Building a Launch GEO Baseline

A specialty pharma company was preparing to launch a new therapy for a moderate-to-severe autoimmune condition. The competitive landscape included three established brands with significant market presence and marketing investment.

The brand team faced several concerns:

• Unknown AI positioning: With over 40 million daily health queries on ChatGPT alone^[2], patients and HCPs were already asking about treatment options. The team didn't know how (or if) AI systems mentioned their therapy, even pre-approval.
• Competitor advantage: Established competitors had years of published content, clinical data, and source footprint. Their therapies appeared prominently in AI answers; the new brand was largely invisible.
• Accuracy concerns: Pre-launch buzz and early clinical publications meant AI might already be making claims about the therapy - potentially inaccurate ones based on incomplete information.
• Launch readiness: The team needed a clear picture of AI visibility before launch to prioritize content investments and ensure approved messaging would be reflected in AI answers.

AI Pulse ran a baseline analysis with 120 question variations covering patient intent (treatment options, side effects, cost, lifestyle) and HCP intent (mechanism, efficacy, dosing, guidelines) across all four major LLMs.

The initial baseline revealed significant visibility gaps:

When the brand was mentioned, it typically appeared after two or three competitors, positioning it as an also-ran rather than a viable first-line option.

Beyond visibility, the Influence Graph revealed why competitors dominated:

• Competitors had 40+ high-authority sources (journal articles, medical education sites, treatment guidelines) being cited by AI.
• The new brand had only 8 citable sources, mostly early clinical trial publications and a corporate press release.
• AI framing used competitor language: "established treatment options include..." with the new therapy mentioned as "an emerging option" or "in late-stage trials."

PI-backed verification identified 14 claims requiring attention:

The Not-in-PI claims were especially concerning: AI was making statements about the therapy that couldn't be defended with approved labeling^{[3, 4]}.

The brand team used the governance queue to route findings to appropriate owners:

1. Medical Affairs (Not-in-PI claims): Immediate review of the 4 Not-in-PI claims. Medical wrote clarifying content for publication on medical education sites and coordinated with clinical teams to address the off-label dosing concern appearing in AI answers.
2. Brand Marketing (visibility gaps): Prioritized content investments to build source footprint. Published 6 new pieces of educational content on high-authority partner sites, refreshed the mechanism of action explainer, and worked with KOLs on accessible publications.
3. Omnichannel (positioning alignment): Reviewed all approved content to ensure consistent positioning language. Updated website copy to use preferred framing that AI could learn from^[6].
4. Communications (narrative prep): Developed talking points for inevitable questions about AI accuracy. Documented baseline findings to show proactive monitoring.

After 8 weeks of content publishing and source building, AI Pulse reran the baseline question sets. Results:

The brand entered launch with a documented AI visibility baseline, clear competitive positioning, and an ongoing monitoring program to catch drift early.

Beyond metrics, the engagement changed how the team operated:

• AI visibility became a KPI: Share-of-answer joined share-of-voice in brand performance dashboards. Monthly reports now include AI Pulse scores.
• Content strategy expanded: The content calendar now explicitly includes "source footprint" goals - not just impressions or engagement, but whether content is likely to be cited by AI systems.
• MLR workflow improved: Not-in-PI findings created a fast-track review process for corrective content. Audit trails made compliance documentation straightforward.
• Cross-functional alignment: Brand, Medical, Omnichannel, and Comms now have a shared view of AI positioning - reducing silos and enabling faster response to drift^[6].