Earnings-call digest for enterprise AEs (L3)
LLM ingests target accounts' earnings calls + 10-Ks; AE gets a 1-pager before exec meetings.
The steps
- 01
Curate the high-priority account list
To ensure high-quality output, start by identifying the 50 accounts with the highest potential Contract Value (ACV) or strategic importance. Do not try to boil the ocean; enterprise research scales poorly if the initial data is messy. • Open your CRM (Salesforce/HubSpot) and create a custom report filtered by 'Account Type = Prospect' and 'Employee Count > 5,000' or 'Revenue > $1B'. • Export this list to a Google Sheet with columns for Account Name, Ticker Symbol (crucial for financial data), and Assigned AE. • The RevOps Manager should own this step, taking approximately 2 hours to validate the list with Sales Leadership. • A common pitfall is including private companies; ensure every company on this list has a ticker symbol for public filings. • Definition of Done: A locked Google Sheet containing 50 validated public enterprise accounts with accurate ticker symbols. • QA Check: Cross-reference five tickers on Yahoo Finance to ensure they match the legal entity names in your CRM.
- 02
Automate financial document ingestion
You need a consistent source for Earnings Call Transcripts and 10-K filings. • Use an API-based financial data provider like Financial Modeling Prep, Alpha Vantage, or Bambu. If you lack a budget, use a web scraper or manually download the 'Management's Discussion and Analysis' section of the 10-K from the SEC EDGAR database. • Set up an automation in Zapier or Make.com that triggers whenever a new filing is detected for your list of tickers. • The Sales Ops Specialist or a Junior Developer owns this setup, requiring 4-6 hours. • Pitfall: Relying on generic news summaries instead of raw transcripts; LLMs provide better insights when given the primary source text. • Definition of Done: A cloud storage folder (Google Drive/S3) automatically populating with .txt or .pdf files of the latest transcripts for all 50 accounts. • QA Check: Verify that the most recent quarterly (10-Q) or annual (10-K) report for the top 3 accounts is present in the folder.
- 03
Configure the LLM extraction schema
Now, configure your LLM (GPT-4 or Claude 3 Opus) to extract specific, actionable insights using a rigorous prompt template. You want the AE to walk into a meeting sounding like a consultant, not a news reader. • In your LLM orchestration tool (e.g., LangChain, Clay, or a custom Python script), use this prompt: 'Analyze the attached earnings transcript. Extract: 1) Top 3 Strategic Pillars for the next 12 months. 2) Key Risks mentioned by the CFO. 3) Exact mentions of "Artificial Intelligence" or "Automation". 4) Recent leadership changes in the C-Suite. Format as a clean 1-page executive brief.' • The RevOps Lead owns this, taking 3 hours to tune the prompt. • Pitfall: Asking for 'general summaries', which results in fluff. Force the LLM to use bullet points and cite specific quotes from the CEO. • Definition of Done: A set of 50 generated briefs stored in a 'Research' field in your CRM or a shared Slack channel. • QA Check: Read one brief and ensure it lists the 'Strategic Pillars' with specific business goals (e.g., 'Reducing OpEx by 15%') rather than vague statements.
- 04
Deliver insights to the AE workflow
The research is useless if it sits in a folder; it must be pushed to the AE exactly when they need it. • Set up an automated workflow where the LLM-generated brief is pushed to the 'Account' record in your CRM. • Additionally, create a Slack or Teams notification that alerts the AE: 'New Earnings Digest for [Account Name] - View here.' • Ensure the brief is also emailed to the AE 24 hours before any calendar event that includes a contact from that account. • The RevOps Manager owns this integration, taking 4 hours using Zapier or a CRM native automation (like Salesforce Flow). • Pitfall: Notification fatigue. Only send the 'digest' alert once per quarter (after the earnings call) rather than every time a news article mentions the company. • Definition of Done: A live automation that updates the CRM 'Earnings Summary' field and notifies the owner. • QA Check: Trigger a test notification for a dummy account and verify the link leads directly to the 1-page brief.
- 05
Track time savings and meeting impact
To measure success, track how this automated research impacts AE behavior and deal velocity. • In your CRM, create a custom field called 'Research Prep Time' on the Opportunity or Event object. • Ask AEs to log their prep time for exec meetings for one month before the project starts, and then compare it to the time logged after the AI briefs are implemented. • Set up a dashboard in your CRM showing 'Exec Meeting Conversion Rate' (Meetings booked vs. Held) for the 50 pilot accounts versus a control group. • The AE and Sales Manager own the data entry, while RevOps owns the reporting, taking 1 hour/week for review. • Pitfall: AEs forgetting to log time. Use a simple 'Was this brief helpful? (Yes/No)' button at the bottom of the digest to track engagement. • Definition of Done: A monthly report showing a 50% reduction in manual research time and a detectable lift in meeting-to-stage-1 conversion. • QA Check: Cross-reference the 'Research Prep Time' field to ensure at least 80% coverage on new enterprise meetings.
Next playbooks
Unfamiliar terms are defined in the AI and Revenue Dictionary. Related frameworks live in the framework library.