Is starting an online shop in the UK worth it on £18,000?

Executive Summary & Verdict

Promising — 7/10. Place the first order. This plan has the three things a stock-holding online shop needs and most first attempts lack: a category narrow enough to be findable, a margin structure that survives postage and payment fees, and a fixed-cost base low enough that a slow month is boring rather than fatal.

Three findings drive the score. First, the margin is real because you are the importer, not a reseller — buying at Japanese wholesale and selling at UK retail is a different arithmetic from buying at UK wholesale, where published retail markups for books and stationery sit at 30–50% [1]. Second, the compliance load on paper, ink and plastic stationery is unusually light: no UKCA-marked product category, no age-restricted goods, and you are almost certainly under both packaging EPR thresholds at this size [3]. Third, returns in this category are low by e-commerce standards — a bottle of ink is bought by someone who already knows what it is — and the 14-day distance-selling cancellation right is a manageable cost when the average parcel weighs 200 grams [4].

What stops it being an 8 or 9 is entirely in section 6: the cash conversion cycle. Japanese distributors will want payment before shipping, your goods take weeks to arrive, and your stock will take most of a year to sell through. That is roughly five months between spending a pound and getting it back, with no supplier credit to soften it. The sequencing in section 7 is written around that single fact.

Idea Validation

The premise holds because the pain is specific and already being paid. Your customer is not being asked to want something new; she already buys Japanese stationery and already tolerates a three-week wait and the possibility of a customs charge on the doormat. You are shortening a supply chain she is currently operating herself, which leaves only one open question: price sensitivity, not desire.

The part of the answers doing the real work is 140 SKUs. A narrow, deep catalogue in a category people search by exact product name is findable; a broad shallow one is not. Breadth is the trap here — every extra SKU is cash on a shelf, and the twentieth ink colour sells a fraction of what the first five do.

What is genuinely unproven is whether your customer will pay the UK-stocked premium rather than wait. Direct-from-Japan sellers can look meaningfully cheaper on the item line before delivery, and your bet is that convenience, VAT-inclusive pricing and no customs risk covers the gap. It probably does — but it is a bet you can test for under £200 before you spend £8,000 on stock, and section 7 says how.

Market & Target Customer

Your market is not "UK stationery". It is the intersection of three small, well-connected groups: fountain pen users, planner and bullet-journal users, and the brush-lettering crowd. All three are concentrated, vocal and already organised into places you can reach without paying for reach — subreddits, Discord servers, YouTube reviewers with five-figure audiences, and two or three physical pen shows a year.

That concentration is the best thing about this plan and it has a hard edge: a bad first six months is public. In a niche this size reputation is the acquisition channel, and one visibly mishandled order in a 40,000-member group costs more than a month of ads would have bought.

What to measure before the second stock order

• Average order value. UK Shopify AOV sits at roughly £62 across all categories [1], but stationery is a low-ticket category and yours will start lower. Everything in section 6 moves when this number moves, so measure it from order one rather than assuming.
• Which five SKUs are 40% of revenue. They always exist. Finding them early is what turns the second order from a guess into a decision.
• Repeat rate at 90 days. Consumables — ink, refills, paper — should produce repeat purchase. If they do not, you have a discovery problem, not a supply problem.
• Actual return rate, not the benchmark. Public UK returns data does not cover this kind of category; the industry proxies quoted for multi-brand retail sit around 12% [1], and the source flags them as proxies rather than measured UK figures. Your own first hundred orders are better data than any of it.

Competitive Landscape

You have three sets of competitors and they behave differently. The direct-from-Japan sellers compete on item price and lose on delivery time and customs risk. The established UK specialists compete on range and trust and will beat you on both for at least a year. And the general marketplaces carry a thin, badly photographed selection at chaotic prices, which is where most of your early customers are currently shopping and being disappointed.

