Five Stages of Pharmaceutical Competitive Intelligence: A Diagnostic Framework from Data to Strategic Decision-Making

From Data Deluge to Decision Dilemma

In the first quarter of 2026, licensing commitments in the obesity and diabetes field have reached $22 billion, exceeding the total for the entire year of 2025. In the same period, Pfizer acquired Metsera for $10 billion after a bidding war, and Roche paid an upfront $1.65 billion to co-develop petrelintide with Zealand Pharma. Behind these deals lies a competitive landscape with over 100 active development programs—in such a field, the gap between identifying an asset early versus one step too late can amount to billions of dollars.

The competitive intelligence (CI) teams supporting these decisions are processing more data than ever. But the key question is: Does this data reach decision-makers in a usable form? In many pharmaceutical companies, the CI workflow has gaps across five stages, preventing intelligence from effectively influencing strategy.

Stage 1: Discovery — Traceability Becomes the New Coverage Problem

Traditionally, the primary search challenge for CI was coverage: not missing key signals such as competitor trial registrations, patent filings, or regulatory submissions. This concern has not disappeared, but the dominant failure mode has shifted. AI-assisted search tools can now present results quickly and in large volumes, but they cannot automatically provide a traceable path from each result back to the primary source—such as registry entries, patent documents, peer-reviewed articles, or regulatory submissions—that can withstand scrutiny in portfolio reviews.

Take a CI report on obesity assets: it needs to integrate ClinicalTrials.gov, multi-jurisdictional patent databases, peer-reviewed pharmacology literature, FDA/EMA/PMDA regulatory filings, and earnings and conference data. Each source uses different data structures and reporting conventions. When CI teams use generic AI summaries to streamline integration, they may produce outputs that appear complete but cannot withstand a business development director's follow-up questioning about specific sources.

Diagnostic question: Can every claim in your current competitive landscape report be traced back to a named primary source?

What high-performing teams do: They treat source traceability as a non-negotiable output standard, not a post-hoc documentation task. Before delivering a landscape report, the team can point to the specific registration, patent, or journal record behind each claim. This is not conservatism—it's maintaining credibility when questions arise.

Stage 2: Analysis — Pattern Detection ≠ Interpretation

Automated tools excel at pattern detection: flagging that six companies have filed for the same target indication, identifying mechanism-of-action clusters, or tracking trial status changes within a competitive set. At the scale of over 100,000 drugs in development and 6.9 million patents, identifying these patterns is indeed a challenging task, and automation provides help.

But interpreting patterns is another matter. Does a cluster in a certain indication represent true target validation, or does it signal crowding risk that should alter the investment thesis for a specific asset? This requires applying contextual judgment to the specific question.The GLP-1 field directly illustrates this distinction. With over 100 active obesity compounds in development, raw numbers are almost meaningless to CI managers. What matters is how these assets differentiate: which have unique mechanisms, which target the same efficacy endpoints as semaglutide and tirzepatide, and which aim to serve those underserved patients.

For example, Roche's CT-388 achieves differentiation through near-equimolar GLP-1/GIP receptor binding (approximately 1:1, compared to tirzepatide's 9:1), an independent peptide backbone patent, and Phase 2 data showing a placebo-adjusted weight loss of 22.5% at 48 weeks without entering a plateau—a profile that simultaneously shaped Roche's BD strategy and Pfizer's decision to pursue a different mechanism via its acquisition of Metsera. Mapping such differentiation requires deep coverage across trials, patents, and clinical data, not just a summary of news headlines.

Diagnostic question: When competitive patterns are identified, does your team have a defined process to reach interpretation, or is this step handled informally?

What high-performing teams do: They assign a named decision question to each analysis before work begins. "Who are the GLP-1 competitors?" yields a different output than "Which GLP-1 assets pose a realistic differentiation threat to our Phase 2 program before 2027?" The second question scopes the analysis, forcing interpretation rather than just enumeration.

Phase 3: Generating Insights—The Missing Perspective

The gap between analysis and insight is perspective. Analysis describes what is happening in the competitive space; insight explains what it means for a specific decision, along with a recommendation.

Most CI deliverables stop at description. Competitors are accurately tracked, trial progress is recorded, data is correct—but the "so what" is left to the reader.

CI teams that consistently influence portfolio and BD decisions go further: they interpret meaning. They don't say "Here are five late-stage entrants in ALK+ NSCLC," but rather "How the Phase 3 timelines of two of them change the risk profile of our X asset licensing decision, and what actions we should take if their data read out before ours."

