The four rating criteria
Every rating on TopKaffeeReviews uses the same grid. That makes scores comparable across products and categories.
1. Build quality (materials and manufacturing quality)
Housing material and stability, quality of the brew group and portafilter (brass or aluminium), design of the boiler or thermoblock, for grinders the precision of the burrs and bearings, panel gaps, the feel of dials and levers. A machine with a rattling tank or a wobbly portafilter cannot be a good buy, however shiny the housing looks.
2. Performance (brew quality and consistency)
Extraction quality and repeatability, temperature stability over several shots, steam power for frothing milk, heating-up behaviour. For grinders, what counts is the uniformity of the grounds, adjustment precision and retention; for filter machines, a brewing temperature in the ideal window of 92 to 96 °C and even wetting.
3. Handling (everyday suitability)
Operating logic, heating-up time, cleaning and descaling effort (removable brew group, rinse programmes, milk system cleaning), grinder noise, tank size and filling, space needed. The best espresso is of little use if the daily care becomes a punishment.
4. Value for money
What do you get for your money, compared with the direct competition? A €300 entry-level espresso machine with small weaknesses can score better here than an almost perfect €1,500 dual boiler.
Where our data come from
We work with three levels of sources: (1) official manufacturer data, data sheets and user manuals (brewing system, boiler, grinder, power, dimensions, care), (2) catalogue data from the Amazon catalogue, such as price and manufacturer description, each with the date it was retrieved, and (3) documented results from independent assessment institutes such as Stiftung Warentest (a German consumer organisation) or ETM Testmagazin, only with source and date and only where they exist. We do not evaluate customer reviews or sales ranks. Where a figure is an estimate, we mark it ("approx.").
We do not invent lab trials, measurements or customer reviews. If no reliable information is available for a criterion, it is left out. A gap beats a claim.
Where AI helps and where it does not
Part of our texts is written with AI assistance: it researches data sheets, organises facts and drafts texts. Nothing is published without editorial review. Certification marks, warranty periods and third-party results are checked against primary sources; retailer listings expressly do not count as evidence. An automatic check blocks publication if such a claim appears in the text without evidence. The English pages are AI-assisted translations of the German pages.
Our images are not AI-generated. Product photos come via the official Amazon interface, category images are licensed stock photography. Should an AI image ever be added, it will carry the label visibly in the image itself. Details are on the AI transparency page (KI-Transparenz, in German).
Badges and awards
Per category we award at most one "Comparison winner 2026" badge (highest overall score) and one "Value pick" badge (best value for money). In addition there are the editorial labels "Beginner's pick", "Premium choice" and "Editor's pick". All badges are based on our rating, none on sales figures. No manufacturer can buy a badge.
Affiliate funding, and why it stays neutral
TopKaffeeReviews is funded through affiliate links (including Amazon PartnerNet). If you buy through one of our links, we receive a small commission; your price does not change. Since the commission rates within a product category are practically identical, we have no incentive to talk up a particular model. That is why our ratings also include clear drawbacks for every product.
Updating
Every rating carries a visible "Last checked" date. Prices are guide values and can change at any time; the click to the shop always shows the current price of the day. On the English pages, prices are in euros from Amazon.de and may differ in your country. Older reports are revised as soon as something relevant changes on the market (successor models, price movements, new results from independent assessment institutes).