Secondary research is the foundation that makes everything else work.
That's not a metaphor — it's operational reality. The quality of your secondary research directly determines the success of your primary collection, the depth of your analysis, and ultimately the value of the intelligence you deliver. Skip it, rush it, or treat it as a box to check, and the rest of your CI process will suffer.
Gathering information is like building a house. If you try to start with the roof, the house will fail. A good house is built on a solid foundation, from the ground up. Secondary research is that foundation. It gives you the knowledge base for the project, helps you identify primary sources, identifies safe topics to discuss with those sources, and teaches you the therapeutic area terminology you need to have credible conversations.
The danger is that secondary research is both indispensable and deceptive. It's easy to lose time searching with little additional benefit. There's a lot of duplicate data and a lot of data that's not relevant. The data needs to be validated — it may be outdated, it may be wrong, and most importantly, it is data, not analysis. The goal is not to gather copious amounts of information but to produce highly relevant analysis that translates directly into improved business results.
That distinction — between gathering and analyzing — is where most secondary research efforts go wrong. They stop at collection. This article covers the complete process: how to plan your research, where to look, how to evaluate what you find, and how to set up ongoing monitoring that keeps your intelligence current.
There are four specific reasons secondary research must precede primary collection, and understanding them changes how you approach the work.
It builds your knowledge base. Before you can have an intelligent conversation with a KOL, a sales rep, or a competitor's scientist, you need to know the landscape. What trials are running? What data has been published? What regulatory milestones are upcoming? Without this baseline, you walk into conversations unprepared, and people can tell.
It identifies primary sources. Secondary research surfaces the names of investigators, presenters, authors, and industry experts you'll want to talk to. ClinicalTrials.gov tells you who the principal investigators are. Conference programs tell you who's presenting. Publications tell you who's publishing in the area. Job postings tell you who the competitor is hiring. All of this feeds your source identification and collection planning.
It identifies safe topics. Knowing what someone has already said publicly — in a presentation, a publication, on their LinkedIn profile — gives you legitimate entry points for conversation. You're not asking them to reveal secrets. You're asking them to expand on work they've already made public. That changes the dynamic entirely.
It teaches you the language. Every therapeutic area has its own vocabulary, its own shorthand, its own clinical endpoints and regulatory nuances. If you're doing CI in oncology, you need to understand ORR, PFS, and OS before your first conversation. If you're in rare diseases, you need to understand natural history studies and regulatory designations. Secondary research builds this fluency.
The higher quality of your secondary research, the more likely you will achieve success gathering human intelligence. That's not an exaggeration — it's a pattern I've seen play out across hundreds of projects. The analyst who shows up to a conference having done thorough secondary research gets more from every conversation than the one who's winging it.
The process has four stages. Each one builds on the last, and skipping any stage creates problems downstream.
Before you search for anything, clarify what you need to find.
Review your KIQs. Your Key Intelligence Questions define what the organization needs to know. Secondary research should be organized around answering those questions — not around browsing interesting topics. Every search should trace back to a specific intelligence need.
Develop your keywords. For each KIQ, develop a set of search terms: compound names, mechanism of action terms, indication names, competitor company names, investigator names. Think about synonyms, brand names vs. generic names, different spellings. The way you ask questions of the web or other sources must change depending on what you're looking for.
Allocate time. Secondary research can expand to fill whatever time you give it. Set boundaries. How many products are you researching? How many companies? How deep does each one need to go? A common pattern: spend 60% of your time on the top 2–3 competitive threats, 30% on the broader landscape, and 10% on emerging or potential competitors.
Organize by structure. Decide how you'll organize what you find — by company, by therapeutic area, by indication, or by product. If a company has a large portfolio, organize by product within company. The organizational structure you choose now will determine how easily you can find and use the information later.
Identify source types to target. Not all sources are equally relevant. Depending on your KIQs, you may need to focus on KOLs, investigators, company insiders, investor relations, field force contacts, or government agencies. Mapping these at the planning stage prevents aimless searching.
