In focus
Prompt sets for service businesses with mixed French and English evidence need a split view: presence, citation, position and description quality first; rewrite recommendations only after the answer pattern is visible.
AI answer measurement
AI visibility is measured by repeated answers, not impressions.
A single screenshot tells you almost nothing. I build prompt ledgers that show how often a business is named, where it sits in the answer, which sources are cited, and whether the description is accurate in French and English. The work is made for French SMBs, agencies, service firms and local operators who need a monthly view they can defend in a meeting before treating AI answers like keyword ranks.
Latest notes — from the field
who writes this
Lucien Varlet
I come from western France, and for 17 years I have worked around search pages and the evidence behind them. My work now is to build repeatable tests: prompt ledgers with date, engine, language, location intent, cited source, answer position and description error in separate columns. French and English prompts are tested separately, because the evidence path often changes by language. An AI answer is useful evidence only when it appears often enough, names the right source often enough, and describes the business accurately enough to guide action.
Measure the answer pattern before changing the page.
Bring a category, a location, a few competitors and the questions buyers actually ask.
Start measurement