How We Research
Cafe & Crema publishes research-based assessments of home coffee equipment. The method is designed to make the evidence visible and to keep different kinds of evidence separate. A manufacturer specification is not an independent test, an owner report is not a failure-rate study, and an affiliate link is not proof that a retailer has an item in stock.
Evidence hierarchy
Manufacturer documentation comes first for current specifications. Current product pages, instruction manuals, service documents and parts diagrams are the primary sources for model identity, dimensions, capacities, controls, materials, maintenance procedures, compatibility and manufacturer warranty or service statements.
Independent testing stays attributed. When an established publication or specialist has measured temperature, grind distribution, retention, workflow or another performance characteristic, Cafe & Crema identifies the source and describes what that source supports. We do not rewrite someone else's test as though we performed it.
Repair and service evidence is evaluated separately. Public parts catalogs, service manuals, authorized-service pathways and repair specialists help show how a product can be maintained or repaired. Parts availability is evidence of serviceability; it is not proof of a longer lifespan or a lower failure rate.
Owner reports are issue signals, not prevalence data. Long-term owner discussions can reveal recurring symptoms, awkward workflows and questions worth investigating. They do not provide the denominator needed to calculate how often a product fails, so Cafe & Crema does not turn forum posts or review counts into population failure rates.
Community sources are supporting context. Reddit and other owner communities can show what people are cross-shopping or troubleshooting. They sit below manufacturer documentation, manuals, independent testing and specialist repair evidence when those stronger sources are available.
First-party testing boundary
Cafe & Crema does not claim first-party testing unless a real test was performed. If the site has not physically tested a product, we do not write “we tested,” “during our testing,” “we measured,” or similar language. A research-based assessment can still compare documented hardware, workflow, serviceability and current market position, but it must not be presented as hands-on evidence.
Where first-party testing is ever performed, the article should say what was tested, how it was tested and what the test cannot establish.
Sources & Evidence on commercial pages
Commercial reviews and comparisons should expose a visible Sources & Evidence section. Each source should have a clear role: manufacturer specification, manual or service guidance, independent testing, repair/service evidence, or qualitative owner context.
If reliable sources conflict, Cafe & Crema either explains the conflict, uses the narrower claim supported by the stronger source, or removes the disputed exact claim. Precision is not useful when the underlying evidence is not precise.
Current models, generations and Amazon identity
Model generations matter. A current product can share a family name with an older machine while using different hardware, controls, accessories or service documentation. We try to name the exact model or generation and avoid applying a successor's specification to a legacy product.
If an exact Amazon product identity cannot be verified, we do not guess an ASIN. A recommendation can remain on the page without an Amazon button when the product is relevant but the exact marketplace identity is uncertain.
Affiliate destinations are monitored for link health, but that is not the same thing as retailer inventory verification. Retailer prices and stock can change, so we do not fabricate price, discount, urgency or availability claims.
How rankings and scores work
Rank is the editorial decision: which product fits the page's stated buying criteria best on the available evidence. The criteria can include workflow, compatibility, serviceability, current model support, independently measured performance where available, and value in the product's current market position.
Displayed scores are deliberately coarse. Cafe & Crema uses the format Cafe & Crema Score: X/10 and does not present Amazon ratings as Cafe & Crema scores. The site does not claim evidence precision it does not have by publishing invented tenths-of-a-point distinctions.
Corrections and updates
Products are revised, renamed and discontinued. Manuals change. Better evidence appears. When that happens, Cafe & Crema updates the existing page when it still serves the same search intent rather than creating a duplicate article.
A material correction should remove or narrow the unsupported claim, update the visible evidence, and advance the page's update date. If you find a wrong specification, stale product generation or unsupported statement, use the contact page.
Affiliate funding and editorial independence
No brand pays for placement and no brand gets to approve an article before publication. Cafe & Crema may earn an affiliate commission when a reader buys through an eligible link. The required disclosure and program details are on the disclosure page.
Affiliate eligibility does not change the evidence standard. A product should not move up a ranking because it has a higher commission, and a product should not receive a fabricated price, stock or urgency claim to encourage a click.