The contrast worth naming, because it is the reason this plan scores higher than the obvious alternative: a dropshipper never touches the stock and therefore never ties up capital, but also never controls delivery time, packaging, quality or the customer relationship — which is exactly what you are selling. Holding stock is the cost of the differentiator. If you were not planning to hold it, you would have no differentiator at all, and you would be competing on price with the same Chinese suppliers as everyone else.

Where the incumbents are actually weak

• Depth in one thing, not breadth in everything. Being the shop that stocks every colour in one ink line, in both bottle sizes, always, is a position that costs them more to copy than it costs you to hold.
• Photography and product information. Ink swatches on actual named paper, nib-width comparisons, paper GSM and bleed-through notes. Free content that ranks, and the incumbents mostly have supplier stock photos.
• Speed and packaging. Same-day dispatch from a spare room beats a larger operation with a next-working-day cutoff, and a parcel that survives Royal Mail without the ink bottle breaking is an advantage you can build in week one.
• Being answerable. A named person who replies within the hour is worth more here than any feature — and it stops working the moment you scale, so use it while you have it.

Business Model & Monetisation

This is a working-capital business dressed as a retail business. Revenue is a function of how much stock you can hold and how fast it turns, not of how hard you work. Doubling your effort does not double your sales; doubling your stock might.

The model itself is plain: buy at Japanese wholesale, land it, hold it, sell at UK retail with VAT included, post it yourself. Gross margin of about 55% on ex-VAT revenue is achievable at a 2.3× markup on landed cost — meaningfully better than a domestic stationery retailer's 30–50% markup [1], and the entire reason the plan survives postage. Contribution per order matters more than margin percentage, though, because postage and payment fees are close to fixed per order rather than per pound.

The revenue lines, in the order they will actually matter

• Consumables. Ink, refills, paper. Low ticket, high repeat, and the reason a customer comes back four times a year without being marketed to.
• Hardware attached to consumables. Pens and notebooks lift AOV but turn slowly. A few, chosen carefully, not a range.
• Bundles and sampler sets. The highest-margin line in the category, because you are selling curation rather than product — and they turn slow stock into a reason to buy.
• Pen shows. Cash on the day, no postage, no returns, and the fastest way to convert slow stock into working capital. A liquidity tool, not a hobby.
• What to refuse for now. Wholesale supply to UK shops. It looks like growth and it is the fastest way to turn a 55% margin into a 25% one while taking on credit risk.

Finances & Budget

This is the section that produces the verdict, and unusually for these reports the numbers come out in your favour — provided you register for VAT voluntarily on day one. The budget below allocates the whole £18,000, with every line priced from a real published figure.

Whether a parcel makes money: the per-order arithmetic

At a £42 average order, ex-VAT revenue is £35. Gross margin at 55% gives £19.25 of gross profit. Take off postage — a 101–250 g Large Letter is £1.90 at stamp prices or £1.81 via Mailmark, a small parcel up to 2 kg is £4.60 stamped or £4.25 Mailmark [4], and Royal Mail Tracked 48 for a small parcel is quoted at £3.99 on online pricing [5] — call it £2.60 on a realistic mix of letters and parcels. Take off 45p of packaging. Take off Shopify Payments at 2% + 25p on the Basic plan [3], which is £1.09 on £42; Stripe's UK standard card rate of 1.5% + 20p [6] is cheaper on paper but check what your platform charges if you route payments outside its own processor.

Contribution per order is therefore about £15.10. Fixed costs are roughly £85 a month — shop, apps, insurance, bookkeeping — which means six orders a month covers every fixed cost you have. That is the number that earns this plan its verdict. A café needs to fill a room every morning to break even; this needs six parcels.

Now the scale question. Selling £8,800 of ex-VAT landed stock at 55% margin produces about £19,500 of ex-VAT revenue, which at £35 an order is roughly 560 orders. Over nine months that is a little over two orders a day, every day, part-time. It is achievable in a niche this concentrated, and it produces about £8,500 of contribution before your own time — a real side income, not a salary. The plan only becomes a job if stock turn improves or the stock base grows, and both of those need capital you do not currently have. Plan for year one to be a paid apprenticeship, not a replacement income.

