How to start selling online in the UK

You've made a thing, a few people have bought it, and now you want to do it properly. The question is what "properly" means and how much of it you have to do today.

Not much of it, is the answer.

You can start selling in the UK as a sole trader, with your own bank account, a page that shows your products and prices, and a way for people to pay. No company, no VAT registration, no accountant. Two rules apply from your first sale, though, and neither is optional: your buyer has a statutory right to change their mind on a distance sale, and once your rolling 12-month taxable turnover crosses the VAT threshold, which gov.uk gave as £90,000 when I checked it in August 2026, you must register. Everything else is a choice.

The rest of this is about the choices, and about the £18 candle that quietly loses money because of how it's packed.

Sole trader now, company if it earns it

You can trade as yourself. Most people selling under a few thousand pounds a month do, and there's no shame in it and no legal problem with it. You tell HMRC, you file a Self Assessment return once a year, you pay tax on your profit.

A limited company gets you limited liability, a name at Companies House, and a structure that some wholesale suppliers and most business customers prefer. It also gets you annual accounts, a confirmation statement, corporation tax, and a director's responsibilities you can't quietly ignore. It's a real commitment with real running costs.

The honest rule of thumb: stay a sole trader until either the tax maths clearly favours a company at your profit level, or you're taking on risk you don't want landing on your personal assets. Ask an accountant once, for an hour, rather than reading forum threads for a fortnight. And check current registration requirements and any small trading allowance figures on gov.uk directly, because those change and a stale number in an article could cost you money. Do you need to register your business walks through the decision without the jargon.

The 14-day rule you cannot write your way out of

This is the part UK sellers most often get wrong, and it's the part with actual legal teeth.

When you sell online to a consumer, that's a distance sale, and the buyer gets a cancellation right that has nothing to do with the item being faulty. They can change their mind. On gov.uk in August 2026, the rules read: you must offer a refund if the customer tells you within 14 days of receiving the item that they want to cancel, they then have another 14 days to send it back, and you must refund within 14 days of getting it.

You cannot put "no returns" in your terms and make that go away. Writing it just makes you look like someone who doesn't know the rules.

There are exceptions in the regulations, and personalised or made-to-order goods and certain perishables are the ones that matter to small sellers. Check the current exceptions on gov.uk rather than assuming your product qualifies, because "I made it in her chosen scent" and "it is personalised" are not always the same thing in law.

Plan your margin around returns you will actually receive, not around a returns policy you wish were enforceable.

What that means practically: build a return rate into your pricing, make your product photos honest enough that fewer people are surprised, and write a refund policy that's clear rather than defensive. A clear policy reduces returns, because it sets expectations before the sale rather than arguing after it. Writing a refund policy people trust has the wording.

VAT: watch the rolling 12 months, not the tax year

The threshold is measured on a rolling 12-month basis, and on your expectations for the next 30 days. That catches people out. You can have a quiet spring, a huge Christmas, and cross the line in February on the strength of the previous December.

gov.uk gave the figure as £90,000 in August 2026, and it said you must register if your total taxable turnover for the last 12 months goes over it, or if you expect to go over it in the next 30 days alone. Check the current figure yourself before you make any decision based on it.

Two things worth knowing before you get near it. Registration is on turnover, not profit, so a business with thin margins can be forced to register while barely making money. And once registered, if you sell mostly to consumers, adding VAT to your prices either cuts your margin or raises your price by a fifth, which is a genuinely difficult month. Sellers who see it coming plan for it. Sellers who don't find out from a letter.

Track it monthly. A single column in a spreadsheet with a rolling 12-month sum is enough, and it takes about four minutes on the first of every month.

Cards, and the arithmetic nobody does

Britain is a card country. Buyers expect a card field, and the friction of asking a British stranger for a bank transfer is higher than it is in Lagos or Bengaluru. So cards are usually the right call here, unlike in a lot of markets.

"Usually" is doing work in that sentence. The maths depends entirely on your volume.

Card payments on Sailo carry Sailo's 1–3% of the goods after discount and excluding delivery and tax, plus your processor's own cut. Stripe's UK pricing page in August 2026 listed standard UK cards at 1.5% + 20p, UK premium cards at 2.8% + 20p, EEA cards at 2.5% + 20p, and cards issued elsewhere at 3.15% + 20p, with a further 2% if a currency has to be converted.

You do not need a paid plan to take cards. The free plan settles them at 3%, and what Business buys you is the rate falling to 1% for $49 a month, billed in US dollars. So the real question isn't whether to turn cards on, it's which plan to be on once they are. Run it at two volumes on an £18 candle:

Card orders a month	Goods value	Free, 3%	Business, 1% + $49	Cheaper
12	£216	£6.48	£2.16 + $49	Free, easily
70	£1,260	£37.80	£12.60 + $49	Free, still

Same product, completely different verdict from the one this table used to give. Business saves you two points of the goods and costs a flat monthly fee, so it pays for itself at $49 ÷ 0.02 — around £1,900 of card goods a month at current rates, well above either row. Below that you are buying the other things Business carries (unlimited products, affiliates, broadcasts, the API), not a cheaper card rate.

