Week after week, we get offers. Manufacturers and retailers write to us: test our machine, we'll pay you 10,000 euros, you'll get the unit, we'll feature you in our campaign. We've had offers of every kind — a car hasn't come up yet, but pretty much everything else has.
Our answer is always the same: no.
We buy every machine ourselves
This isn't a marketing claim. We buy our test units, often at full retail price, sometimes at the B2B price a dealer would pay. But we pay. We insist on an invoice — and sometimes we have to ask three times before we finally get one. Some manufacturers seem to mean well by not sending one, but that's exactly what we don't want.
Anyone who seriously doubts this is welcome to hire an independent trustee to review our test units and the corresponding invoices. We're completely open to that.
Why we order directly from manufacturers
We used to buy machines anonymously through retail. That kept landing us with units running outdated firmware, or leftover stock that had been sitting in a warehouse for a while. Not the best starting point for a fair test.
Today we order directly from the manufacturer whenever possible. That ensures we're testing the current model with the latest software. At the same time, we have a direct line to them if we have questions or notice something off.
Does that mean we get specially fine-tuned units? No. The machines still come with some amount of offset. Sometimes we'd even wish someone had double-checked them one more time.
What happens when we find a fault
If we find a fault during a test, we contact the manufacturer. In these cases we order a second unit — we don't buy that one; it's provided to us and goes back afterward. We then check whether the fault is confirmed or whether it was a one-off.
With modern machines that get OTA updates, it also happens that the manufacturer fixes the fault via a software update within a few days. We factor that into our assessment, because it shows a real quality of the machine: it can be updated. A static machine can't do that.
We then publish the test. But that's not the end of it. Many of these machines show up again later in our home barista courses. That's how we see, over time, whether our test results hold up in practice. We'll also redo some tests — not because the original tests were bad, but because our testing protocol has become stricter and some manufacturers have changed how consistently they execute on quality.
How we fund ourselves — without selling out the tests
Our business model draws a clear line between what we test and what we sell. We test espresso machines and espresso grinders. We sell coffee and accessories. We drew that line deliberately. We don't sell machines and grinders, because we test them.
It wasn't always this way. We used to have a shop in Basel where we also sold machines. We added the machines we'd tested well to the shop — not the other way around. Even so, we noticed that the role of a neutral referee doesn't sit well with selling. So we stopped.
For the same reason, we no longer do consulting for individual machines. We can't help a manufacturer develop a product and then review that same unit. The two don't mix.
There are no paid partnerships with coffee companies in our videos. No sponsorship format for machine brands. We looked into it, and it doesn't fit what we want to do. We're open to sponsorships outside the coffee world — sportswear, personal care products, the barber shop down the street in Basel. But nothing that touches our independence when it comes to coffee equipment.
Our testing protocol has a version number
We follow a documented testing protocol, which we explained in detail in a dedicated video. This protocol recently got a version number, linked in every test report. That way you can see exactly which version of the protocol a machine was evaluated against.
Still, we're shaped by our own history and preferences too. We've known some manufacturers for years, we've visited their factories, we've shared meals with them. That's just reality. That's exactly why we measure ever more precisely and look for data that can be objectively verified. Our word alone shouldn't be enough — the numbers need to carry the assessment.
Don't buy blindly on our recommendation
We provide a lot: temperature stability, energy measurements, noise levels, extraction data. We want you to be able to walk into a specialty shop with that data and make your own decision. But form your own picture too. Come to one of our courses, go to a trade show, try machines out at a specialty retailer. Gather your own experience and treat our data as one foundation among several.
And if you want us to test a specific machine: tell us in the comments. The more often we hear a request, the sooner that unit lands on our testing calendar. The list is long, and a test takes time. But we're listening.
If you want to support our work: buy our coffee or our accessories. The cleaning powder, for example. That's how we fund the tests — and stay independent.