MVP & What to Build First

There is no pivot needed here, which is unusual. What is needed is sequencing, because the order in which you spend the £18,000 changes how much of it you get to spend twice.

First, before any stock: prove the price gap is survivable. Spend under £200. Buy one sample set from a direct-from-Japan seller and one from a UK specialist, time both, then put up a one-page site with a waitlist and honest UK-stocked prices beside the direct-from-Japan alternative. If two hundred people in the relevant groups see that and nobody joins the list, you have learned the most expensive thing you needed to know for the price of a takeaway.

Then, in this order

• Register for VAT voluntarily, before the first import. You are nowhere near the £90,000 threshold [1], but voluntary registration is what makes the £1,760 of import VAT on your first order reclaimable rather than sunk, and postponed VAT accounting lets you declare and recover it on the same return instead of paying it at the border [2]. Get the EORI number in the same week — you cannot move the goods without it [3].
• Order narrow and deep, not broad. Half the SKU count you think you need, double the depth on the ones you are confident about. You can restock a winner in six weeks; you cannot un-buy a loser.
• Photograph everything before listing anything. Swatches on named paper are the content that ranks and converts, and one session costs half what five do.
• Solve the breakage problem in week one. Post six ink bottles to a friend in another postcode before you take a single order. A leaked bottle is a refund, a return and a public complaint in the same email.
• Defer: the second supplier, the pen show table, paid ads, and any SKU you cannot explain in one sentence. Reasonable in month six, expensive in month one.

What you are building in the first ninety days is not a shop. It is a reliable answer to one question: how fast does a pound of stock come back as £2.30 of revenue? Once you know that number for real, every subsequent decision — reorder size, range width, whether to borrow — becomes arithmetic instead of nerve.

Go-to-Market & Launch Plan

Your market is small, concentrated and allergic to marketing. That rules out most of a standard launch playbook and leaves four things that work.

• Be useful in the communities before you sell in them. Six weeks answering paper and nib questions in the relevant subreddits and Discords, under your own name, with no link. Then launch. The reverse order gets you removed and remembered.
• Seed two reviewers, honestly. Product to two UK stationery reviewers with no conditions and no script. One genuine review here outperforms any budget you could set, and a scripted one is spotted immediately.
• Launch on availability, not discount. "Every colour in the line, in the UK, posted today" is the message. A launch discount trains a price-sensitive niche to wait for the next one and eats the margin the second stock order needs.
• Book one pen show for month four, not month one. You need to know your top sellers first. A table converts slow stock to cash and teaches you more about objections in a day than three months of analytics.

Marketing & Growth

Two channels matter and the rest are a distraction at this size.

Search, by exact product name. Your customer knows what she wants and types it. A properly built page for each of 140 SKUs, with real swatches, real paper notes and honest in-stock status, is a hundred and forty separate chances to be the answer. Slow, free and compounding — the only channel that keeps working when you stop paying attention to it.

Email, to people who have already bought. A restock notification list for specific colours is the highest-converting email any shop in this category can send. Collect the address at checkout, ask what they want back in stock, tell them when it is. That is the whole programme.

On paid: median cost per acquisition across paid channels is quoted around $33, and Meta's median around $38 [1]. At a £42 average order and £15 contribution, paid acquisition does not pay for a first order at anything like those numbers — it only works if the second and third orders are reliable. Measure repeat rate first. Spending the £1,500 launch budget on ads before you know that number is the most common way a plan like this loses its working capital.

Pricing Strategy

Price at UK retail, inclusive of VAT, and do not try to match the direct-from-Japan price. You cannot — they are not paying UK postage and not carrying the stock. Competing on the item line means giving up the margin that funds the whole operation to win a customer who was never buying convenience in the first place.

