Earnings Call Transcripts Scraper

Introduction

The Earnings Call Transcripts Scraper extracts and structures earnings call transcripts for publicly traded companies. Point it at one or more ticker symbols and it returns clean, machine-readable JSON with the full transcript, speaker-by-speaker breakdowns, Q&A exchanges, extracted financial metrics (revenue, EPS, margins, guidance), and forward-looking statements automatically flagged and categorised.

Transcripts are sourced from Motley Fool (the most scraper-friendly public source), SEC EDGAR 8-K filings (the gold-standard public record), and Seeking Alpha preview pages. The actor tries your preferred source first and falls back automatically so you always get maximum coverage.

Primary use cases:

• Financial modelling teams pulling earnings data into spreadsheets or data pipelines
• AI and LLM teams building training datasets from real financial language
• Hedge funds and asset managers monitoring portfolio companies quarterly
• Financial journalists researching earnings trends and executive commentary
• Academic researchers studying corporate communications and guidance accuracy
• Individual investors doing fundamental analysis on watchlist companies
• Fintech startups building financial data products without expensive subscriptions

Key differentiators: multi-source with automatic fallback, structured JSON parsing (not just raw text), per-speaker segmentation, Q&A section identification, regex-based financial metric extraction, forward-looking statement categorisation, and pay-per-use pricing — no subscription required.

---

Why Use This Actor

Financial transcript data has historically been locked behind expensive subscriptions. This actor gives you programmatic access to public transcript content at a fraction of the cost.

Feature	This Actor	Seeking Alpha Premium	AlphaSense	Financial Modeling Prep API
Price	$10.40 / 1,000 transcripts	$239/year subscription	$2,000+/year	$29/month subscription
Per-transcript pricing	Yes (PPE)	No	No	No
Structured JSON output	Yes	HTML/PDF	Proprietary	JSON (limited)
Speaker identification	Yes	Yes	Yes	Basic
Q&A section parsing	Yes	Manual	Yes	No
Financial metrics extraction	Yes	No	Yes (proprietary NLP)	No
Forward-looking statements	Yes	No	Yes	No
SEC EDGAR fallback	Yes	No	No	No
Zero-config first run	Yes	Account required	Enterprise sales	API key required
Apify ecosystem integration	Full	None	None	None

Cost comparison: Analysing one company across four quarters costs $0.04 with this actor. Doing the same for a 50-stock portfolio across one year (200 transcripts) costs $2.08 — versus hundreds of dollars per month on any subscription platform.

---

Proxy Requirements

This actor requires residential proxy to access Motley Fool and SEC EDGAR, which block datacenter IP addresses.

• Apify paid plans: Residential proxy is included. The actor uses it by default.
• Apify free plan: Free plan does not include residential proxy. You can provide your own residential proxy URL in the proxy configuration, or the actor will attempt to run without proxy (results may be empty or blocked).
• Without residential proxy: The actor will still run but may return zero results due to IP blocking by Motley Fool and SEC EDGAR.

How to Use

Zero-config (just add tickers)

{
  "tickers": ["AAPL", "MSFT"]
}

This retrieves the most recent transcripts for Apple and Microsoft from all configured sources with full parsing enabled.

Multi-company portfolio monitoring

{
  "tickers": ["AAPL", "MSFT", "GOOGL", "AMZN", "META", "NVDA"],
  "maxTranscriptsPerCompany": 1,
  "quarters": ["Q1 2025"]
}

Historical research across multiple years

{
  "tickers": ["AAPL"],
  "maxTranscriptsPerCompany": 12,
  "fiscalYears": [2023, 2024, 2025],
  "extractMetrics": true,
  "extractForwardLooking": true
}

Metadata-only lightweight run (smaller output)

{
  "tickers": ["TSLA", "NFLX", "UBER"],
  "includeFullText": false,
  "maxTranscriptsPerCompany": 4
}

Via the Apify API (Node.js example)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_API_TOKEN' });

const run = await client.actor('junipr/earnings-call-scraper').call({
  tickers: ['AAPL', 'MSFT'],
  maxTranscriptsPerCompany: 4,
  extractMetrics: true,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

