Home

/

Case Study

From ~8% to ~28%: How Mando turbocharged its AI search Visibility

AT A GLANCE

Mando was invisible amongst men’s deodorants on AI search. In four months it became the source 2nd most cited by AI, driving visibility from ~8% to 28%. Bodhium wrote 5 of the top 6 cited Mando articles.

The challenge

Mando was nearly absent from AI search in its highest-intent category: comparison questions like "best deodorant for sensitive skin."

AI answers are winner-take-most. The brands an assistant names are the shortlist, and there is no page two.

In March, Mando appeared in under 8% of those answers and close to zero on ChatGPT. Its owned pages were built to sell, not to answer the question shoppers were asking.


The approach

Bodhium ran a precision Answer Engine Optimization program built for how AI assistants actually source and cite recommendations. Mando makes whole-body deodorant for men (72-hour, aluminum-free, mandelic-acid based), and the work paired that product truth with content engineered for AI consumption, not for a human skimming a results page.

  • Map the questions, then own them. We identified the highest-intent questions real shoppers ask in Mando's category and built content to answer them directly: 10 comparison and head-term articles on March 27, then 11 whole-body use-case articles on May 27 to 29. 21 live articles in total.


  • Engineer each page to be cited. Five choices did the heavy lifting: match each article to the sources a given engine already trusts; open with a self-contained, quotable answer; repeat one named causal mechanism (mandelic acid lowers skin pH so bacteria cannot turn sweat into odor); back claims with specific, compliant clinical detail (72-hour control, tested by Princeton Consumer Research; target pH 4.7 to 5.75); and own the category vocabulary and the under-served zones competitors skip.

The challenge

Mando was nearly absent from AI search in its highest-intent category: comparison questions like "best deodorant for sensitive skin."

AI answers are winner-take-most. The brands an assistant names are the shortlist, and there is no page two.

In March, Mando appeared in under 8% of those answers and close to zero on ChatGPT. Its owned pages were built to sell, not to answer the question shoppers were asking.


The approach

Bodhium ran a precision Answer Engine Optimization program built for how AI assistants actually source and cite recommendations. Mando makes whole-body deodorant for men (72-hour, aluminum-free, mandelic-acid based), and the work paired that product truth with content engineered for AI consumption, not for a human skimming a results page.

  • Map the questions, then own them. We identified the highest-intent questions real shoppers ask in Mando's category and built content to answer them directly: 10 comparison and head-term articles on March 27, then 11 whole-body use-case articles on May 27 to 29. 21 live articles in total.

  • Engineer each page to be cited. Five choices did the heavy lifting: match each article to the sources a given engine already trusts; open with a self-contained, quotable answer; repeat one named causal mechanism (mandelic acid lowers skin pH so bacteria cannot turn sweat into odor); back claims with specific, compliant clinical detail (72-hour control, tested by Princeton Consumer Research; target pH 4.7 to 5.75); and own the category vocabulary and the under-served zones competitors skip.

Map the questions, then own them. We identified the highest-intent questions real shoppers ask in Mando's category and built content to answer them directly: 10 comparison and head-term articles on March 27, then 11 whole-body use-case articles on May 27 to 29. 21 live articles in total.

● Engineer each page to be cited. Five choices did the heavy lifting: match each article to the sources a given engine already trusts; open with a self-contained, quotable answer; repeat one named causal mechanism (mandelic acid lowers skin pH so bacteria cannot turn sweat into odor); back claims with specific, compliant clinical detail (72-hour control, tested by Princeton Consumer Research; target pH 4.7 to 5.75); and own the category vocabulary and the under-served zones competitors skip.

The challenge

Mando was nearly absent from AI search in its highest-intent category: comparison questions like "best deodorant for sensitive skin."

AI answers are winner-take-most. The brands an assistant names are the shortlist, and there is no page two.

In March, Mando appeared in under 8% of those answers and close to zero on ChatGPT. Its owned pages were built to sell, not to answer the question shoppers were asking.


The approach

Bodhium ran a precision Answer Engine Optimization program built for how AI assistants actually source and cite recommendations. Mando makes whole-body deodorant for men (72-hour, aluminum-free, mandelic-acid based), and the work paired that product truth with content engineered for AI consumption, not for a human skimming a results page.

  • Map the questions, then own them. We identified the highest-intent questions real shoppers ask in Mando's category and built content to answer them directly: 10 comparison and head-term articles on March 27, then 11 whole-body use-case articles on May 27 to 29. 21 live articles in total.

  • Engineer each page to be cited. Five choices did the heavy lifting: match each article to the sources a given engine already trusts; open with a self-contained, quotable answer; repeat one named causal mechanism (mandelic acid lowers skin pH so bacteria cannot turn sweat into odor); back claims with specific, compliant clinical detail (72-hour control, tested by Princeton Consumer Research; target pH 4.7 to 5.75); and own the category vocabulary and the under-served zones competitors skip.

Map the questions, then own them. We identified the highest-intent questions real shoppers ask in Mando's category and built content to answer them directly: 10 comparison and head-term articles on March 27, then 11 whole-body use-case articles on May 27 to 29. 21 live articles in total.

● Engineer each page to be cited. Five choices did the heavy lifting: match each article to the sources a given engine already trusts; open with a self-contained, quotable answer; repeat one named causal mechanism (mandelic acid lowers skin pH so bacteria cannot turn sweat into odor); back claims with specific, compliant clinical detail (72-hour control, tested by Princeton Consumer Research; target pH 4.7 to 5.75); and own the category vocabulary and the under-served zones competitors skip.

“The shopper asks one question and acts on the answer. We made Mando the answer, not an afterthought.”

Jim Hanrahan

Senior Brand Director, Mando (Mammoth Brands)

The results

We tracked 2 clear wins: Search Visibility and ChatGPT User-Bot Traffic.

Non-branded AI visibility more than tripled in four months, from 7.8% to 28.1%.

Mando's own site became a top source, not just a mention: shopmando.com is now the #2 most-cited domain on non-branded questions across the category, ahead of Reddit, YouTube - only trailing the incumbent Dove.

Gains were broad, not one lucky channel. Gemini rose 10.8% to 44.4%, Perplexity 11.6% to 30.5%, and Google AI Overview reached 33.7%.

"The team at Bodhium Labs helped us precisely identify which queries to track and care about. Beyond just measurement, they strategically helped us identify key gaps and opportunities. Their actionable and tested content strategy has helped us eliminate guesswork and control our AI visibility and brand narrative."

[Name], [Title] · Mando

All case studies