Croda dominates AI beauty ingredient discovery as suppliers face GEO gaps

The Pitman Comms Visibility Index Report has named Croda as the winner in the beauty industry’s behind-the-scenes Generative Engine Optimization (GEO) search results. The B2B consultancy analysis of ChatGPT’s beauty sourcing recommendations highlights wide disparities across AI visibility. 

The UK-focused test posed ChatGPT with six unbranded supplier-discovery questions, focused on buyers’ cosmetic ingredient search specifications across five ingredient categories. Croda stood out with a clear lead as the only company to achieve consistent recognition across all five categories. 

Each of the five ingredient questions was run three times, producing 15 shortlists and 75 recommendation places in total. Croda appeared across all 15, while the closest runners-up — Lubrizol and Provital — sat at four. The categories listed are: actives, functional, botanical, sensory, and biotech. 

Distributors also emerged as significantly more visible in some AI-driven ingredient sourcing, potentially overshadowing the manufacturers whose products they sell. In the functional ingredients category, distributors occupied 80% of recommendation places, exemplifying their visibility over manufacturers. 

“For manufacturers, the practical check is whether partner listings clearly identify who makes the ingredient,” says Simon Pitman, founder at Pitman Comms. 

Croda leads, competitors fill niches

No other company listed in the report came close to Croda’s breadth of visibility across categories. 

Other companies’ performances were more specialized: Croda, Givaudan Active Beauty, and Provital appeared in every run of the actives category. In the botanicals category, Naturallythinking and O&3 appeared in every search, joining Croda.

In sensory searches, Elementis, Lubrizol, and Seppic appeared every time together with Croda. Meanwhile, Vytrus Biotech was the only name other than Croda in the biotechnology category. 

In functional ingredients distributors, Azelis, Brenntag, and Harke Chemlink appeared every time, leaving Croda as the only manufacturer in the category. 

Croda appeared in all 15 ingredient shortlists tested by Pitman Comms. 

Understanding the divide

Croda’s results demonstrate unusually broad recognition across the five ingredient needs tested. However, the report does not establish why ChatGPT selected Croda so consistently. 

Pitman Comms notes that broad and specialist portfolios require different visibility targets, cautioning that Croda’s breadth should not be treated as a universal benchmark for more specialized suppliers. 

According to the report, specialists like Vyrtus Biotech have carved out a niche and should not be expected to perform the same way that a broader portfolio company like Croda does. 

Additionally, Croda showed consistent response. The report highlights the key impact of repeated discovery over occasional favorability. 

For example, Lubrizol and Provital had four appearances each. Lubrizol received three appearances in sensory ingredients and one in active ingredients. Meanwhile, Provital hit three in actives and one in biotechnology. 

Pitman Comms recommends suppliers check if AI search recognizes their priority products and launches when buyers search for specific needs. If recognition is missing or inconsistent, the consultancy suggests confirming the question fits the offer and investigating the gap. 

“These observations show which suppliers the test responses returned, not observed buyer behavior, supplier quality, or overall performance. Recognition can vary by search and capability; distributor visibility may not identify the manufacturer. Media examples show ways to explain expertise without establishing why AI selected a supplier,” says the report. 

AI sourcing results revealed significant visibility gaps across beauty ingredient suppliers. 

Distributor v manufacturer, a problem of visibility 

Croda also stood apart in the functional ingredient search as the only manufacturer named alongside distributors. Across the three responses, the specialty ingredients company appeared every time, while distributors occupied the remaining 12 of 15 recommendation places. 

The finding raises a question about whether AI sourcing results attribute ingredients to manufacturers or to suppliers who may be more visible in association with the ingredient. The company being the only manufacturer shortlisted in the functional ingredients category may indicate that ChatGPT associates Croda, not one of its distribution partners, with the ingredient requirement.

The report underscores that distributor sites may be becoming an important avenue for GEO discovery. As a solution, it states that manufacturers can ensure that their partner pages identify the manufacturer, application, expertise, and a way to access samples or technical support. 

The study examined a total of 26 companies, 11 of them did not appear in any of the six sourcing searches despite the report recording relevant category links for them. 

The 11 included AAK Personal Care, Active Concepts, Cobiosa, Groupe Berkem, IMCD Beauty & Personal Care, INOLEX, Kobo Products, Lipoid Kosmetik, SILAB, SOPHIM, and Symrise Cosmetic Ingredients. 

The discrepancy between the listed companies’ established industry recognition and GEO performance highlights whether online presence portrays ingredient expertise in a way that generative search can recognize. 

“An AI answer can introduce a supplier before the buyer visits its website,” says the report. 

Pitman Comms suggests investigating whether product information, applications, supporting evidence, media coverage, distributor listings, and wider search presence consistently connect a supplier with its priority areas of expertise.