Most European brands approach Amazon the same way: allocate a modest budget, list a handful of SKUs, and see what happens over a quarter. The logic feels prudent — spend little, learn fast, scale if it works. In practice it produces the opposite of learning. A launch that runs out of inventory in week six, or gets suppressed because a Responsible Person field was left blank, tells you nothing about whether the market wants your product. It tells you the plan was incomplete. At INNELS we have watched enough of these quarters end inconclusively to be direct about it: Amazon rewards preparation rather than experimentation, and almost every input a brand needs to plan properly is publicly knowable before launch. Fees are published. Compliance deadlines are legislated. Registration lead times are documented. What separates a viable Amazon business from an expensive test is whether somebody did the arithmetic first.
A test tells you what happened. A model tells you what has to be true for this to work — and you can check those assumptions before you spend anything.”— Andrejs Klimovskis, Founder
A market entry model is not a forecast, and it is not a pitch deck. It is a structured answer to the four questions a board will ask anyway: what will this cost, when does it break even, how long until we are live, and what could legally stop us. We build ours around those four inputs because each one is independently verifiable — the cost stack from Amazon’s published fee schedules, break-even from your own landed costs, the timeline from documented registration lead times, and the requirements from EU regulation. When a brand asks whether Amazon is worth entering, we do not answer with an opinion about their category. We populate the model with their real numbers and let the output speak. Sometimes the honest conclusion is that a SKU does not clear the fee stack at its current retail price, and that answer is worth far more delivered in a planning session than discovered after a year of inventory sitting in a German fulfilment centre.
Every number in a market entry model should be traceable to a source. The moment you are estimating, you have stopped modelling and started hoping.”— Mark Daniel Zalomajev, Founder
There is no mystery in what Amazon charges. A Professional selling account costs €39 per month excluding VAT and covers the European stores under one unified account. Referral fees run from five to forty-five percent by category, though most consumer goods land in the eight to fifteen percent band, and once fulfilment, storage, returns and surcharges are layered on, the total take across most categories sits somewhere around twenty to twenty-five percent of the sale price. Amazon reduced European referral and FBA fees for 2026 by an average of seventeen cents per unit, then added a 1.5% fuel and logistics surcharge to FBA fulfilment fees from 17 April 2026 — exactly the kind of detail a spreadsheet built last year quietly misses. Fulfilment structure matters just as much, because Pan-EU lowers the per-unit cost but obliges you to register for VAT and EPR in every country holding your inventory.
Sellers underestimate Amazon by about ten points of margin, almost every time. Not because the fees are hidden, but because nobody added them up before agreeing the retail price.”— Niks Saknitis, PPC Manager
Break-even on Amazon is contribution margin arithmetic, and it behaves predictably once you stop treating advertising as a variable to be figured out later. Take the landed cost, subtract the referral fee, the fulfilment fee, monthly storage, a realistic returns provision, any promotional spend, and then the share of revenue advertising will consume while the listing still has no organic history. That last input is where most models quietly fail, because launch-phase advertising is structurally more expensive than mature-phase advertising — a new listing has no ranking, no reviews and no conversion history, so it buys its early velocity. Industry experience consistently puts launch efficiency well below mature efficiency, but those figures are working rules of thumb rather than surveyed benchmarks, so we model them as ranges rather than points. If a SKU only clears break-even at mature advertising efficiency, it does not have a launch plan. It has a hope that the expensive months will be short.
EU compliance is the part of Amazon entry that cannot be tested, only satisfied. The General Product Safety Regulation has required an EU-established Responsible Person since December 2024, with their details shown on the listing and on the product itself. Extended Producer Responsibility is national rather than European, so packaging, electronics and battery obligations are handled country by country, and Amazon has enforced battery EPR numbers across eight European markets since August 2025. From 12 August 2026, marketplaces must verify packaging EPR registration in every EU country where a seller sells or stores inventory, which ends any remaining ambiguity about registering selectively. Add VAT registration lead times that range from roughly a fortnight in Italy to several months in France, plus Brand Registry, which needs a registered or properly filed trademark before it will approve, and the realistic gap between deciding to enter and holding sellable inventory is rarely under four to eight weeks. None of that is a risk to discover. It is a schedule to work backwards from.
Responsible Person details, EPR marks and language requirements all land on the packaging and the listing. If compliance arrives after the design is signed off, you pay for the artwork twice.”— Yulia Hurenko, Graphic Designer
A launch budget that funds inventory and advertising but not listing quality is a budget designed to waste the advertising. Amazon’s ranking system is A9 with a COSMO layer interpreting shopper intent, and while ranking decides what gets surfaced, conversion decides whether that traffic pays for itself. Amazon’s own data indicates standard A+ can lift sales by up to around eight percent and Premium A+ by up to roughly twenty percent, but A+ text is not directly indexed, so it supports conversion rather than ranking. We include creative production in the entry model as a dated, costed workstream for the same reason we include VAT registration — photography, localisation, A+ modules and compliant packaging artwork all carry lead times, and every week a listing runs without them is a week of paid traffic converting below its potential. Traffic is the expensive part. Arriving unprepared to receive it is the avoidable part.
Paid traffic does not fix a listing that is not ready. It just makes the gap more expensive per click.”— Niks Saknitis, PPC Manager
A complete model answers four questions with sourced numbers rather than estimates. The investment question needs the landed cost per unit, the full Amazon fee stack for your specific categories and fulfilment choice, opening inventory volume, creative production, registration costs, and a launch advertising budget held as a range rather than a single figure. The break-even question needs contribution margin per unit modelled at launch-phase advertising efficiency and again at mature efficiency, so the destination and the cost of the journey are both visible. The timeline question needs dated lead times for VAT registration in each relevant country, EPR and GPSR registration, trademark and Brand Registry status, creative production, and inbound freight through to sellable stock. The requirements question needs a category-specific compliance list — packaging, batteries, cosmetics notification, supplement claims, toy standards — checked against the marketplaces you actually intend to sell in. If any of the four is missing, what you have is not a plan. It is a budget with optimism attached.
We would rather tell a brand in week one that the numbers do not work than in month twelve. Both are honest answers. Only one of them is cheap.”— Andrejs Klimovskis, Founder
Amazon in Europe is no longer a speculative channel. In 2025, EU-based small and medium businesses passed forty billion euros in sales on Amazon for the first time, with more than 127,000 European SMEs selling through the platform, and a market that mature does not reward exploratory launches. The brands that struggle are rarely the ones with the wrong product; they are the ones that committed capital before they understood what the channel would charge them, how long it would take to reach break-even, and which registration would gate their listing in which country. A model does not guarantee the launch succeeds. What it guarantees is that if the answer is no, you find out in a planning document rather than in a warehouse — and if the answer is yes, you enter with a budget calibrated to the months that actually cost money rather than the ones that do not.
%20(1).png)
.png)
.jpg)
.jpg)
.jpg)
.jpg)