Achieving this requires setting a decision question at the outset. "Oncology competitive landscape" is a topic; "Should we accelerate Phase 2 enrollment before competitor A's data readout?" is a decision question. These two starting points will produce deliverables that are used in very different ways in portfolio committee meetings.

Diagnostic question: Do your CI deliverables always include clear recommendations, or do they present data and leave interpretation to the reader?What high-performing teams do: They align with stakeholders on the decision question before—not after—building the deliverable. In practice, this means the CI manager has a 20-minute conversation with the BD or portfolio lead before starting to write the landscape report, asking: "What specific decision will this analysis support? When does it need to be made?" This conversation shapes everything that follows, from which data is prioritized to how the output is structured.

Stage 4: Communication – Format Determines Whether CI Gets Used

Even if the CI team excels in the first three stages, their output may still fail to influence decisions. The most common reason is format: a comprehensive briefing document organized by topic rather than built around a specific question.

BD and portfolio leads read reports organized by decision question with clear action recommendations, not data manuals organized by data category. This is not to say format is more important than content; rather, if decision-makers cannot quickly find the part relevant to their decision, the content never gets a chance to be applied.

The best practice is: every CI report opens with a "Decision Summary"—not an executive summary, but one that explicitly states, "We recommend Option A over Option B for the following reasons, and this recommendation is based on these four key data points." Then, the detailed analysis is grouped by decision question, not by data source.

Stage 5: Scenario Modeling – From Reporting to Strategic Consulting

The highest-maturity CI teams answer not only "What is the current situation?" but also "How will the competitive landscape change under different conditions?" Scenario modeling transforms CI from a descriptive function into a strategic consulting function.

In the obesity space, a CI team might need to build multiple scenarios: What happens if Novo Nordisk's oral amycretin shows a safety signal? What if Roche's CT-388 proves superior to tirzepatide in Phase 3 but is less tolerable? These are not hypothetical questions—they are scenarios the BD team uses to evaluate deals, set milestones, and decide internal investment priorities.

Scenario modeling requires three inputs: a deep understanding of the current competitive landscape, an analysis of scientific and regulatory dynamics that could change that landscape, and a framework—a method for turning dynamics into concrete scenarios. Most CI teams lack the third input, so their reports describe the present but cannot support "what-if" decisions.

Diagnostic question: To what extent does your CI function provide "what-if" analysis, or does it stop at "as-is" reporting?

What high-performing teams do: They build a small number of scenarios (typically 3–4) for each key decision, clearly labeling the probability and trigger factors for each scenario. Then they provide recommendations tailored to each scenario, enabling decision-makers to prepare in advance.

What AI Changes, and What It Cannot Change## What AI Changes and What It Cannot Change

A common misconception about AI in CI is that it will automate the entire workflow from discovery to recommendation. In reality, AI is highly effective in Phase 1 (by accelerating search and integration) and Phase 2 (by enhancing pattern recognition). However, in Phases 3 and 5, AI serves as a support tool—providing data-driven events and correlations—but the final perspective and scenario construction require human judgment and strategic context.

In Phase 4 (Communication), AI can help generate initial drafts and visualizations, but the structure that decision-makers need—organized by decision questions—still requires human understanding of the audience’s priorities.

In clinical decision-making, the positive impact of human input on system performance is well documented; in competitive intelligence, a similar logic holds: AI augments human analysts, not replaces them.

Diagnostic Self-Questioning

For pharmaceutical CI teams, the question is not “Do we have enough data?” but rather "Does the data reach the right person in the right form at the right time?" Most CI teams have bottlenecks in one or more stages of the five-phase workflow. Identifying these bottlenecks and applying targeted improvements—whether in traceability standards, front-loading decision questions, or scenario modeling capabilities—can significantly enhance CI’s actual impact on strategic decisions.

Every CI team can ask itself a series of diagnostic questions today. The answers will reveal where the next improvement should be:

• Is every assertion traceable to a primary source?
• Does our analysis start with enumeration or with decision questions?
• Are our deliverables accompanied by clear recommendations?
• Does the form and structure help decision-makers quickly find the relevant sections?
• Do we provide "what-if" scenarios to support forward-looking decisions?

In an era where competition in the pharmaceutical industry is escalating at the scale of billions of dollars, the ultimate measure of competitive intelligence is not the number of reports, but its measurable impact on every transaction and every portfolio decision.

---

*This article is based on the document "Pharmaceutical Competitive Intelligence Workflow Problems: A Diagnostic Framework" published by Clarivate, with data and analysis derived from that original source.*