Searching is a two-level process. The first level establishes the broad picture. The second level fills in the details.
Level 1 — The Broad Sweep. Start with Google, company websites, and product websites. This gives you the public narrative: what the competitor says about themselves, what they're highlighting to investors, what their marketing messages look like. It also surfaces news, press releases, and recent developments. Company websites are particularly valuable for their news pages, investor relations sections, government filings, investor presentations, and R&D updates.
Level 2 — The Targeted Dig. Move to specialized databases: ClinicalTrials.gov, PubMed, LinkedIn. This is where you find the clinical data, the investigator names, the publication history, and the professional networks that support primary collection.
The mistake is spending too much time at Level 1. Google searches and news scanning are necessary but they give you what everyone else already has. Public data is available to everyone. It is necessary but not sufficient. The value starts at Level 2, where you're finding specific clinical details, investigator names, and regulatory signals that require effort and expertise to interpret.
Information you can't find again is information you don't have.
Three categories of information need structured storage:
Contact information. Names, titles, organizations, contact details, and — critically — notes on what they know, what they've said publicly, and how they might be approached. This is the seed for your source network.
Product information. Clinical data, pipeline status, regulatory filings, market share data, pricing information. Structured by product and updated continuously.
Company information. Financial results, organizational structure, strategy statements, partnership announcements, leadership changes. Structured by company.
The filing system matters more than most people think. Projects should be organized around therapeutic areas — not just individual products — because multiple projects often draw on the same source data. A well-structured filing system means any team member can quickly retrieve relevant background, not just the person who originally found it.
Secondary research is not a one-time project. After the initial research phase, you need systematic ongoing monitoring to catch changes as they happen.
RSS feeds from key journals, news outlets, and competitor websites deliver updates automatically. They reduce the daily scanning burden and ensure you don't miss important developments.
Google Alerts for competitor names, product names, key investigator names, and indication terms capture mentions across the web. Set them up early and review regularly.
Subscription services — SeekingAlpha for earnings call transcripts, PubMed alerts for new publications, regulatory agency notifications for filing updates — provide structured, domain-specific monitoring.
The monitoring stage is where secondary research becomes an ongoing CI capability rather than a one-time exercise. Done well, it provides a continuous stream of raw material that feeds regular updates to competitor profiles, landscape assessments, and intelligence alerts.
Google is the starting point for most secondary research, and most people underuse it. Basic searches are fine for getting oriented. But for CI, you need to know the advanced operators.
Start with combinations that surface competitive intelligence specifically:
Google's advanced operators refine searches dramatically. The syntax is `Operator:search_term` — no space between the operator and the colon.
`site:` — Restricts results to a specific website. `site:a top-10 pharma company.com pipeline` returns only the pharma company's own pipeline content. Particularly useful for mining competitor websites systematically.
`inurl:` — Searches within URLs. `inurl:investor-relations` combined with a company name finds their IR pages directly.
`filetype:` — Searches for specific file types. `filetype:pdf "competitive landscape" oncology` surfaces PDF reports and presentations that often contain more detailed analysis than web pages.
`intitle:` — Searches within page titles. `intitle:"Phase III results" competitor_name` finds pages specifically announcing trial results.
Competitor websites are structured information goldmines when you know where to look:
News page — press releases on trial initiations, data readouts, regulatory submissions, partnerships, and leadership changes. The cadence and content of press releases is itself intelligence — a company that suddenly goes quiet after months of regular updates may be dealing with setbacks.
Investor relations page — earnings presentations, annual reports, SEC filings, investor day transcripts. These documents often contain more candid strategic commentary than marketing materials. When a CEO tells investors "we're prioritizing Indication A over Indication B," that's intelligence about resource allocation you won't find anywhere else.
Government filings — SEC filings (10-K, 10-Q, 8-K) for US-listed companies contain risk factors, competitive landscape descriptions, and financial details that are legally required to be accurate. For European companies, equivalent filings through regulatory authorities serve the same purpose.
