In this tutorial, we demonstrate how to leverage ’s powerful scraping tools in combination with Gemini AI to automate the collection, parsing, and analysis of competitor information. By using ScrapeGraph’s SmartScraperTool and MarkdownifyTool, users can extract detailed insights from product offerings, pricing strategies, technology stacks, and market presence directly from competitor websites. The tutorial then employs Gemini’s advanced language model to synthesize these disparate data points into structured, actionable intelligence. Throughout the process, ScrapeGraph ensures that the raw extraction is both accurate and scalable, allowing analysts to focus on strategic interpretation rather than manual data gathering.
%pip install --quiet -U langchain-scrapegraph langchain-google-genai pandas matplotlib seabornWe quietly upgrade or install the latest versions of essential libraries, including langchain-scrapegraph for advanced web scraping and langchain-google-genai for integrating Gemini AI, as well as data analysis tools such as pandas, matplotlib, and seaborn, to ensure your environment is ready for seamless competitive intelligence workflows.
import getpass
import os
import json
import pandas as pd
from typing import List, Dict, Any
from datetime import datetime
import matplotlib.pyplot as plt
import seaborn as snsWe import essential Python libraries for setting up a secure, data-driven pipeline: getpass and os manage passwords and environment variables, json handles serialized data, and pandas offers robust DataFrame operations. The typing module provides type hints for better code clarity, while datetime records timestamps. Finally, matplotlib.pyplot and seaborn equip us with tools for creating insightful visualizations.
if not os.environ.get("SGAI_API_KEY"):
os.environ["SGAI_API_KEY"] = getpass.getpass("ScrapeGraph AI API key:n")
if not os.environ.get("GOOGLE_API_KEY"):
os.environ["GOOGLE_API_KEY"] = getpass.getpass("Google API key for Gemini:n")We check if the SGAI_API_KEY and GOOGLE_API_KEY environment variables are already set; if not, the script securely prompts the user for their ScrapeGraph and Google (Gemini) API keys via getpass and stores them in the environment for subsequent authenticated requests.
from langchain_scrapegraph.tools import (
SmartScraperTool,
SearchScraperTool,
MarkdownifyTool,
GetCreditsTool,
)
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.runnables import RunnableConfig, chain
from langchain_core.output_parsers import JsonOutputParser
smartscraper = SmartScraperTool()
searchscraper = SearchScraperTool()
markdownify = MarkdownifyTool()
credits = GetCreditsTool()
llm = ChatGoogleGenerativeAI(
model="gemini-1.5-flash",
temperature=0.1,
convert_system_message_to_human=True
)
Here, we import and instantiate ScrapeGraph tools, the SmartScraperTool, SearchScraperTool, MarkdownifyTool, and GetCreditsTool, for extracting and processing web data, then configure the ChatGoogleGenerativeAI with the “gemini-1.5-flash” model (low temperature and human-readable system messages) to drive our analysis. We also bring in ChatPromptTemplate, RunnableConfig, chain, and JsonOutputParser from langchain_core to structure prompts and parse model outputs.
class CompetitiveAnalyzer:
def __init__(self):
self.results = []
self.analysis_timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
def scrape_competitor_data(self, url: str, company_name: str = None) -> Dict[str, Any]:
"""Scrape comprehensive data from a competitor website"""
extraction_prompt = """
Extract the following information from this website:
1. Company name and tagline
2. Main products/services offered
3. Pricing information (if available)
4. Target audience/market
5. Key features and benefits highlighted
6. Technology stack mentioned
7. Contact information
8. Social media presence
9. Recent news or announcements
10. Team size indicators
11. Funding information (if mentioned)
12. Customer testimonials or case studies
13. Partnership information
14. Geographic presence/markets served
Return the information in a structured JSON format with clear categorization.
If information is not available, mark as 'Not Available'.
