Send a case for careful review

Levertrace Lab studies which optimisation changes appear to alter LLM descriptions of French businesses. The team compares before-and-after answers, public source changes, language versions and discrepancies. Its work is research and interpretation, not a guarantee of placement.

The lab does not promise rankings, force citation, manage review profiles or sell instant AI visibility fixes. It also avoids cases where there is too little public evidence to compare. A useful case needs a visible source trail.

The notes should be read as careful case evidence, not universal law. A lever that produced adoption in one setting may produce partial echo or no visible movement in another. The value is in the pattern, the source context and the limits stated beside the finding.

Yes. The lab accepts suggestions when the case involves French business information, public sources and a clear optimisation change or discrepancy. The team may use a suggestion as a private reply, a research lead or a future anonymised note.

It answers selected requests when the question fits its method. Strong requests describe the before state, the after change, the model wording and the language condition. Thin requests that only ask “why am I not visible?” are usually too broad to assess.

Replies are not instant because the team reads the source trail before answering. Simple messages may receive a brief response; cases requiring comparison take longer and may not be accepted for review. The lab prefers a slower answer over a tidy guess.

The default site language is English, and French materials may appear where the source context requires it. The lab studies both English and French answers because language transfer is often part of the problem.