How we test.
How we test espresso machines: every machine runs the same five-part gauntlet below — same criteria, same weightings, same reference recipe — scored against measured performance rather than marketing copy. Ratings are earned in shots, not spec sheets.
01.Shot quality & consistency
Every machine is graded against the same reference recipe — roughly 18g in, 36g out, 25–30 seconds. What matters isn't the best shot a machine can produce on a good day; it's how close the average shot lands to that best one. We weigh published shot-by-shot measurements, long-run owner reports, and the hardware that drives repeatability: basket size, pump behavior, pre-infusion. A machine that nails Tuesday and sours Wednesday gets told on.
02.Temperature stability
Sour-then-bitter shots are almost always temperature swings, so we go looking for the numbers: documented group-head readings across back-to-back shots and recovery time after steaming. Machines with PID control should hold within a couple of degrees. Machines without it get judged on how predictable their drift is — some you can surf, some you can only forgive.
03.Steam & milk performance
Steam gets scored on the boiler behind it (dedicated boiler, heat exchanger, or a thermoblock doing its best), wand design, and the reported time-to-texture for a small pitcher of whole or oat milk. The question is whether the wand can actually stretch milk or only heat it. Auto-frothers are measured against what a competent human gets from the same machine. The human usually wins. Not always — which is exactly the kind of thing we report.
04.Workflow & annoyance audit
Tank access, cup clearance, drip tray capacity, warm-up time, cleaning cycles, beeping. We comb long-term owner reports for the stuff nobody mentions on day one and keep a petty little list of everything that makes people sigh six months in. The petty list becomes the 'What's annoying' section. You're welcome.
05.Value, honestly
The final rating weighs performance against what the machine costs and what else that money buys. A $150 machine that makes decent espresso can outscore a $700 machine that makes good espresso, because ratings answer the only question that matters: should YOU buy it?
The independence part
No brand pays for coverage, no brand sees a review before you do, and no brand has ever changed a verdict. Affiliate commissions are the same whichever machine you buy, which means we have no financial reason to steer you wrong. Our only incentive is that you trust the next review too.