"""
try:
result = smartscraper.invoke({
"user_prompt": extraction_prompt,
"website_url": url,
})
markdown_content = markdownify.invoke({"website_url": url})
competitor_data = {
"company_name": company_name or "Unknown",
"url": url,
"scraped_data": result,
"markdown_length": len(markdown_content),
"analysis_date": self.analysis_timestamp,
"success": True,
"error": None
}
return competitor_data
except Exception as e:
return {
"company_name": company_name or "Unknown",
"url": url,
"scraped_data": None,
"error": str(e),
"success": False,
"analysis_date": self.analysis_timestamp
}
def analyze_competitor_landscape(self, competitors: List[Dict[str, str]]) -> Dict[str, Any]:
"""Analyze multiple competitors and generate insights"""
print(f"
Starting competitive analysis for {len(competitors)} companies...")
for i, competitor in enumerate(competitors, 1):
print(f" Analyzing {competitor['name']} ({i}/{len(competitors)})...")
data = self.scrape_competitor_data(
competitor['url'],
competitor['name']
)
self.results.append(data)
analysis_prompt = ChatPromptTemplate.from_messages([
("system", """
You are a senior business analyst specializing in competitive intelligence.
Analyze the scraped competitor data and provide comprehensive insights including:
1. Market positioning analysis
2. Pricing strategy comparison
3. Feature gap analysis
4. Target audience overlap
5. Technology differentiation
6. Market opportunities
7. Competitive threats
8. Strategic recommendations
Provide actionable insights in JSON format with clear categories and recommendations.
"""),
("human", "Analyze this competitive data: {competitor_data}")
])
clean_data = []
for result in self.results:
if result['success']:
clean_data.append({
'company': result['company_name'],
'url': result['url'],
'data': result['scraped_data']
})
analysis_chain = analysis_prompt | llm | JsonOutputParser()
try:
competitive_analysis = analysis_chain.invoke({
"competitor_data": json.dumps(clean_data, indent=2)
})
except:
analysis_chain_text = analysis_prompt | llm
competitive_analysis = analysis_chain_text.invoke({
"competitor_data": json.dumps(clean_data, indent=2)
})
return {
"analysis": competitive_analysis,
"raw_data": self.results,
"summary_stats": self.generate_summary_stats()
}
def generate_summary_stats(self) -> Dict[str, Any]:
"""Generate summary statistics from the analysis"""
successful_scrapes = sum(1 for r in self.results if r['success'])
failed_scrapes = len(self.results) - successful_scrapes
return {
"total_companies_analyzed": len(self.results),
"successful_scrapes": successful_scrapes,
"failed_scrapes": failed_scrapes,
"success_rate": f"{(successful_scrapes/len(self.results)*100):.1f}%" if self.results else "0%",
"analysis_timestamp": self.analysis_timestamp
}
def export_results(self, filename: str = None):
"""Export results to JSON and CSV files"""
if not filename:
filename = f"competitive_analysis_{datetime.now().strftime('%Y%m%d_%H%M%S')}"
with open(f"{filename}.json", 'w') as f:
json.dump({
"results": self.results,
"summary": self.generate_summary_stats()
}, f, indent=2)
df_data = []
for result in self.results:
if result['success']:
df_data.append({
'Company': result['company_name'],
'URL': result['url'],
'Success': result['success'],
'Data_Length': len(str(result['scraped_data'])) if result['scraped_data'] else 0,
'Analysis_Date': result['analysis_date']
})
if df_data:
df = pd.DataFrame(df_data)
df.to_csv(f"{filename}.csv", index=False)
print(f" Results exported to {filename}.json and {filename}.csv")
Starting competitive analysis for {len(competitors)} companies...")
for i, competitor in enumerate(competitors, 1):
print(f"
Analyzing {competitor['name']} ({i}/{len(competitors)})...")
data = self.scrape_competitor_data(
competitor['url'],
competitor['name']
)
self.results.append(data)
analysis_prompt = ChatPromptTemplate.from_messages([
("system", """
You are a senior business analyst specializing in competitive intelligence.
Analyze the scraped competitor data and provide comprehensive insights including:
1. Market positioning analysis
2. Pricing strategy comparison
3. Feature gap analysis
4. Target audience overlap
5. Technology differentiation
6. Market opportunities
7. Competitive threats
8. Strategic recommendations
Provide actionable insights in JSON format with clear categories and recommendations.