• Target 2.3× landed ex-VAT cost as a default, which produces the 55% gross margin section 6 is built on. Published retail practice is keystone — 100% markup, or 50% gross margin [1] — so 2.3× is a modest importer's premium, not gouging.
• Show prices VAT-inclusive, always. You are selling to consumers; a price that grows at the checkout is both a conversion killer and a consumer-law problem.
• Set a free-delivery threshold above your natural AOV, not below it. At a £42 average order, free delivery over £50 pays for itself in basket size. Free delivery on everything gives away £2.60 on a £12 bottle of ink.
• Do not discount to move slow stock — bundle it. A sampler set at full margin moves the same units as a 30% sale and does not teach anyone to wait.
• Review landed cost quarterly. You buy in yen and sell in sterling; a 10% currency move is a 10% margin move, and quarterly review is the difference between noticing and discovering.
• Keep one honest cheap thing. A £4 add-on item is how a niche shop gets a first order out of a cautious customer.

Legal Setup & IP

The structure itself is simple. You must register for Self Assessment as a sole trader once you earn more than £1,000 in a tax year, and you can start trading before registering [1]. A limited company is worth a conversation with an accountant before you import, but it is not a precondition for a shop of this size.

The part most people get wrong: importing makes you the producer. Under the General Product Safety Regulations 2005, "producer" includes importers and own-branders, not just manufacturers. That means your obligations are to supply a safe product, provide clear, legible and durable warnings and instructions, show the name and address of the producer in Great Britain on the product or packaging along with batch or reference information, sample and test what you sell, keep a register of safety complaints, and notify your local Trading Standards service in writing immediately if you discover a product is unsafe [2]. Failing these is an offence. For ink, paper and pens this is largely paperwork and a label — but it is paperwork you now own rather than the Japanese manufacturer.

Which product categories change this picture

• Your chosen category is on the easy side. The GPSR are a "mop-up" regime that does not apply where other UK law covers all safety aspects [2], and paper, notebooks and most pens sit outside the conformity-marking regimes entirely.
• Where UKCA or CE marking bites. More than twenty product types are covered by the UK conformity-marking regime, and as the importer into Great Britain you must verify the manufacturer completed the conformity process and obtained the correct marking, keep the technical documentation, ensure storage and transport do not compromise compliance, and stay in contact with the manufacturer. If you sell under your own name or trademark, or modify the product, you take on the manufacturer's full obligations, including affixing the marking yourself [3]. That is the trap in own-branding an imported product.
• Categories to avoid at this size. Anything electrical or battery-powered (electric erasers, USB desk lamps), toys or anything marketed to children under 14, cosmetics and anything skin-contact, and anything with a blade. Each drags in its own regime, its own testing and its own documentation, and one of them can cost more in compliance than your entire stock budget.
• Own-branding. Putting your logo on a notebook you import is commercially tempting and legally expensive. Do it only with a category you are certain carries no conformity obligation, and take advice first.

The rest of the legal setup

• Consumer Rights Act 2015. Every goods contract includes a term that the goods are of satisfactory quality, judged by what a reasonable person would consider acceptable, taking into account description, price and all other relevant circumstances — and including fitness for purpose, appearance and finish, freedom from minor defects, safety and durability [4]. The short-term right to reject runs for 30 days from the day after ownership, delivery and any installation have all happened [5]. That is separate from and additional to the 14-day cancellation right in section 14.
• Trade marks. Search the UK register before you pay for packaging. A shop name that collides with an existing stationery mark is a rebrand you will pay for twice.
• Terms, returns policy and business address. Written before launch, not after the first dispute. Section 14 explains why the returns policy specifically is a cash decision, not a legal formality.
• Insurance. Product liability matters more than people expect once you are the producer, and household contents cover usually excludes business stock. Tell the insurer what is in the spare room.

