The U.S. pharmaceutical industry spent an estimated $8 billion on direct-to-consumer advertising in 2024, making it the largest DTC ad category in the country. New research from 5W AI Communications shows that AI engines largely ignore that spending. When patients ask about drugs, the tools name Eli Lilly and Novo Nordisk - not the companies running the most prime-time spots.
The finding comes from the Pharma / Rx AI Visibility Index 2026, released June 19. 5W ran more than 60 patient and consumer prompts across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews in clean sessions during Q2 2026. The result: a ranking of the top 25 pharmaceutical companies by AI citation share.
The citation leaders and the revenue disconnect
Eli Lilly captured an estimated 12.5% of all pharma citations in the test set. Novo Nordisk followed at 11.5%. Together they hold roughly a quarter of the AI answer surface for the industry. The combined 2025 sales of their flagship GLP-1 drugs - Ozempic, Wegovy, Mounjaro, and Zepbound - reached approximately $70 billion.
Merck, whose Keytruda is the world's best-selling prescription drug, landed fifth at 6.0%. AbbVie, historically the highest DTC spender behind the Humira franchise, placed sixth. "Revenue rank does not predict citation rank," the report states. "The highest-earning drugmakers are not the most-cited."
The full top fifteen by estimated AI citation share:
- Eli Lilly - 12.5%
- Novo Nordisk - 11.5%
- Pfizer - 8.5%
- Johnson & Johnson - 7.0%
- Merck - 6.0%
- AbbVie - 5.0%
- AstraZeneca - 4.0%
- Moderna - 3.5%
- Amgen - 3.0%
- Novartis - 2.6%
- Bristol Myers Squibb - 2.3%
- GSK - 2.0%
- Sanofi - 1.8%
- Gilead Sciences - 1.5%
- Roche - 1.3%
Why ad dollars don't buy the answer surface
The mechanism is the difference between paid attention and what the report calls "AI-retrievable record." A 30-second television spot can shift share-of-voice for a quarter. But a peer-reviewed trial in The New England Journal of Medicine, followed by sustained coverage across Reuters, STAT, and Fierce Pharma, moves the answer surface for years. The engines retrieve the durable record, not the campaign.
The conditions patients self-research drive the citations. Weight, diabetes, and immunology generate traffic. Specialist-prescribed oncology does not, which explains why Merck's blockbuster cancer drug does not lift the company's citation rank to match its revenue. The report also notes that diffuse corporate structures and private company status dilute citation share, while a single blockbuster drug pulls the parent company name with it.
Ronn Torossian, Founder and Chairman of 5W AI Communications, said: "Pharma is the largest direct-to-consumer ad category in America. The AI engines are not impressed. They name the companies whose drugs patients actually research - Eli Lilly and Novo Nordisk - not the companies running the most prime-time spots. AbbVie spent the most. Merck makes the best-selling drug on earth. Neither is the answer the chatbox returns. That is the gap."
He added: "The pharmaceutical industry is paying for share-of-voice on a channel - television - that no longer routes to the buyer the same way. The buyer is in the chatbox, asking ChatGPT to compare Ozempic and Mounjaro. The companies that built the AI-retrievable record around those names own that answer. The ones still buying spots and hoping it carries through to the engines are funding the wrong channel."
What the data documents for communications strategy
The Index surfaces six structural truths. Self-researched conditions drive citations. DTC advertising compounds into citations only when it feeds the editorial record - the spot itself is not retrievable, but the coverage it generates is. Patients now research before the appointment, which means corporate citation share sits upstream of the prescription conversation. And diffuse corporate structures dilute the signal.
The methodology matters for communicators evaluating the findings. 5W ran each prompt five times per engine in clean sessions, tracking citations across business and health editorial press, reference sources like Wikipedia and Drugs.com, and brand-owned corporate content. The full methodology is available at 5wpr.com.
The report carries an important framing: it measures AI citation share for corporate communications and marketing strategy only. It is not medical advice and does not rank drugs or companies on safety, efficacy, or value.
Why this matters for PR and communications professionals
For communicators, the Index reframes the DTC spend question entirely. The channel that built pharmaceutical brand awareness for two decades - linear television - generates no retrievable record inside the tools where patients now start their research. The report suggests that AI for PR & Communications is no longer a future capability; it is the current answer surface that determines whether a brand appears when a patient asks a comparative question. The companies that invested in sustained editorial coverage, Wikipedia presence, and trial coverage own that surface. The ones funding campaigns without feeding the retrievable record are, as Torossian put it, funding the wrong channel. For PR teams, the implication is direct: earned media coverage that lands in the databases the engines crawl compounds in value long after the ad buy expires.
Your membership also unlocks:
