to $50K or more per month in revenue, but it is capital intensive and demands marketing skill or the budget to hire it. Understand the real startup costs before committing.
Where AI actually helps, and where it just adds noise
AI has genuinely changed the economics of a one-person store. A few years ago these tasks needed a team or expensive contractors. Now a solo operator with the right stack handles them alone.
Product research. Tools like Helium 10 ($39/month) and Jungle Scout ($49/month) analyze demand, competitor gaps, and pricing in hours instead of the weeks manual spreadsheet work used to take.
Listing and copy. AI generates titles, bullet points, and descriptions at scale, and Amazon’s own listing tools suggest machine-learned improvements. This is real time saved, though the output still needs a human editor with taste, because generic AI copy reads generic to buyers.
Product photography. Tools like Flair.ai and Photoroom produce studio-quality product and lifestyle images without a camera, cutting creative costs sharply for direct-to-consumer brands.
Customer service. AI chatbots resolve most routine inquiries (order tracking, returns), so you delay hiring support until you are processing hundreds of orders a month.
Where AI adds noise: it will not fix a product nobody wants, and it will not manufacture demand. It compresses the work around a good product; it does not replace the judgment that picks one. A related shift worth watching is how AI shopping tools now surface products, which is changing where the buying decision happens. That is its own topic, covered in how AI shopping agents are sending customers back to your store.
Five mistakes that sink new stores
1. Skipping demand validation. Falling for a product idea and ordering $5,000 of stock before confirming anyone will buy. Always test with a small run, pre-orders, or a marketplace listing before committing capital.
2. Ignoring unit economics. Revenue per sale, minus cost of goods, shipping, platform fees, import duties, and ad cost per acquisition, equals profit. If that number is negative, more sales just means bigger losses. Calculate it before spending on ads, and in 2026 that calculation has to include the new import duties.
3. Underfunding marketing. Building a beautiful store and waiting for visitors. Organic traffic takes 6 to 12 months to build. Budget 30% to 50% of your starting capital for paid ads, or plan for a much longer runway on content and SEO.
4. Selling on one platform only. Platform dependency is a business risk. An algorithm change, a fee increase, or an account suspension can erase revenue overnight. Diversify to at least two channels once the first is stable.
5. Competing on price alone. Racing to the bottom attracts bargain hunters with zero loyalty. Compete on brand, experience, and quality. The stores clearing $10K a month almost always have a story and a base of repeat buyers.
What holds true regardless of platform
Product-market fit beats platform choice. A great product on the wrong platform still sells. A mediocre product on the “best” platform still fails.
Customer acquisition is the actual business. Building the store is easy. Getting people to visit and buy is the hard, expensive part. Mobile commerce alone reached $2.51 trillion in 2025, close to 59% of all e-commerce, so a store that converts poorly on a phone is leaving most of its buyers on the table.
Own the relationship you can keep. Marketplaces rent you their audience and can raise the rent (see Etsy’s fees and Amazon’s 30%-plus take). An email list is the one asset that travels with you across platforms.
Start with the model that matches your money. Do not stretch for a $15,000 private-label launch on a $3,000 budget. Begin with what you can afford, learn the fundamentals, and scale into higher-margin models as profits allow.
Start here
Pick the model that matches your capital and risk tolerance, read the detailed playbook linked in that section, and do the first action item in that guide today. Do not try to learn all five. Master one, prove it works, then diversify.
If you are starting with less than $1,000, the first $1,000 online roadmap lays out a week-by-week plan for building income with minimal capital.
Common questions
How much does it cost to start an online store in 2026? As little as $100 with print on demand or $200 with dropshipping, though in 2026 you also need to price in import duties on any overseas-sourced goods. Etsy runs about $100 to $1,000 in materials and listing fees. A Shopify store with paid advertising runs $3,000 to $10,000 or more including ad spend and inventory. Amazon FBA needs $5,000 to $15,000 for sourcing and initial stock.
Which model is most profitable? Private-label products on your own Shopify store carry the highest margins (40% to 70%) because you cut out marketplace fees, but they demand the most capital and inventory risk. Amazon FBA offers scale but hands 30% to 35% of revenue back to the platform. Print on demand is the lowest risk with a lower ceiling. The right answer depends on your capital and how much inventory risk you can stomach.
Are the 2026 import tariffs a dealbreaker for beginners? Not a dealbreaker, but a math change. The end of the $800 de minimis exemption means low-cost imported goods now carry duties they used to avoid, which erodes thin dropshipping margins the most. Domestic suppliers and print on demand sidestep most of that exposure. Calculate landed cost, duties included, before you commit.