So: don't upgrade because it feels professional. Upgrade when two points of your card goods is worth more than the monthly fee, or when you actually need one of the features. If you're anywhere near the line, do you need card payments to sell online works the numbers properly. Watch the 20p as well as the percentage, because on a £6 order the fixed fee is doing more damage than the rate and almost nobody notices which one is hurting.

Bank transfer, meanwhile, still works fine in the UK. Faster Payments is instant and free to the buyer, and repeat customers who already trust you will use it without complaint. The cost is that you confirm every one by hand.

Postage is priced on depth, and it will cost you

The most useful hour a new UK seller can spend is at a Post Office counter with three different boxes.

Postage here is banded by dimensions as well as weight, and the jump between a large letter and a small parcel is decided by thickness, measured against a physical gauge. A product that fits the thinner band costs meaningfully less to send than the identical product in a box 5mm too deep. Over 70 orders a month, that difference is somebody's grocery shop.

Which means packaging is a pricing decision, not a presentation decision. Sellers redesign inserts, swap bubble wrap for corrugated card, and lie flat what they used to stand upright, and the saving lands on every single order for the rest of the year.

Get the current price bands and dimensions from Royal Mail and from one courier directly. Don't take them from an article, including this one. Then work out your actual delivery charge from real numbers rather than guessing at "£3.95 flat" and absorbing the difference. How to work out delivery charges is the method.

And decide in advance what happens when a parcel goes missing, because it will. You're responsible to the buyer for delivery regardless of what the courier says, so "the courier lost it" is not an answer they have to accept. Know the wait period before your carrier will treat an item as lost, know what proof they'll want, and tell the buyer that timeline on day one rather than inventing it under pressure on day nine.

Worked example: Ellie, soy candles, Sheffield

Ellie makes soy candles in her kitchen. £18 for a 200g tin, £32 for a pair, six scents. Around 70 orders a month, mostly through Instagram, a market stall twice a month, and a slowly growing repeat list.

Her numbers, per candle: £5.40 in wax, wick, fragrance and tin. £1.30 in packaging. Postage at the small parcel band. She charges £3.50 delivery and it costs her a bit more than that, so delivery runs at a small loss she's decided to accept because free-ish delivery converts.

She turned cards on in month seven, at around 50 orders a month, and it was the right time. Before that she took bank transfer from her repeat buyers and cash at the market, and reconciled once a day for about six minutes.

The thing that changed her margin most had nothing to do with payments. She was shipping each tin in a box with a cardboard insert and it fell into the small parcel band. She swapped to a rigid mailer, redesigned the insert flat, and moved most single-candle orders into the cheaper band. That change was worth more per year than every payment fee she pays combined, and it took an afternoon and £40 of test packaging.

Her returns run at about two a month, mostly "the scent isn't what I expected". She reduced those by writing scent descriptions that name what it actually smells of rather than what it's called. "Fig" became "green fig, cedar, slightly sharp". Returns halved.

Where Sailo fits, and where it doesn't

Sailo gives you a link like sailo.store/yourname, live at signup, with your products, options and prices on it. Bank transfer, cash on delivery, WhatsApp, Instagram and email orders cost you nothing in commission, because on those Sailo never touches the money. Card is the exception: 1–3% of the goods, on the Business plan, through your own Stripe account, so the money lands with you rather than with Sailo.

The honest limitations for a UK seller:

Cards need $49 a month before the button appears. At low volume that is the dominant cost and it will be worse than doing nothing. The table above is not marketing; it's the reason to wait.

Nothing here handles VAT for you, and nothing here files anything with HMRC. Sailo is a shop, not an accountant. Your rolling turnover, your returns, your Self Assessment: yours.

There's no native app. You run it in a browser, on your phone or your laptop.

The free plan caps at 10 products. Six scents in two sizes is twelve, which fits. Six scents in two sizes plus seasonal one-offs does not, so use options rather than separate products where you can.

What to do next

Two jobs this week, and the first one takes an afternoon.

Take your three best sellers to a Post Office counter and find out exactly which band each one falls into as currently packed. Then find out what it would take to move each one down a band. That's the single highest-return afternoon available to a UK seller and almost nobody spends it.

Second, open a spreadsheet and put a rolling 12-month turnover total in it, updated monthly. If that number is anywhere within twenty percent of the current VAT threshold, book an hour with an accountant now, before the quarter that pushes you over.

If you're still deciding which payment rails to switch on at all, how to take payment as a small seller compares every option by what it costs you and what it leaves on your desk afterwards.