R&D pages — pipeline charts, technology platform descriptions, research publications. These provide the clinical and scientific context that feeds product-level competitive analysis.
In pharma competitive intelligence, secondary sources fall into distinct categories, each serving a different purpose.
ClinicalTrials.gov is the single most important secondary source for pharma CI. It's where you find competitor trial designs, endpoints, enrollment status, investigator names, and expected completion dates.
A basic search gets you started — search for a drug name or indication and scan the results. But the advanced search is where the real value is. You can refine by phase of development, sponsor company, and recruiting status. The enrollment status and enrollment rate of competitor trials give you timeline intelligence. The primary endpoints tell you what efficacy claims they're planning to make. The investigator names feed your source identification for primary collection.
WHO International Clinical Trials Registry Platform (ICTRP) captures trials registered outside the US, extending your visibility into global development programs.
PubMed is the primary source for published clinical data. Refining by date, author, and keyword helps you track competitor publication strategies. If a competitor suddenly publishes three papers on a specific safety endpoint, that signals something about their differentiation strategy.
FDA — Advisory committee documents, approval letters, complete response letters, label information, and post-marketing requirements. FDA.gov contains a wealth of competitive intelligence, particularly for products in late-stage development or recently approved.
EMA — European Public Assessment Reports (EPARs) provide detailed regulatory assessments that include clinical data summaries competitors would prefer you not read. Country-level agencies (like AIFA in Italy) add local regulatory context.
SEC filings — As discussed above, 10-K annual reports, 10-Q quarterly reports, and 8-K current reports contain risk factors that often name competitors, describe market dynamics, and quantify financial exposure.
USPTO / European Patent Office — Patent searches reveal where competitors are investing in R&D. The European Patent Office publishes all patent applications within 18 months of filing, giving you earlier visibility than many other patent databases. Patent filings on new formulations, delivery devices, or manufacturing processes signal lifecycle management strategies.
LinkedIn is CI source material, not just a networking tool. Search by product name, brand name, company name, or job title. Competitor employees' profiles reveal organizational structure, recent hires (which signal capability-building), and professional backgrounds.
Job portals — competitor job postings are one of the most underused secondary sources. If a competitor is hiring a dozen MSLs in your indication, that signals launch preparation. If they're posting for a head of market access, they're getting serious about commercialization. Job postings reveal strategic priorities before earnings calls do.
CafePharma — anonymous forums where pharmaceutical sales representatives discuss their companies, products, and competitors. The information needs validation, but it often provides ground-level intelligence about sales force morale, competitive messaging, and market dynamics that you won't find in official sources.
SeekingAlpha — analyst commentary and earnings call transcripts. Analyst questions during earnings calls often surface competitive dynamics that management might not volunteer. When an analyst asks "how are you responding to Competitor X's pricing strategy?", the response is intelligence.
Investor presentations — quarterly results presentations, investor day decks, and analyst conference presentations. These documents reveal management's view of the competitive landscape and their strategic priorities.
AdisInsight — comprehensive pipeline database with drug profiles, development status, and timelines.
TrialTrove — clinical trial analytics, enrollment data, and trial design comparisons.
Cortellis — competitive intelligence platform covering drugs, patents, deals, and regulatory information.
These paid sources add depth, structure, and historical perspective that free sources can't match. Whether they're worth the investment depends on the scope and frequency of your CI needs.
Here's where most secondary research falls short. Finding information is relatively easy. Knowing whether to trust it is the skill that separates useful intelligence from noise.
Every source should be evaluated on four dimensions:
Past performance. Has this source provided accurate information before? Government agencies and peer-reviewed journals have strong track records. Anonymous forum posts do not.
Proximity. How close is the source to the actual information? A clinical trial investigator commenting on trial enrollment is a high-proximity source. A journalist summarizing that trial is lower proximity. A blog post citing the journalist's article is lower still. Each layer of distance introduces potential distortion.
Access. Does the source actually have access to the information they're reporting? An investor relations officer discussing their own company's pipeline has high access. An industry analyst speculating about the same pipeline has lower access. Understanding the difference prevents you from treating speculation as fact.