"""),
("human", "Analyze this competitive data: {competitor_data}")
])
clean_data = []
for result in self.results:
if result['success']:
clean_data.append({
'company': result['company_name'],
'url': result['url'],
'data': result['scraped_data']
})
analysis_chain = analysis_prompt | llm | JsonOutputParser()
try:
competitive_analysis = analysis_chain.invoke({
"competitor_data": json.dumps(clean_data, indent=2)
})
except:
analysis_chain_text = analysis_prompt | llm
competitive_analysis = analysis_chain_text.invoke({
"competitor_data": json.dumps(clean_data, indent=2)
})
return {
"analysis": competitive_analysis,
"raw_data": self.results,
"summary_stats": self.generate_summary_stats()
}
def generate_summary_stats(self) -> Dict[str, Any]:
"""Generate summary statistics from the analysis"""
successful_scrapes = sum(1 for r in self.results if r['success'])
failed_scrapes = len(self.results) - successful_scrapes
return {
"total_companies_analyzed": len(self.results),
"successful_scrapes": successful_scrapes,
"failed_scrapes": failed_scrapes,
"success_rate": f"{(successful_scrapes/len(self.results)*100):.1f}%" if self.results else "0%",
"analysis_timestamp": self.analysis_timestamp
}
def export_results(self, filename: str = None):
"""Export results to JSON and CSV files"""
if not filename:
filename = f"competitive_analysis_{datetime.now().strftime('%Y%m%d_%H%M%S')}"
with open(f"{filename}.json", 'w') as f:
json.dump({
"results": self.results,
"summary": self.generate_summary_stats()
}, f, indent=2)
df_data = []
for result in self.results:
if result['success']:
df_data.append({
'Company': result['company_name'],
'URL': result['url'],
'Success': result['success'],
'Data_Length': len(str(result['scraped_data'])) if result['scraped_data'] else 0,
'Analysis_Date': result['analysis_date']
})
if df_data:
df = pd.DataFrame(df_data)
df.to_csv(f"{filename}.csv", index=False)
print(f"
Results exported to {filename}.json and {filename}.csv")The CompetitiveAnalyzer class orchestrates end-to-end competitor research, scraping detailed company information using ScrapeGraph tools, compiling and cleaning the results, and then leveraging Gemini AI to generate structured competitive insights. It also tracks success rates and timestamps, and provides utility methods to export both raw and summarized data into JSON and CSV formats for easy downstream reporting and analysis.
def run_ai_saas_analysis():
"""Run a comprehensive analysis of AI/SaaS competitors"""
analyzer = CompetitiveAnalyzer()
ai_saas_competitors = [
{"name": "OpenAI", "url": "https://openai.com"},
{"name": "Anthropic", "url": "https://anthropic.com"},
{"name": "Hugging Face", "url": "https://huggingface.co"},
{"name": "Cohere", "url": "https://cohere.ai"},
{"name": "Scale AI", "url": "https://scale.com"},
]
results = analyzer.analyze_competitor_landscape(ai_saas_competitors)
print("n" + "="*80)
print(" COMPETITIVE ANALYSIS RESULTS")
print("="*80)
print(f"n Summary Statistics:")
stats = results['summary_stats']
for key, value in stats.items():
print(f" {key.replace('_', ' ').title()}: {value}")
print(f"n Strategic Analysis:")
if isinstance(results['analysis'], dict):
for section, content in results['analysis'].items():
print(f"n {section.replace('_', ' ').title()}:")
if isinstance(content, list):
for item in content:
print(f" • {item}")
else:
print(f" {content}")
else:
print(results['analysis'])
analyzer.export_results("ai_saas_competitive_analysis")
return results
COMPETITIVE ANALYSIS RESULTS")
print("="*80)
print(f"nThe above function initiates the competitive analysis by instantiating CompetitiveAnalyzer and defining the key AI/SaaS players to be evaluated. It then runs the full scraping-and-insights workflow, prints formatted summary statistics and strategic findings, and finally exports the detailed results to JSON and CSV for further use.
def run_ecommerce_analysis():
"""Analyze e-commerce platform competitors"""
analyzer = CompetitiveAnalyzer()
ecommerce_competitors = [
{"name": "Shopify", "url": "https://shopify.com"},
{"name": "WooCommerce", "url": "https://woocommerce.com"},
{"name": "BigCommerce", "url": "https://bigcommerce.com"},
{"name": "Magento", "url": "https://magento.com"},
]
results = analyzer.analyze_competitor_landscape(ecommerce_competitors)
analyzer.export_results("ecommerce_competitive_analysis")
return resultsThe above function sets up a CompetitiveAnalyzer to evaluate major e-commerce platforms by scraping details from each site, generating strategic insights, and then exporting the findings to both JSON and CSV files under the name “ecommerce_competitive_analysis.”
