Your product on SleekNova? 99% don't make it. Try anyway

August 22, 2026

What We Reject Stays Visible to the Next Search

A product we rejected in December was researched again from scratch in January, because the input our discovery pass reads listed only what was already live.

A product we have already judged should never be judged again by accident. That rule did not exist here until a soap we rejected in December came back through the front door in January and passed.

On January 27, 2026, while building a new feature for our Context Exporter plugin, we noticed the same product sitting in our Gap Analysis output twice: once rejected, once approved. Binu Binu Shaman Charcoal Soap, two separate Lab Notes, two different verdicts, five weeks apart.

Both verdicts were right for the day they were made. The gap was that nothing carried the first one forward.

The Two Lab Notes

December 30, 2025: Fifty Stockists and Zero Reviews. We found 53 premium retailers carrying the soap. Nordstrom, Dover Street Market, SSENSE, the Hammer Museum gift shop. We found three customer reviews. Confidence Score: 28/100. Below the 35 minimum. Rejected.

January 22, 2026: What Korean Shamans Knew About Charcoal. Same product. But now with verified customer reviews on Thingtesting, editorial validation from Vogue, NYT, and Architectural Digest, and a 10-year track record. Trust Score: 7/10. Final Score: 8.2. Passed.

The product is now live: Binu Binu Shaman Charcoal Soap.

What Happened

The December rejection was correct at the time. Three reviews across 53 stockists meant we could not verify customer satisfaction. Our system requires evidence, not assumptions.

Between December and January, conditions changed. Reviews accumulated. The product crossed our trust threshold.

In January, our Scout agent found the soap again during a bath and wellness hunt. Scout checked our live database for duplicates. No match. The product had been rejected, not approved. There was no live product page to flag.

Scout proceeded with research. Researcher verified the sources. Judge scored it. The soap passed.

The gap: our agents could not see rejected Lab Notes. The Context Exporter showed live products and taxonomy gaps. It showed nothing about what we had already evaluated and rejected.

Scout re-researched a product we had already documented. The second evaluation passed because circumstances changed. But we burned research hours on something we had already analyzed.

What We Built

In January 2026 the Context Exporter gained a “Previously Evaluated Products” section, and every Gap Analysis export has carried it since. When the discovery pass receives the export, it sees:

  • All rejected products with product name, brand, category, date, reason, and Lab Note URL
  • All passed products with Lab Notes
  • Any conflicts between rejected Lab Notes and live products

The discovery pass now checks this table before researching any candidate. If a brand appears in the rejected list, it is flagged immediately. No wasted research. No duplicate evaluations without explicit intent.

What the Duplicate Actually Cost

In this specific case, the re-evaluation was justified. Circumstances changed. Reviews accumulated. The trust score improved from 2.8 to 7. The product now passes our full verification.

The problem was not the outcome. The problem was the process.

If Scout had known about the December rejection, it could have noted “previously rejected, check if conditions changed” and fast-tracked the re-evaluation. Or flagged it for human review before investing research hours.

Now it can.

What This Means for Readers

Nothing changes about our recommendations. Every product in our database still passes the same verification. Binu Binu Shaman Charcoal Soap earned its place through verified customer reviews, editorial validation, and a decade of consistent quality.

What changes is the standard we hold ourselves to. A rejection is a judgment, and a judgment we cannot see is one we are free to contradict without noticing. When we reject something now, that decision persists into the input the next search reads, so a re-evaluation is a decision we take on purpose rather than one we stumble into.

The Receipts

This rule change was approved byMark van OosterwijckFounder and editor, SleekNova Labs