A while back, I blogged about how analysts and investors are increasingly using AI tools to read and analyze earnings reports (10-Ks, 10-Qs, earnings releases and transcripts) to gain faster insights and identify investment opportunities. Here are 10 tips to consider when drafting an earnings release with the AI reader in mind:
1. Use clear and consistent terminology: Avoid jargon and ambiguous phrases, and use standard terms consistently. For example, say: “Adjusted EBITDA increased 12%” – and avoid, “We saw meaningful progress in operating performance.”
2. Structure your narrative clearly: Use bullet points, clear headings and consistent section ordering (e.g., Results ? Drivers ? Outlook). AI tools digest structured, predictable formats more accurately.
3. Include key metrics in both text and tables: AI extracts better from numerical tables, but duplicating metrics in text ensures redundancy. Include EPS, revenue, margins and cash flow in both formats.
4. Avoid overuse of boilerplate risk language: AI can detect hedging or overly generic disclosures (e.g., “may,” “could,” “possibly”). Add specific context to risk factor and MD&A disclosures.
5. Be precise in forward-looking statements: Include quantified guidance ranges, time frames and assumptions. Instead of: “We expect growth to continue,” say, “We expect 4-6% revenue growth in FY2025, driven by … .”
6. Minimize PDF formatting issues: AI parsing tools struggle with poorly tagged PDFs or unusual layouts. Publish in machine-readable HTML or ensure tagged PDF/XBRL formatting is clean.
7. Cross-link sections thoughtfully: Use internal hyperlinks between key financial tables, footnotes and narratives when digital formats allow. This improves navigation for both human and machine readers.
8. Tag disclosures consistently in XBRL: Ensure key metrics and narrative blocks are properly tagged. Inaccurate or missing tags hinder AI scraping and comparison.
9. Keep earnings call transcripts human and honest: Investors use AI to analyze tone and evasiveness. Be authentic and avoid robotic scripts.
10. Monitor how your report is being read: Use market surveillance tools to see how your disclosures are flagged, quoted or rated by AI-driven investor platforms.
Authored by
Broc Romanek