Authority. How authoritative is the source on this specific topic? A hematologist commenting on a cardiovascular trial design is less authoritative than one commenting on a hematology trial, even if both are credentialed physicians. Authority is topic-specific, not general.
No single secondary source should be treated as confirmed intelligence. The way to validate is through triangulation — confirming the same finding across multiple independent sources.
If a ClinicalTrials.gov entry shows a competitor trial completing enrollment, and a conference presentation shows interim data consistent with that timeline, and an earnings call mentions expected Phase III results in the next quarter, those three independent sources converge on the same picture. That's intelligence you can act on.
If you have one anonymous forum post claiming a competitor is having manufacturing problems and nothing else confirms it — that's a lead to investigate, not intelligence to report.
The dangers of secondary information bear repeating. There is a lot of duplicate data out there — the same press release republished by ten different outlets is still one source, not ten. The data may be outdated — a competitor's pipeline chart from 18 months ago doesn't reflect today's reality. And most critically, data is not analysis. A collection of facts about a competitor is not the same as an assessment of what those facts mean for your strategy.
Good secondary research has three qualities. It is targeted — organized around specific KIQs, not general browsing. It is validated — triangulated across multiple sources with reliability explicitly assessed. And it connects to action — every finding is framed in terms of what it means for the organization's decisions, not just what it is.
The analyst who produces a secondary research package that answers "what do we know?" is adequate. The analyst who produces one that answers "what does this mean, and what should we do about it?" is exceptional.
The shift from project-based secondary research to ongoing monitoring is where CI matures from a series of exercises into a sustained capability.
The basic stack:
Google Alerts — set up alerts for competitor company names, product names (brand and generic), key investigator names, indication terms, and regulatory agency names. Use the "as-it-happens" frequency for critical competitors and "daily" for the broader landscape.
RSS feeds — subscribe to competitor press release feeds, key journal tables of contents, FDA and EMA alert feeds, and patent office publication feeds. An RSS reader consolidates these into a daily scanning exercise.
Calendar-driven checks — quarterly earnings call dates, major conference dates, regulatory decision dates, and patent expiry dates should be mapped on a calendar. These are predictable competitive events that trigger focused research.
Subscription notifications — set up alerts within whatever paid databases you use: AdisInsight pipeline updates, TrialTrove enrollment changes, SeekingAlpha earnings transcript alerts.
Not everything needs daily attention. A practical cadence:
Daily scanning (15–30 minutes): news alerts, RSS feeds, Google Alerts. Quick triage — flag anything that requires deeper analysis.
Weekly review (1–2 hours): review flagged items in depth, update product profiles with new data, check clinical trial registries for status changes.
Event-triggered deep dives (as needed): major conference presentations, regulatory decisions, competitor earnings calls, pipeline announcements. These are the events that warrant pulling up the competitor profile and assessing whether threat levels, timelines, or scenarios need updating.
The evolution of secondary research monitoring is one of the areas where AI has had the biggest practical impact on CI.
Three years ago, we had people doing news monitoring who would spend probably about three hours per day logging into email, Google Alerts, different access feeds and different databases. All of that can now be centralized. AI-powered platforms monitor 3,000+ sources every 15 minutes, process the information, filter for relevance, and deliver alerts with preliminary analysis. What used to be a three-hour daily task becomes a 20-minute review of curated, prioritized intelligence.
of a mid-size pharmaceutical company — a smaller company — describes AI as an equalizer in this context. Before AI, his team relied heavily on external vendors for competitive analysis. Expensive and slow. Now they can establish strategy and test different assumptions relatively instantly. The secondary research that used to require a dedicated analyst spending days on manual searching can now be automated at a fraction of the cost.
But — and this matters — automating secondary research doesn't eliminate the need for human judgment. AI monitors and filters. Humans assess significance, connect findings to strategy, and decide what to do. The technology handles the gathering. The analyst handles the "so what."
from the pharma company's Enterprise Strategy captures it precisely: "I want AI to do the data gathering so I can spend time on strategic thinking." That's the model. Secondary research as the automated foundation. Strategic analysis as the human contribution. Each making the other more valuable.
