How we pick

Start with what we do not do

We have not tested any of this equipment. We do not own it, we run no test kitchen and no lab, and there is no bench in a garage somewhere with fourteen espresso machines on it. Nothing on this site will ever say “in our testing”, because it would not be true.

That is unusual to lead with, and it is deliberate. Most sites in this category open with a testing claim. When you look for the protocol behind those claims — what was measured, with what, how many units, over what period — it is very often not there. We would rather be the site that tells you what it actually did.

• We accept no free products.
• We take no payment for placement, and no brand has ever seen a page before it shipped.
• We publish no star ratings and no scores of our own, and we emit none in structured data.
• We do not quote a price we have not verified. Where we have not, the page says “check price”.

What we do instead

Four things, all of which you can check yourself.

1. Compile published specifications, with the source named

Every specification on a product card links to the manufacturer page it came from. If a maker does not publish a figure, the row does not exist — we do not fill the gap with a number from a review or a retailer listing. That is why some product cards carry a full spec table and some carry none, and why the ones with none say so in plain words rather than quietly showing something else.

2. Do the arithmetic, and show it

The clearest example is what a setup costs to run. Not one competitor we looked at publishes a cost-per-shot figure. We compute it from a bean price you choose, a dose the manufacturer states, and an electricity price from the U.S. Energy Information Administration — currently 18.34 cents per kilowatthour, June 2026, U.S. residential average. Every formula on that page is shown so you can substitute your own numbers and get an answer for your kitchen rather than ours.

3. Cite published standards rather than opinion

Where a page states a brewing parameter — a ratio, a shot time, a grind description — it comes from a named published source, most often the National Coffee Association, and it is linked at the bottom of the page. “Most people say” is not a source.

4. Say who each pick is wrong for

Every product on this site carries a “skip it if” line — 47 of 47 products, which is all of them. A ranked list where everything is recommended is an advertisement. A list that tells you what not to buy is advice, and it costs us commission, which is the point.

How the picks are ordered

In this priority, and ties break downward:

1. Does it do the job for a beginner? A machine with a non-standard basket or a grinder that cannot reach espresso fineness is disqualified regardless of price.
2. Does it fail the buyer in a way they will notice within a month? Anything that pushes someone toward giving up ranks lower, whatever its spec sheet says.
3. Is it standard? A 54 mm or 58 mm basket means every accessory and every recipe written by anyone else applies. Non-standard sizing is a real cost.
4. Is the specification published? A maker that documents its product properly is one we can hold to what it said.
5. Availability and parts. Espresso machines have gaskets and seals that wear. A product with no parts supply has an expiry date.

Price is not in that list, because the pages are organized by budget instead. Within a budget page, the ordering above applies to what is available at that money.

Where the prices come from

A single data layer, refreshed on a schedule, that writes one file the whole site reads. It has one rule that matters: a price older than 48 hours is not shown. The page falls back to “check price on Amazon” instead. A number with a date you can see is useful; a number with no date is a liability, and the reason most sites in this category show neither is that keeping prices current is work.

There is no live price data on this build, so every buy link currently reads “check price on Amazon”. That is the system working as designed rather than a page that failed to load.

How often things are refreshed

• Prices: on a daily schedule, with the verification date shown next to every price.
• Pick order: monthly, and immediately if a product is discontinued.
• Specifications: quarterly, against the manufacturer page they came from.
• Cost-to-run inputs: quarterly, and whenever the EIA publishes a new average.
• This page and the other policy pages: every six months, or immediately if anything about how we earn changes.

How we earn

Affiliate commission. Crema Method participates in the Amazon Services LLC Associates Program, and links to products go through a redirect on this domain that adds our affiliate tag. If you buy something, we get a percentage and you pay exactly what you would have paid anyway.

Commission rates vary by product category, and we will state the obvious risk that creates: it would be possible to rank higher-paying products first. We do not, the ordering criteria above are the whole ordering, and the “skip it if” line on every product is the most visible evidence of that — it exists to talk some readers out of buying. The full disclosure is here.

Who writes this

One person. Scooter M., an enthusiast who reads the manuals, compiles the published specs and does the math. No lab coat, no test kitchen, no pretence of having owned forty espresso machines. More about that here.

How to tell us we are wrong

Email info@cremamethod.com. If a figure on this site is wrong we will check it against the source, correct the page, and note the correction on it — that is written into our editorial policy as a commitment rather than an intention.

This site currently has 72 content pages plus its trust pages. If a claim on any of them cannot be traced to the source we cite, that is a defect and we would like to know about it.