🛒 Selling on Amazon in 2026: how to build a profitable business

Amazon doesn't sell a significant share of its products on its own: more than 60% of sales in the Amazon store come from independent sellers, and this is confirmed by the company's own report. The marketplace is open to an ordinary entrepreneur, not just to major brands.

The problem is that beginners stumble at the start: they pick the wrong product, get confused about logistics models, and underestimate fees. Money goes into warehousing and advertising, but profit never arrives. Below is everything you need, no fluff: how to choose a product, which is more profitable between FBA and FBM, how much is left after fees, and what the latest reports show.

How to start selling on Amazon

💡 Quick overview:

• Step 1: open a Seller Central account and choose a niche with steady demand.
• Step 2: build your listing for the A9 algorithm: keywords in the title, clean photos, clear bullets.
• Step 3: compare FBA and FBM by margin, not just by convenience.
• Step 4: launch PPC ads and track ACOS from day one, otherwise the budget will evaporate.
• Step 5: check reviews, conversion, and inventory levels once a week.

Getting started requires one account and one product, not a perfect strategy. You open a Seller Central account, complete verification, and choose a niche, and only then do pricing, advertising, and logistics come into play. Before ordering your first batch, check the step-by-step business launch plan from the SBA so you don't miss registration, accounting, and legal compliance details.

Account verification also requires attention: prepare your documents, bank details, and tax information in advance so you don't get stuck in review when your product is already ready to ship. The cleaner your data at the start, the fewer pauses before your first ad launch.

Demand matters more than inspiration. A good product solves a clear problem, fits in a compact box, and is priced so that margin remains after fees. Warehouse and delivery mechanics are easy to check against the Fulfillment by Amazon guide, and the article on how Amazon works helps you understand the platform's overall model.

How to choose a product and build a listing that sells

Listings rank by relevance and sales, so the first word in working on one is: search queries. The A9 algorithm reads the title, bullets, and backend fields, while the buyer looks at photos and price. If the two layers don't align, the position drops even for a quality product.

A good listing stands on four pillars:

• A title with the main keyword and a clear benefit, without piling up synonyms.
• Clean photos from multiple angles: the main one on a white background, the rest showing usage scenarios.
• Bullets about the outcome for the buyer, not specs for the sake of specs.
• A price that leaves room for fees and advertising.

The principles of Amazon search optimization are covered in the official Amazon SEO guide, and you can find keywords with search volume using Ahrefs' free Amazon keyword tool. The main beginner mistake is the opposite: they invest in pretty packaging and forget about keywords, without which the product simply isn't found.

Niche validation is also data-driven, not intuitive. Look at the number of competitors in search results, price spread, and demand seasonality. A product with a narrow audience and a clear job to be done is almost always easier to make profitable than a trending gadget with a hundred identical competitors. Separately, calculate size and weight: a compact item is cheaper to store and ship, while fragile packaging eats into margin through returns. It's useful to compare several niches using the same criteria and enter the one where demand is stable and competition hasn't turned into a price war.

The business format is also chosen before the first purchase. Private label gives you control over the brand but requires investment in inventory and promotion. Reselling existing products starts faster, but margins are lower and price competition is higher. Online arbitrage and handmade occupy their own niches but scale less well, so a beginner is better off starting with one model and not spreading too thin.

Photos work before the text is read. The main image should be readable from a phone in a second, so show the product large, on a contrasting background, and without unnecessary details. Additional images address objections: size in hand, usage scenario, packaging, and contents. If a buyer leaves the listing without understanding the product, your ad spend is wasted.

FBA vs FBM: which logistics model is more profitable

The choice of logistics model directly affects margin and the Prime badge. With FBA, you send your product to an Amazon warehouse, and the platform handles storage, packing, and delivery. With FBM, you ship orders yourself. The Prime badge noticeably boosts conversion, so lightweight, fast-moving products are usually run through FBA.

Calculate not revenue, but margin after all fees: that's what determines whether a product survives.

A simple rule: a lightweight, fast-moving product is usually more profitable through FBA, while a bulky or high-priced one-off item is often cheaper to ship through FBM. A warehousing mistake costs more than an advertising mistake, because frozen inventory eats up working capital.

The second source of margin erosion: returns and storage fees. On Amazon, buyers return products easily, so fragile items, products that are complex to assemble, or ones with unclear sizing cost more than they seem to at the start. Before sending a batch to the warehouse, calculate not only the cost of goods and inbound shipping, but also the referral fee, return cost, and monthly storage fee. These expense items don't show up on day one, but at the end of the month, when the warehouse has already sent the bill.