---

Input Configuration

Parameter	Type	Default	Description
tickers	array	["AAPL", "MSFT"]	Stock ticker symbols. Max 50.
companyNames	array	[]	Company names as fallback when ticker lookup fails.
cikNumbers	array	[]	SEC CIK numbers for direct EDGAR lookup.
quarters	array	[]	Filter by fiscal quarter, e.g. ["Q1 2025"]. Empty = all.
fiscalYears	array	[]	Filter by fiscal year, e.g. [2024, 2025]. Empty = all.
maxTranscriptsPerCompany	integer	4	Max transcripts per company (1–20).
dateFrom	string	null	ISO 8601 lower date bound, e.g. "2024-01-01".
dateTo	string	null	ISO 8601 upper date bound, e.g. "2025-12-31".
sources	array	["motley_fool", "sec_edgar"]	Sources to try. Options: motley_fool, seeking_alpha, sec_edgar.
preferredSource	string	"motley_fool"	Primary source tried first. Falls back if unavailable.
extractMetrics	boolean	true	Extract financial metrics (revenue, EPS, margins, guidance).
extractForwardLooking	boolean	true	Flag forward-looking statements.
includeFullText	boolean	true	Include the complete transcript text. Set to false for smaller output.
splitBySpeaker	boolean	true	Parse transcript into per-speaker segments.
proxyConfiguration	object	{ "useApifyProxy": false }	Proxy settings. Residential proxy recommended for Seeking Alpha.

---

Output Format

Each dataset item represents one earnings call transcript. Key fields:

{
  "ticker": "AAPL",
  "companyName": "Apple Inc.",
  "cik": "0000320193",
  "quarter": "Q1 2025",
  "fiscalYear": 2025,
  "callDate": "2025-01-30",
  "callType": "earnings",
  "source": "motley_fool",
  "sourceUrl": "https://www.fool.com/earnings-call-transcripts/...",
  "transcript": {
    "fullText": "...",
    "wordCount": 8421,
    "sections": {
      "operatorIntro": "Good afternoon and welcome to the Apple Q1 2025 earnings call...",
      "preparedRemarks": [
        {
          "speaker": "Tim Cook",
          "title": "CEO",
          "affiliation": "Apple",
          "text": "Thank you. We're pleased to report another record quarter..."
        }
      ],
      "questionAndAnswer": [
        {
          "questionSpeaker": "Erik Woodring",
          "questionAffiliation": "Morgan Stanley",
          "questionText": "Can you talk about iPhone demand trends in China?",
          "answerSpeaker": "Tim Cook",
          "answerTitle": "CEO",
          "answerText": "We remain very excited about the opportunity in China..."
        }
      ],
      "closingRemarks": "Thank you all for joining. That concludes today's call."
    }
  },
  "participants": [
    { "name": "Tim Cook", "title": "CEO", "affiliation": "Apple", "role": "executive" },
    { "name": "Luca Maestri", "title": "CFO", "affiliation": "Apple", "role": "executive" },
    { "name": "Erik Woodring", "title": "Analyst", "affiliation": "Morgan Stanley", "role": "analyst" }
  ],
  "metrics": [
    { "name": "revenue", "value": "revenue of $124.3 billion", "period": "Q1 2025", "context": "...total revenue of $124.3 billion for the quarter..." }
  ],
  "forwardLookingStatements": [
    { "statement": "We expect revenue to grow in the mid-single digits in Q2.", "speaker": "Luca Maestri", "category": "guidance" }
  ],
  "scrapedAt": "2025-03-11T12:00:00Z",
  "errors": []
}

A run summary is written to the Key-Value Store under OUTPUT with totals, per-company breakdowns, and source attribution.

---

Tips and Advanced Usage

Building financial models: The metrics array gives you structured value strings (e.g. "$124.3 billion") with surrounding context. Feed these into a normalisation script or LLM prompt to extract clean numbers for spreadsheet ingestion.

Sentiment analysis from Q&A tone: The questionAndAnswer array lets you compare analyst question sentiment versus executive answer length and tone across consecutive quarters — a signal used in quantitative research.

