Skip to content
Methodology

How we test. Seven steps to a recommendation.

AI-powered data analysis, professional source evaluation and our own product testing—combined into recommendations you can actually trust.

Wissenschaftliche Haaranalyse mit Pipette und Notizbuch
7
Process Steps
5
Source Types
1–10
Rating Scale

We don't rely on a single source. We combine AI-powered data analysis, professional test reports, thousands of real customer reviews and, where needed, our own product tests.

Our process
01

Market Analysis With AI

We map the entire market with AI. That means reading studies, real product reviews and social media posts. What comes out is a complete picture of the category.

02

Defining the Rating Criteria

Every product category gets its own rating criteria. Then we weight them by how much they actually matter. Usually that means value for money, customer feedback and category-specific factors.

03

Research and Source Review

We work through every relevant product review, test report and comparison. The sources: published data from regulators and standards bodies, plus test reports from independent testing organizations. Real customer feedback and social media posts round it out.

04

AI-Assisted Product Analysis

For each product, the AI reads every customer rating—on Amazon and other platforms. Then come social media comments, ingredient lists and technical specs. The result is a score from 1 to 10.

05

Our Own Hands-On Tests

When the data doesn't line up, we buy the product with our own money. Then we test it ourselves. That's how we find out whether the reviews and the research hold up in real life.

06

Structured Test Reports

Every product becomes a structured test report. Pros and cons, all of them—so you get the full picture.

07

Building the Comparison

Every score feeds into one big comparison article, where we rank the products at a glance. You find the right one—without spending hundreds of hours on research yourself.

Why this approach works

Most comparison sites lean on a single source of information. We combine four: AI-powered data analysis, professional test reports, real customer feedback and our own testing.

The result: you find the right product without spending hundreds of hours on research yourself.

Frequently asked questions

We work in seven steps. First we analyze the market with AI. Then we set the rating criteria. Step three is the source review: peer-reviewed research, published data from regulators and standards bodies, independent test reports, customer reviews and social media. After that, the AI scores every product on a scale of 1 to 10. When the data is unclear, we test the product ourselves. Finally come the test reports—and out of those, the comparison articles.
Yes. The AI maps the market and works through thousands of customer reviews. It checks ingredient lists and scores products on the data. Real product tests and an editorial review round that out.
Yes. Sometimes reviews, studies and test reports contradict each other. When they do, we buy the product with our own money and test it ourselves—that's how we verify what the research turned up.
We use five types of sources. First, scientific studies via PubMed and CIR. Second, professional test reports from independent testing organizations, regulators and standards bodies. Third, platform ratings on Amazon and other major retailers. Fourth, social media feedback from Instagram, YouTube and Reddit. And fifth, our own product tests.
Every product gets a score from 1 to 10. It rests on rating factors matched to the category at hand. Usually that means efficacy, ingredient quality, value for money and customer feedback.
Regularly. New product versions, fresh study results or a shift in the market all trigger an update. Every correction carries a visible date.
No. The rankings come out of the 7-step research process alone. Affiliate revenue changes nothing—not the placements, not the recommendations.