How to Use AI and 10-K Reports to Uncover Sales Insights That Win More Deals
Kevin Lawson and Sean O'Shaughnessy discuss how sales professionals can use AI tools to extract actionable sales intelligence from publicly available 10-K annual reports. They argue that most salespeople ignore this resource despite it containing explicit information about a company's strategic priorities, risks, hiring plans, and revenue outlook. They provide specific AI prompting strategies to make this process fast and accessible for any seller.
Summary
The episode opens with Sean O'Shaughnessy framing the broader context: sales methodology has evolved dramatically decade by decade, and AI represents the next major shift entering the second half of the 2020s. Both hosts promote their community, the B2B Sales Lab, as a resource for salespeople and sales leaders navigating these changes.
The core topic is how to use AI to analyze 10-K annual reports — the SEC-mandated annual filing required of all public companies. The hosts note that most salespeople never read these documents, despite them being freely available online via company investor relations pages or SEC.gov. They argue this represents a major missed opportunity, as 10-Ks contain forward-looking statements about company strategy, risk factors, hiring intentions, revenue expectations, and legal challenges.
Kevin explains the distinction between SEC filing types: the 10-K is the comprehensive annual report, the 10-Q is a quarterly financial update, and the 8-K covers significant business events. He emphasizes that sellers should also look for earnings call transcripts, which capture executive commentary in plain language and complement the formal 10-K document.
On AI prompting, Kevin shares his preferred approach: instructing the model to 'act as a senior analyst tasked with developing key selling insights,' then asking for the top five company initiatives for the upcoming year, while explicitly telling the model not to make assumptions and to cite only exact quotes from the uploaded documents. He stresses that this constraint on the model prevents hallucinated or generic output.
Sean builds on this by recommending that sellers first give the AI context about their own product or service before asking questions, so the model can filter insights through the lens of what is actually relevant to the seller's offering. He provides a list of specific high-value questions to ask the AI: expected company revenue for the next year, hiring plans and challenges, how geopolitical events or tariffs affect the company, how fuel prices impact operations, and what active lawsuits could affect company health. Sean notes that tariffs, for instance, help some companies while hurting others, and understanding which side a prospect falls on can sharpen targeting.
Both hosts conclude by positioning this as a foundational habit for enterprise sellers, arguing that building this practice into regular account planning gives salespeople a perspective most competitors will simply not have.
Key Insights
- Kevin argues that the single most important constraint in an AI research prompt is explicitly instructing the model not to make assumptions and to report only exact quotes from the source document — this prevents the generic or hallucinated output that makes AI-generated summaries unreliable for sales use.
- Sean contends that sellers should prime the AI with a brief description of their own product or service before asking questions about a prospect's 10-K, because the relevance of what the model surfaces depends entirely on understanding what problems the seller actually solves.
- The hosts argue that 10-K reports are uniquely valuable because companies are legally required to disclose not just financial performance but also forward-looking risks — including geopolitical exposure, pending litigation, and macroeconomic vulnerabilities — giving sellers a rare window into what executives are actually worried about.
- Kevin claims that combining the formal 10-K PDF with the earnings call transcript gives sellers two complementary lenses on the same company: the structured financial and legal record plus the unscripted executive voice, and that uploading both to an AI chat session yields richer insight than either document alone.
- Sean argues that tariff and geopolitical analysis within a 10-K can be used offensively in prospecting — identifying which companies are positioned to benefit from current trade conditions and prioritizing outreach to those accounts over ones facing headwinds.
Topics
Full transcript available for MurmurCast members
Sign Up to Access