Packaging is also part of logistics. A neat box, barcodes, and on-time shipping protect your account rating more than a weak product does. Late shipments hit your metrics faster than anything else, so check the FTC delivery deadline rule.

Advertising, analytics, and brand promotion

Amazon advertising works as an auction where you bid for impressions, so the first metric to watch is ACOS, the share of ad spend relative to sales. PPC campaigns bring traffic to a new listing while there are no organic sales yet, and then the goal is to convert that traffic into rankings. Ad formats and launch rules are covered in the 2026 Amazon ad types guide.

Analytics turns guesswork into decisions. The minimum set of metrics to review weekly:

• Listing conversion rate and CTR for key search terms.
• ACOS and the share of ad-driven sales in total revenue.
• Inventory turnover rate, so you avoid stockouts or overpaying for storage.
• Review volume and sentiment, especially negative reviews.

A separate growth lever: brand registration and your own content. The principles of building an ecommerce brand and working with an audience are collected in HubSpot's ecommerce marketing overview. A brand gives you protection from copycats and repeat purchases, which are cheaper than advertising.

Bid management is easier to control with negative keywords. They cut out irrelevant impressions that get clicks without purchases, keeping ACOS within your target range. Start with an automatic campaign, pull the search term report, move the working terms into a manual campaign, and add the junk terms to your negative list.

Reviews work as free advertising: a listing with a clear rating and fresh reviews converts better on the same traffic. Ask for reviews carefully, through Amazon's built-in tools, and respond to negative feedback not with an argument but with a solution. One resolved customer issue often brings more sales than an extra bid.

How much sellers actually earn: numbers and unit economics

Earnings on Amazon are measured not by revenue, but by margin after fees. The platform's scale sets the ceiling: in 2024, Amazon's net sales grew 11% to $638 billion, and in 2025 they added another 12% to reach $716.9 billion, according to Amazon's annual report.

The benchmark for an individual seller is also tangible: according to Amazon data, in 2024 independent sellers supported more than 2 million jobs in the US. And in an earlier independent sellers report, the company noted that US sellers on average sell more than $230,000 worth of goods per year.

Let's calculate unit economics based on ranges, not advertising promises:

Sustainable profit comes after the listing gains organic traction: first the listing collects reviews, then the share of paid traffic declines. So the winner is not the one who got in first, but the one who calculates unit economics and adapts the listing, price, and inventory faster.

A separate expense line that beginners often miss: working capital. Buying the second batch, paying for ads, and setting aside a reserve for returns require money that is tied up in inventory and does not come back instantly. So tie your growth plan not to revenue, but to free cash after restocking.

The benchmark of "sold a hundred thousand" without margin means nothing. Two sellers with the same revenue can have different profit: one pays for storage and returns, the other carries a lightweight product and steady demand. So track your numbers from net margin per unit, and use revenue only as the top line.

⁉️🤔 Frequently asked questions

How much money do you need to start on Amazon?

A realistic starting budget is measured in several thousand dollars: the first batch of product, photos, barcodes, and PPC ads for the ramp-up. You can start cheaper through FBM and small batches, but it's better to set aside a reserve for ads and unexpected expenses from the start, otherwise the listing will be left without traffic.

What is more profitable for a beginner: FBA or FBM?

For a lightweight, fast-moving product, FBA is usually more profitable: the Prime badge boosts conversion, and Amazon handles logistics. Bulky or one-off expensive items are often cheaper to run through FBM, where you control packaging and costs.

Is it realistic to become profitable in the first year?

Yes, but not for everyone right away. Profit depends on product selection, accurate fee calculation, and the ad budget at launch. The main reasons for failure: a weak product, underestimated fees, and no reserve for promotion.

How do you promote a product without a big budget?

First, build an accurate listing around key search terms and quality photos, then launch a careful PPC campaign with ACOS control. In parallel, collect reviews and register your brand: organic sales and repeat purchases are cheaper than paid traffic.

Do you need to buy a lot of inventory right away?

No. The first batch should be large enough to test demand, but not so large that it freezes working capital in the warehouse. Launch a small batch, collect data on conversion and sales velocity, and only then increase your purchase.

Summary and conclusions

Selling on Amazon remains a viable way to build a business in 2026, but the winners win on numbers, not luck. A specific recommendation: for a lightweight, fast-moving product, start with FBA and calculate margin after fees, not revenue; for bulky and expensive items, compare with FBM before paying for storage. Keep ACOS under control from day one, and don't launch the second batch until the first has confirmed demand through reviews and steady sales.

The main pitfall: frozen inventory and ads without calculation are the fastest way to push a project into the red. Apply the principles covered here to your next product: build a listing around key search terms, calculate unit economics in a spreadsheet, and launch your first campaign with ACOS control.