Forward-looking statement accuracy tracking: Collect forwardLookingStatements from one quarter and compare against actual reported figures the following quarter to score guidance accuracy over time.

LLM summarisation pipeline: Pass transcript.fullText directly to GPT-4 or Claude with a prompt like "Summarise this earnings call in 5 bullet points with the key guidance." The word count field helps you chunk long transcripts for context window limits.

Event study research: Combine callDate with stock price APIs (e.g. Yahoo Finance) to study abnormal returns around earnings call dates. The structured data makes it easy to tag calls with positive vs negative guidance and correlate with price movements.

Combining with other actors: Pair this actor with a news scraper to correlate earnings call themes with subsequent news coverage, or with a financial data API for quantitative signal generation.

---

Pricing

This actor uses Pay-Per-Event (PPE) pricing: $10.40 per 1,000 transcripts scraped.

Pricing includes all platform compute costs — no hidden fees.

A billable event (transcript-scraped) is charged only when a transcript is successfully located, text extracted, and the result pushed to the dataset. Failed lookups, blocked sources, and companies with no transcripts available are not billed.

Scenario	Transcripts	Estimated Cost
Single company, latest quarter	1	$0.01
Portfolio (10 stocks), latest quarter	10	$0.10
Annual review (1 company, 4 quarters)	4	$0.04
Sector analysis (50 companies, 1 quarter)	50	$0.52
Historical research (10 companies, 3 years)	120	$1.25
Quant fund quarterly refresh (500 companies)	2,000	$20.80

Compare to Seeking Alpha Premium at $239/year or AlphaSense starting at $2,000/year — pay only for what you actually use.

---

FAQ

What sources does it scrape transcripts from?

The actor scrapes Motley Fool (most scraper-friendly, good coverage), SEC EDGAR 8-K filings (official public filings, highest legal certainty, US public companies only), and Seeking Alpha public preview pages (best coverage but anti-bot protection means residential proxy is recommended). You can configure which sources to use and in what order via sources and preferredSource.

Can I get transcripts for international companies?

SEC EDGAR only covers US-listed companies. Motley Fool and Seeking Alpha cover major international companies listed on US exchanges (e.g. ASML, NVO, TSM). For companies not in the built-in ticker map, provide the ticker and the actor will attempt to find transcripts using the ticker as a search term.

How far back do transcripts go?

Coverage varies by source. Motley Fool has transcripts going back to approximately 2016. SEC EDGAR EDGAR filings go back decades. In practice, the most reliable coverage starts around 2018–2019 across all sources.

Does it work for private companies?

No. Earnings call transcripts are a public company phenomenon — only companies required to report quarterly results (US exchange-listed companies) hold public earnings calls. Private companies do not have public transcripts.

What if a transcript is behind a paywall?

Seeking Alpha limits full transcript access to subscribers. When partial content is returned, the actor sets an PARTIAL_TRANSCRIPT error in the errors array and automatically falls back to Motley Fool or SEC EDGAR for a more complete version. You will never be billed for a failed or partial-only extraction.

Can I use this to train AI models?

Yes. The structured JSON output (speaker turns, Q&A exchanges, section boundaries) is well-suited for fine-tuning LLMs on financial language, training earnings call summarisation models, or building financial NLP datasets. SEC EDGAR content is explicitly public domain US government data with no copyright restrictions.

How accurate is the speaker identification?

Speaker identification uses pattern matching against common transcript formatting conventions (e.g. "Tim Cook — CEO" or "Tim Cook (Apple)"). Accuracy is high for well-formatted Motley Fool transcripts and moderate for SEC EDGAR filings which have more variable formatting. When a speaker cannot be identified, the segment is labeled "Unidentified" rather than silently dropped.

Does it extract specific financial numbers?

Yes. The metrics array uses regex patterns to find mentions of revenue, EPS, gross margin, operating income, net income, and guidance figures. Each match includes the raw value string (e.g. "$89.5 billion") and surrounding context for verification. These are best-effort extractions — always verify critical figures against official SEC filings before making financial decisions.