Tax, Invoicing & Admin

VAT is the single most consequential decision in this plan, and the answer is counter- intuitive: register before you have to. Compulsory registration bites when your total taxable turnover over the last 12 months goes over £90,000, with 30 days from the end of that month to register, or when you expect to cross it within 30 days [1]. You will not be near that. But a stock-holding importer who is not registered pays 20% import VAT at the border on every consignment and never gets it back — on this plan that is £1,760 on the first order alone, which is nearly twice your contingency.

Register voluntarily and the arithmetic inverts. Import VAT becomes input tax you reclaim, and postponed VAT accounting lets you declare and recover it on the same VAT return rather than paying it at the border and waiting [2]. No approval is needed; you select it on the import declaration, quote your VAT number, and pick up your postponed import VAT statements online. The VAT is recorded against your EORI number at declaration level, and you cannot change the method after the declaration is submitted — so brief your customs agent in writing before the first shipment, not after.

The cost of registering is that you then charge 20% output VAT on UK consumer sales. At a 55% gross margin with a large stock purchase, reclaiming input VAT on that purchase plus Shopify, packaging, postage and freight comfortably beats the output VAT you hand over in year one. Run the numbers with an accountant for your actual first-year sales forecast before you commit, because the answer flips if your stock purchase is small relative to your sales.

Importing: the mechanics, in the order they happen

• EORI number. You need one to move goods between Great Britain and any other country [3]. Apply before you place the order, not when the goods are on a plane.
• Commodity codes. Every product needs one — an internationally recognised reference number, of which the first six digits are used worldwide — and it determines the rate of customs duty and import VAT, plus licensing requirements, preferential rates, anti-dumping measures and duty reductions [4]. Classifying is genuinely difficult; GOV.UK itself recommends the Trade Tariff tool or a customs professional. Get the codes right once, at the start, for all 140 SKUs.
• Customs duty and the UK–Japan agreement. Do not assume zero. Most tariff commitments between the UK and Japan transitioned into CEPA without change from the previous EU–Japan arrangement, with duty eliminated on a limited additional set of industrial lines [5]. Check the actual rate for your codes. Useful detail for a small importer: goods valued under £1,000 entering the UK need no statement of origin [5], so your first small test consignment is simpler than your first real one.
• Import VAT value. Not just the invoice. It is the customs value plus all charges payable on importation such as customs duty, plus incidental expenses, excluding the VAT itself [6]. Freight is in the base. Budget accordingly.

Income tax and record-keeping

• Self Assessment deadlines. Tell HMRC by 5 October if you need to file for the previous year and have not filed before; paper returns by 31 October; online returns and payment by 31 January; a second payment on account, where it applies, on 31 July [7].
• Making Tax Digital for Income Tax. Sole traders and landlords with qualifying income over £50,000 should have started using it from 6 April 2026; the threshold drops to £30,000 from 6 April 2027 and £20,000 from 6 April 2028 [8]. You will almost certainly be inside the £20,000 band before 2028, so buy MTD-compatible bookkeeping software now rather than migrating later.
• Stocktake discipline. Your profit is not your bank balance. With over half your capital sitting on a shelf, a monthly stock count reconciled to the shop's inventory is the only way you will know whether the business is working. This is the admin task most likely to be skipped and most expensive to skip.
• Separate business bank account and a yen-cost record. Log the exchange rate on every purchase order. Landed cost per SKU, in sterling, on the day you paid, is the number your pricing depends on.

Tech, Hosting & Security

You have two sensible routes and the difference is not money, it is where your attention goes.

Shopify. Basic is £25 a month billed monthly or £19 on annual billing, with Shopify Payments at 2% + 25p for online standard cards; Grow is £65 or £49 annual and drops the card rate to 1.7% + 25p [1]. At your volume Basic is correct, and at 560 orders a year the Grow card rate saves a trivial amount — move up when order volume, not ambition, justifies it.

