similarweb-analytics

# SimilarWeb Analytics Comprehensive website and domain analysis using SimilarWeb traffic data. ## Core Capabilities - **Traffic Analysis**: Total visits, unique visitors, traffic trends - **Engagement Metrics**: Bounce rate, pages per visit, average visit duration - **Global Ranking**: Website ranking over time - **Traffic Sources**: Marketing channels (desktop and mobile) - **Geographic Distribution**: Traffic breakdown by country ## API Usage All examples use `ApiClient` from a runtime directory supplied by environment configuration. Set `SIMILARWEB_RUNTIME_PATH` or `UNIVERSAL_AI_SANDBOX_RUNTIME` to the directory that contains `data_api.py`. ```python import os import sys runtime_path = ( os.environ.get("SIMILARWEB_RUNTIME_PATH") or os.environ.get("UNIVERSAL_AI_SANDBOX_RUNTIME") ) if not runtime_path: raise RuntimeError("Set SIMILARWEB_RUNTIME_PATH to the directory containing data_api.py") sys.path.append(runtime_path) from data_api import ApiClient ``` Common parameters: - `domain`: Website domain (e.g., "google.com") - `start_date`: Start date (YYYY-MM). Max 12 months ago - `end_date`: End date (YYYY-MM). Max 12 months ago, default is 1 month ago (most recent complete month) - `main_domain_only`: Exclude subdomains if True (default: False) **Default time ranges vary by API:** - Global Rank, Visits Total, Unique Visit, Bounce Rate: default **6 months** - Traffic Sources (Desktop/Mobile), Traffic by Country: default **3 months** ### Get Global Rank ```python client = ApiClient() result = client.call_api('SimilarWeb/get_global_rank', path_params={'domain': 'amazon.com'}) ``` ### Get Website Visits Total ```python client = ApiClient() result = client.call_api('SimilarWeb/get_visits_total', path_params={'domain': 'amazon.com'}, query={'country': 'world', 'granularity': 'monthly', 'start_date': '2025-07', 'end_date': '2025-12'}) ``` ### Get Unique Visit ```python client = ApiClient() result = client.call_api('SimilarWeb/get_unique_visit', path_params={'domain': 'amazon.com'}, query={'start_date': '2025-07', 'end_date': '2025-12'}) ``` ### Get Bounce Rate ```python client = ApiClient() result = client.call_api('SimilarWeb/get_bounce_rate', path_params={'domain': 'amazon.com'}, query={'country': 'world', 'granularity': 'monthly', 'start_date': '2025-07', 'end_date': '2025-12'}) ``` ### Get Traffic Sources - Desktop Returns breakdown by channel: Organic Search, Paid Search, Direct, Display Ads, Email, Referrals, Social Media. ```python client = ApiClient() result = client.call_api('SimilarWeb/get_traffic_sources_desktop', path_params={'domain': 'amazon.com'}, query={'country': 'world', 'granularity': 'monthly', 'start_date': '2025-07', 'end_date': '2025-12'}) ``` ### Get Traffic Sources - Mobile ```python client = ApiClient() result = client.call_api('SimilarWeb/get_traffic_sources_mobile', path_params={'domain': 'amazon.com'}, query={'country': 'world', 'granularity': 'monthly', 'start_date': '2025-07', 'end_date': '2025-12'}) ``` ### Get Total Traffic by Country Returns traffic share, visits, pages per visit, average time, bounce rate and rank by country. - `limit`: Number of countries to return (default: 1, max: 10) - **Date range limit**: max 3 months (unlike other APIs) ```python client = ApiClient() result = client.call_api('SimilarWeb/get_total_traffic_by_country', path_params={'domain': 'amazon.com'}, query={'start_date': '2025-10', 'end_date': '2025-12', 'limit': '10'}) ``` ## When to Use Invoke APIs when users mention: - Domain names: "google.com", "amazon.com" - Traffic queries: "traffic", "visits", "visitors" - Ranking queries: "rank", "ranking", "how popular" - Engagement queries: "bounce rate", "engagement" - Source queries: "traffic sources", "marketing channels" - Geographic queries: "countries", "geographic" - Comparison queries: "compare", "vs" ## Data Limitations - Historical data: max 12 months - Geography: worldwide only - Granularity: monthly only - Latest data: last complete month ## Important: Save Data to Files API calls may fail mid-execution due to credit depletion. **Always save all retrieved data to files immediately** to avoid data loss and prevent redundant API calls.