A few patterns I see repeatedly when secondary research programs underperform:
Searching without a plan. If you don't know what questions you're trying to answer, you'll find plenty of interesting information that doesn't connect to any decision. Plan first, search second.
Treating all sources equally. A peer-reviewed journal article, a press release, and an anonymous forum post are not equivalent. Build source assessment into your process, not as an afterthought.
Stopping at collection. The biggest pitfall. Gathering 200 pages of competitor information is not secondary research — it's data hoarding. Secondary research produces targeted analysis that answers specific questions.
Not setting up monitoring. One-time research efforts produce rapidly depreciating intelligence. If you did excellent secondary research last quarter but haven't monitored since, your picture of the competitive landscape is already outdated.
Ignoring the connection to primary collection. Secondary research exists to enable better human intelligence gathering. If your secondary research doesn't surface source names, safe topics, and conversation entry points, it's not doing its job.
Secondary research is the systematic gathering and analysis of information from published and publicly available sources — company websites, clinical trial registries, patent databases, regulatory filings, scientific publications, financial reports, and news media. It's called "secondary" because the information already exists; you're not generating new data through surveys or interviews. In CI, it serves as the foundation for all other intelligence activities, including primary (human) collection.
It depends on the industry. In pharma, the most valuable sources are ClinicalTrials.gov (for trial designs and timelines), FDA/EMA filings (for regulatory intelligence), PubMed (for published clinical data), SEC filings (for financial and strategic intelligence), patent databases (for IP intelligence), and investor presentations and earnings call transcripts (for strategic direction). Job postings are among the most underused yet revealing sources. For any industry, competitor websites, government filings, industry publications, and professional networks like LinkedIn form the core source set.
Through triangulation — confirming findings across multiple independent sources. Evaluate each source on four dimensions: past performance, proximity to the information, access to the actual data, and authority on the specific topic. Never treat a single source as confirmed intelligence. Watch for common traps: republished content that looks like multiple sources but originates from one, outdated information presented without date context, and opinion presented as fact.
For a new competitive intelligence project, initial secondary research typically takes 2–5 days per major competitor, depending on how much published information exists. Ongoing monitoring should take 15–30 minutes daily for scanning, plus 1–2 hours weekly for deeper review. Event-triggered deep dives (new data presentations, regulatory decisions) are additional. AI-powered monitoring can significantly reduce the daily scanning time by automating source crawling and relevance filtering.
Secondary research uses published and publicly available information — anyone could find it with the right tools and effort. Primary research involves gathering non-public information directly from human sources through conversations, interviews, and elicitation. Secondary research is the foundation; primary research builds on it. Public data is available to everyone and necessary but not sufficient. Only non-public data gathered through human sources creates true competitive advantage.
AI automates the monitoring layer — crawling thousands of sources, filtering for relevance, flagging changes, and delivering curated alerts. What used to require hours of daily manual scanning now takes minutes. AI can also help with initial categorization and summarization of large volumes of information. But validation, interpretation, and connecting findings to strategy remain human tasks. The most effective model uses AI for gathering and humans for analysis — not AI as a replacement for skilled secondary research analysts.
When your secondary research has established a solid baseline but identified gaps that public sources can't fill. Specifically: when you need to understand a competitor's intentions or strategic thinking (not just their public actions), when published data is outdated or insufficient for decision-making, when you need to validate or contradict secondary findings, or when you need perspective from domain experts that hasn't been published. The secondary research should prepare you for primary collection by identifying who to talk to, what to ask about, and what topics are safe entry points.
Salvador Carlucci is the founder of the Atacana Group and has spent over two decades building competitive intelligence functions for pharmaceutical and healthcare companies worldwide. He is co-author of the seminal articles on CI in biotech published in Nature Biotechnology.
© CompetitiveIntelligence.com