WooCommerce. Cheaper in licences, more expensive in hours. Published breakdowns put a starter store at roughly $75–$900 a year for hosting, domain, SSL and a theme, and a fully functional store at $200–$1,500+ once essential plugins are included, with WooCommerce Payments at 2.9% + $0.30 per transaction [2]. The money is comparable; the difference is that on a bad Sunday you are fixing a plugin conflict instead of packing parcels. On a part-time plan with a day job, that decides it, and it points at Shopify.

The unglamorous list that actually matters

• Inventory accuracy above everything. Overselling a bottle of discontinued ink to a collector is worse than any site outage. One source of truth, counted monthly.
• Never store card details yourself. Use the platform's hosted checkout and keep your PCI scope as close to nothing as it allows.
• Two-factor authentication on everything — shop admin, email, bank, registrar, supplier portals. A compromised admin is a compromised customer list.
• Back up the product photography separately. Shooting 140 SKUs is your second-biggest non-stock investment; losing it is £600 and three weekends.
• Fraud review on high-value orders, and address validation at checkout — a failed delivery costs postage twice plus the goodwill.

Regional Compliance & Privacy

Three layers apply to a UK online shop holding its own stock, and one of them is new enough that most guides still do not mention it.

Distance selling: the 14-day right, and who pays the postage

Consumers have 14 days to cancel most distance contracts, and for goods that period begins the day after the items arrive. Once they have told you they are cancelling, they must return the goods within 14 days. You must refund without undue delay and within 14 days from the day after they told you, or — where goods are being returned — within 14 days of receiving them or of proof of postage [1].

The part that costs money: normally the consumer bears the cost of returning the goods, but if you failed to tell them that beforehand, you must pay those costs yourself [1]. One missing sentence in your pre-contract information converts every return into a postage bill. There is a second, smaller trap in delivery refunds: you must refund the basic delivery cost of getting the goods to the customer, but if they chose a premium service you only have to refund your standard rate, not the upgrade [1].

Note that this 14-day change-of-mind right sits on top of the 30-day short-term right to reject faulty goods under the Consumer Rights Act [2]. They are different rights with different triggers and different consequences, and a returns policy that conflates them will be wrong in both directions.

Packaging EPR: the cost nobody budgets for

If you supply or import packaging in the UK, extended producer responsibility can apply. The obligation is triggered by an annual turnover of £1 million or more and supplying or importing more than 25 tonnes of packaging [3]. You are a "small" organisation if turnover is £1–2 million with over 25 tonnes, or turnover over £1 million with 25–50 tonnes; you are "large" if turnover is £2 million or more and you handle more than 50 tonnes. Small producers record and report data annually; large producers report on a six-monthly schedule, pay waste disposal fees, pay scheme administrator and registration charges, obtain PRNs or PERNs, and report nation data [3].

On this plan you are comfortably outside both thresholds, and that is worth knowing rather than guessing about. But it is worth knowing what it would cost if you grew into it, because the fees are material: the confirmed disposal-cost fees for the 2025–26 year run from £192 a tonne for glass and £196 for paper and card up to £423 for plastic and £461 for fibre-based composite, with the same rates applying to small and large producers [4]. Illustrative year-two fees have been published and are broadly similar to year one overall, though individual materials have moved sharply — wood rose 60%, from £280 a tonne to a 2026 amber fee of £450 [5]. The practical consequence for a shop your size is a design decision you can make for free right now: choose paper and card mailers over plastic, and you have both a cheaper future EPR position and a better-looking parcel.

Privacy and data

UK GDPR applies from your first customer. The discipline that matters is restraint: collect the name, address, email and order history you need to fulfil and support the order, and nothing else. Date of birth, phone number and "how did you hear about us" are all data you will have to protect, justify and eventually delete. Have a privacy notice on the site before launch, a cookie banner that actually honours a refusal, and a written answer to what you do when someone asks for their data to be erased. A one-person shop can comply with all of this in an afternoon; it just has to be the afternoon before launch, not the one after the first complaint.

