Is It Worth Selling on Amazon UK? What the Survival Data Says

Every article about whether Amazon is worth it is an opinion piece, an affiliate funnel, or an anecdote about one person's FBA journey. None of them answer the question with data. The site's own UK Online Seller Business Index does: it is built from Companies House records on incorporated UK internet-retail businesses and it shows, cohort by cohort, how many survive. The odds are sobering, and the sellers who beat them share one habit.

The short answer

Selling on Amazon can be worth it, but the odds are humbling. Companies House data on incorporated UK online-retail businesses shows barely a third of the 2021 lockdown-boom cohort survived three years, and fewer than half of 2018 starters are still active. The survivors share one habit: they run it as a taxed business from day one.

What "worth it" really means: the survival odds

The question "is Amazon worth it" is really a question about odds. The UK Online Seller Business Index tracks formation-year cohorts of SIC 47910 incorporated internet-retail companies using Companies House data. These are not Amazon-specific figures and they do not include marketplace sole traders, who never appear on the Companies House register. They are the best available public proxy for what happens to online-retail businesses that start in the UK.

The table below shows each year's cohort: how many companies incorporated, how many remain on the active register today, and what share had survived without formal dissolution one, two, and three years after formation.

Source: UK Online Seller Business Index, derived from Companies House Advanced Search API (SIC 47910), data pulled 2026-07-15.

One reading trap in this table: the "active now" column looks better for recent cohorts (2024 shows 34.7%) but that is simply because those companies have not had time to fail. The fair comparison is the "survived X yr" columns, which compare cohorts at the same age. By that measure, the trend is clearly downward: the 2018 cohort's three-year survival rate was 47.1%; the 2021 lockdown-boom cohort's three-year rate was 31.6%; the 2022 cohort's was 23.4%. The index methodology notes this explicitly, and any interpretation of the data that compares raw active percentages across cohorts of different ages is misleading.

The lockdown year stands out on its own. 2021 brought 50,699 new incorporations in SIC 47910 as sellers rushed to capture the pandemic shift to online spending. By three years in, only 31.6% of that cohort had survived without formal dissolution. Over 39,000 incorporated businesses from that single boom year have already dissolved.

Why so many fail: the undercosting trap

The survival data points to a structural problem rather than a random one. The pattern is consistent across cohorts: first-year survival rates are reasonably high (70 to 87%), then the curve falls steeply in years two and three. That shape fits a specific failure mode: sellers launch, see early revenue, and then discover that their unit economics do not work once all costs are counted.

The costs that catch sellers out are not usually hidden. They are:

• Platform fees: referral fees, fulfilment fees, storage fees, advertising spend. These are visible on each transaction statement, but sellers often look at gross revenue rather than net margin per unit.
• Cost of goods: purchase price, import duties, freight, inspection costs. For FBA sellers sourcing from overseas, landed cost can be substantially higher than the factory price.
• VAT and income tax: sellers who cross the VAT threshold without planning lose 20% of gross sales to output VAT overnight. Income tax on profit sits on top of that. Neither is modelled at startup.

The problem is that these costs are rarely modelled together on the same sheet before launch. The seller take-home calculator combines all of them into a single figure: after fees, after cost of goods, after tax, what actually lands in your pocket per pound of gross sales. For most sellers who have not run the numbers this way, the result is a revision downward.

The hidden tax line most starters miss

Of all the costs that catch marketplace sellers out, the VAT threshold is the most common surprise, and the most expensive.

Under HMRC's VAT registration rules, registration becomes compulsory once taxable turnover exceeds £90,000 in any rolling 12-month period. For marketplace sellers, "turnover" is gross sales, not the platform payout. If you sell £100,000 of goods on Amazon but receive £70,000 after Amazon takes its fees, the VAT threshold test bites on the £100,000. Sellers who watch their bank balance routinely under-count and register late.

There is also a forward-look test: if you expect to exceed £90,000 in the next 30 days alone, registration is required immediately, regardless of where you are in the rolling 12 months.

A second threshold trigger that almost no guides mention: under HMRC Notice 741A, when a UK business buys services from overseas suppliers (Amazon's advertising services, overseas SaaS tools, platform fees billed from abroad), those are reverse-charge services. The UK buyer self-accounts for VAT, and the value of those reverse-charge services counts toward the £90,000 threshold. A seller running, say, £15,000 a year of overseas platform fees and ad spend has £15,000 of threshold exposure they may not be counting.

For the mechanics of how the threshold works and a real-time tracker against your own numbers, see the VAT threshold tracker and the VAT threshold: gross vs payout page, which covers the gross-sales rule in full.

What the survivors do differently

The cohort data does not tell us why individual companies survive, but the pattern across the curves is consistent with one explanation: sellers who survive treat the business as a managed, taxed entity from the start rather than discovering the economics retrospectively.

In practice that means three things:

1. Model true take-home before launching a product line, not after. Gross margin minus fees minus COGS minus tax is the real number. The seller take-home calculator runs this calculation before you commit to stock.
2. Track the VAT threshold actively from the first pound of sales. Crossing it unplanned creates a backdated VAT liability and a pricing problem: prices have been set on a VAT-exclusive basis and cannot easily absorb 20% output tax without margin impact.
3. Make the structure decision deliberately, not by default. Sole-trader status is simpler to start but limits what you can do with retained profit and means your personal allowance may already be used by other income before the first pound of selling profit is counted. The sole-trader-vs-ltd-sellers calculator runs the 2026/27 numbers for your specific extraction ratio.

The Amazon sellers hub covers the platform-specific tax and VAT questions in more depth, including the deemed-supplier rules for overseas sellers and the reverse-charge position on fees.

About the survival data: method and caveats

The cohort figures above come from the UK Online Seller Business Index, which uses the Companies House Advanced Search API to count SIC 47910 companies (retail sale via mail order houses or via internet) by formation year and current register status. The index is updated quarterly and published under the Open Government Licence.

Three caveats matter for how to read these numbers:

• Incorporated companies only. Marketplace sole traders, Vinted resellers, and side-hustle sellers who never incorporated do not appear on the Companies House register. This index measures the incorporated layer of UK online retail. It is the closest available public proxy for business survival odds in the sector, but it is not a count of all marketplace sellers.
• Not Amazon-specific. SIC 47910 covers internet-retail companies across all platforms and channels. Some companies in the index never sold on Amazon; some sold across multiple platforms. The index is a sector proxy, not an Amazon survival rate.
• Formal dissolution lags trading closure. The index records formal dissolution dates, which typically come weeks to months after a business has actually ceased trading. Real attrition runs earlier than the curves show.

For the full methodology, quarterly incorporation and dissolution flows, and the ONS internet-retail demand overlay, see the UK Online Seller Business Index.