@chain
def social_media_monitoring_chain(company_urls: List[str], config: RunnableConfig):
"""Monitor social media presence and engagement strategies of competitors"""
social_media_prompt = ChatPromptTemplate.from_messages([
("system", """
You are a social media strategist. Analyze the social media presence and strategies
of these companies. Focus on:
1. Platform presence (LinkedIn, Twitter, Instagram, etc.)
2. Content strategy patterns
3. Engagement tactics
4. Community building approaches
5. Brand voice and messaging
6. Posting frequency and timing
Provide actionable insights for improving social media strategy.
"""),
("human", "Analyze social media data for: {urls}")
])
social_data = []
for url in company_urls:
try:
result = smartscraper.invoke({
"user_prompt": "Extract all social media links, community engagement features, and social proof elements",
"website_url": url,
})
social_data.append({"url": url, "social_data": result})
except Exception as e:
social_data.append({"url": url, "error": str(e)})
chain = social_media_prompt | llm
analysis = chain.invoke({"urls": json.dumps(social_data, indent=2)}, config=config)
return {
"social_analysis": analysis,
"raw_social_data": social_data
}
Here, this chained function defines a pipeline to gather and analyze competitors’ social media footprints: it uses ScrapeGraph’s smart scraper to extract social media links and engagement elements, then feeds that data into Gemini with a focused prompt on presence, content strategy, and community tactics. Finally, it returns both the raw scraped information and the AI-generated, actionable social media insights in a single structured output.
def check_credits():
"""Check available credits"""
try:
credits_info = credits.invoke({})
print(f" Available Credits: {credits_info}")
return credits_info
except Exception as e:
print(f" Could not check credits: {e}")
return None
Available Credits: {credits_info}")
return credits_info
except Exception as e:
print(f"
Could not check credits: {e}")
return NoneThe above function calls the GetCreditsTool to retrieve and display your available ScrapeGraph/Gemini API credits, printing the result or a warning if the check fails, and returns the credit information (or None on error).
if __name__ == "__main__":
print(" Advanced Competitive Analysis Tool with Gemini AI")
print("="*60)
check_credits()
print("n Running AI/SaaS Competitive Analysis...")
ai_results = run_ai_saas_analysis()
run_additional = input("n Run e-commerce analysis as well? (y/n): ").lower().strip()
if run_additional == 'y':
print("n Running E-commerce Platform Analysis...")
ecom_results = run_ecommerce_analysis()
print("n Analysis complete! Check the exported files for detailed results.")
Advanced Competitive Analysis Tool with Gemini AI")
print("="*60)
check_credits()
print("n
Running AI/SaaS Competitive Analysis...")
ai_results = run_ai_saas_analysis()
run_additional = input("n
Run e-commerce analysis as well? (y/n): ").lower().strip()
if run_additional == 'y':
print("n
Running E-commerce Platform Analysis...")
ecom_results = run_ecommerce_analysis()
print("n
Analysis complete! Check the exported files for detailed results.")Finally, the last code piece serves as the script’s entry point: it prints a header, checks API credits, then kicks off the AI/SaaS competitor analysis (and optionally e-commerce analysis) before signaling that all results have been exported.
In conclusion, integrating ScrapeGraph’s scraping capabilities with Gemini AI transforms a traditionally time-consuming competitive intelligence workflow into an efficient, repeatable pipeline. ScrapeGraph handles the heavy lifting of fetching and normalizing web-based information, while Gemini’s language understanding turns that raw data into high-level strategic recommendations. As a result, businesses can rapidly assess market positioning, identify feature gaps, and uncover emerging opportunities with minimal manual intervention. By automating these steps, users gain speed and consistency, as well as the flexibility to expand their analysis to new competitors or markets as needed.
Check out the. All credit for this research goes to the researchers of this project. Also, feel free to follow us on and don’t forget to join our and Subscribe to .