Risks & Common Pitfalls

• Dead stock — the risk that actually kills shops like this. Not lack of sales; the wrong sales. Twenty SKUs that never move tie up £1,500 you needed for the reorder of the five that do. Review sell-through by SKU every month from month two and be ruthless about discontinuing.
• Buying breadth instead of depth. 140 SKUs is already at the upper end for £8,000 of stock. Every plan like this drifts wider because wider feels like a better shop. It is a worse balance sheet.
• Not registering for VAT, then importing. £1,760 of unrecoverable import VAT on the first order [1], repeated on every consignment. This is the single most expensive avoidable mistake available to you, and it is invisible until an accountant points at it a year later.
• Getting commodity codes wrong. The code determines duty and import VAT and licensing [2]. A wrong classification across 140 SKUs is a retrospective correction across every consignment you have ever landed.
• The missing sentence in the returns policy. Fail to tell customers they pay return postage and the cost falls on you [3]. At 12% returns on 560 orders that is roughly 67 return postage labels you did not budget for.
• Becoming the producer without noticing. Importing makes you the producer under the GPSR, with traceability, warning, monitoring and notification duties [4]. Own-branding an imported product can hand you the manufacturer's full conformity obligations [5]. Both are manageable; neither is optional.
• Leakage and damage in transit. A glass ink bottle in a padded envelope is a refund waiting to happen. Solve this before launch, not after the first one-star review.
• Currency exposure. Buying in yen, selling in sterling, with no hedge and a five-month cash cycle. A 10% move against you erases a fifth of your gross margin on stock already ordered.
• The spare room is a hard ceiling. 2.6 m × 3.1 m holds a certain amount of stock and no more. At the point this business starts working, your next constraint is physical, and storage costs money you were planning to spend on inventory.
• The day job. Part-time is the right call and it caps your throughput. Two orders a day is fine in a spare hour; twenty is not, and the transition happens faster than the house can absorb.

Action Plan: 30 / 60 / 90 Days

This plan assumes you accept the sequencing in section 7: cheap demand test first, then registrations, then stock. Reversing that order is how a funded plan becomes an underfunded one.

Days 1–30 — test the price gap and get the paperwork moving

• Put up the one-page site with honest UK-stocked prices against the direct-from-Japan alternative, and a waitlist. Under £200.
• Start the six weeks of being useful in three communities, under your own name, no link.
• Register for Self Assessment as a sole trader [1]; open a separate business bank account.
• Register for VAT voluntarily [2] and apply for an EORI number [3] — both before any order is placed.
• Get quotes from both Osaka distributors, confirm payment terms in writing, ask for the commodity codes they use [4], then verify them yourself against the Trade Tariff.
• An hour with an accountant to sanity-check the VAT decision against your real forecast.

Days 31–60 — buy narrow, land it properly

• Place the first order: half the SKU count, double the depth, around £8,000 at wholesale, split across two consignments so the second can correct the first.
• Brief the customs agent in writing to use postponed VAT accounting — the method cannot be changed after the declaration is submitted [2].
• Build on Shopify Basic [5]: product pages first, design last.
• Write the terms, privacy notice and returns policy, and make sure the returns policy states explicitly who pays return postage [6].
• Insurance — stock in the home, public and product liability. Tell the insurer what is in the spare room.
• Test the packaging: six ink bottles to two postcodes, opened by you.

Days 61–90 — launch, then measure the only number that matters

• Photograph and list all 140 SKUs in one push. Swatches on named paper.
• Launch on availability, not discount. Seed two reviewers, no conditions.
• Record AOV, conversion, return rate and repeat rate from order one. Compare them to last week, not to benchmarks.
• Count stock at month end and reconcile it to the shop. Then every month.
• File the first VAT return and confirm the £1,760 of import VAT actually came back [2].
• At day 90, calculate sell-through by SKU and your real cash conversion cycle. If a pound of stock is coming back as £2.30 in under five months, place the second order and start planning for the day the spare room runs out. If it is slower, the answer is narrower range and deeper stock, not more range.