Every product on Pickion goes through the same process: an AI-assisted screening model built by Pickion reads the public signals around a product — buyer reviews, complaints, refund mentions, offer terms and marketplace data — and a human reviewer checks every result before anything is published.
Last updated: October 13, 2026 · Version 1.0 (experimental)
We score every product on five pillars — buyer sentiment, complaint & refund signals, offer transparency, product substance and market traction — using publicly available data. The AI does the heavy reading; a human makes the call. Commission size is never part of the score.
We don’t review everything. Each candidate product passes through a four-stage screening funnel before it earns a full review:
Products that fail the gates or score below 2.5 in the preliminary scan are not featured. If a product we already cover drops below that line, we update the review to say so.
Each pillar is scored from 0 to 5 using the KPIs below. All sources are public, so anyone can check our work.
What real buyers say after purchase — not the testimonials on the sales page.
Sources: Trustpilot, Reddit, YouTube comments, independent forums and review sites.
How often buyers run into problems, and how hard it is to get money back.
Sources: BBB (Better Business Bureau), Trustpilot, Reddit, consumer forums.
Whether you know exactly what you’re paying for before you pay.
Sources: the official sales page and checkout, ClickBank product terms.
Whether the product delivers what it promises.
Sources: official product information, buyer reports, product previews.
Whether the product has a real, sustained track record — a supporting signal, never the main one.
Sources: ClickBank marketplace, Google Trends, Wayback Machine.
Gravity reflects how many different affiliates recently earned a sale from a product. It shows demand, not quality — so we cap its influence and slightly lower the score for extremely high values, which often signal aggressive promotion rather than happy customers.
| Gravity | Traction sub-score | How we read it |
|---|---|---|
| Below 5 | 1.5 | Little recent sales activity or a brand-new product |
| 5 – 20 | 3.0 | Niche but active |
| 20 – 60 | 4.0 | Established demand |
| 60 – 150 | 4.5 | Strong, proven demand |
| Above 150 | 4.0 | Very high promotion pressure — checked extra carefully |
The result is rounded to one decimal and then checked against the red flags below. The final verdict follows these bands:
| Score | Verdict | What it means for you |
|---|---|---|
| 4.5 – 5.0 | Excellent | Strong signals on every pillar. Easy to recommend for its target buyer. |
| 4.0 – 4.4 | Recommended | Solid choice for the right person, with known and manageable downsides. |
| 3.0 – 3.9 | Situational | Worth it only in specific cases. Read the “who should skip it” section. |
| Below 3.0 | Not recommended | Too many warning signs. We say so clearly and suggest alternatives. |
A score is only as good as the data behind it, so every review also shows how much evidence we had:
Some problems are serious enough that a high average shouldn’t hide them. When we find one, the score is capped:
A human reviewer reads a sample of the source material behind every pillar, corrects misclassifications and can adjust a pillar score by up to ±0.5 with a written reason.
The Pickion Signal Score is an original methodology developed by Pickion. It is not a scientific study or a certification, and it is not affiliated with or endorsed by ClickBank or any other platform mentioned. We refine it as we learn and publish every change below.
See the method in action in our latest review.
Read the Ted’s Woodworking review →