Google Search

The Google Search node fetches real-time Google search results (SERP) for any query, with support for multiple search types, geo-targeting, and URL exclusion.

What does the Google Search node do?

The Google Search node fetches real-time search engine results page (SERP) data from Google for any query. It supports multiple search types (web, news, images, videos, shopping, places, maps, scholar, patents) and returns structured results including organic listings, People Also Ask, related searches, and more.

Common use cases:

• Tracking organic rankings for target keywords across countries and locales
• Analyzing competitor SERP presence and visibility for strategic queries
• Finding top-ranking content on a topic for content research and inspiration
• Monitoring news coverage for brands, products, or industry topics
• Discovering local businesses with geo-targeted Places searches

Quick setup

Configuration parameters

Required fields

Optional fields

What does the node output?

The node returns a JSON string whose structure depends on the Type parameter. For a standard Search, results land in the organic array; News populates news, Images populates images, Maps and Places populate places.

{
  "searchParameters": {
    "q": "site:example.com",
    "gl": "fr",
    "hl": "fr",
    "type": "search"
  },
  "organic": [
    {
      "title": "Page Title",
      "link": "https://example.com/page",
      "snippet": "Description of the page...",
      "position": 1
    }
  ],
  "peopleAlsoAsk": [
    {
      "question": "What is...?",
      "snippet": "Answer...",
      "title": "Source Title",
      "link": "https://example.com"
    }
  ],
  "relatedSearches": [
    { "query": "related search term" }
  ]
}

Usage examples

Example 1: SEO keyword ranking check

Track where your domain ranks for a target keyword in France, page 1.

Configuration:

• Query: best SEO tools 2025
• Type: Search
• Country: France (fr)
• Locale: French (fr)
• Page: 1

Connect the output to a JSON Path Extractor with $.organic[*].link to pull the list of ranking URLs, then to a Conditional node to check whether your domain is present.

Example 2: Brand news monitoring across regions

Watch for press coverage of your brand in two countries by passing the country dynamically.

Configuration:

• Query: 'Draft and Goal' OR 'dng.ai' (use single quotes inside the query value)
• Type: News
• input_gl: connected to a Text Input cycling through us, gb, fr

Run inside a Loop node to iterate over a list of country codes.

Example 3: Local business discovery

Find Italian restaurants in a specific city using geo-targeting.

Configuration:

• Query: restaurant italien
• Type: Places
• Country: France (fr)
• Location: Paris, France

The output places array includes name, address, rating, and CID for each business.

Example 4: Filtered competitor research

Exclude noisy domains while researching a query so the SERP focuses on real competitors.

Configuration:

• Query: seo audit checklist
• Type: Search
• Blacklist URLs: pinterest.com, reddit.com, quora.com

Each blacklist entry is appended as a -inurl: operator at runtime.

Common issues

The node fails with a Serper Dev integration error

Cause: The Serper Dev integration is not configured or the API key is invalid.

Solution: Open Settings > Integrations, locate Serper Dev, and add or refresh the API key from your serper.dev dashboard.

No results returned even though the query is valid on google.com

Cause: Your Country and Locale combination, the Location value, or restrictive blacklist URLs filter every result out.

Solution: Drop the Location first, broaden Country and Locale, and remove blacklist URLs one by one until results reappear.

The node returned a different structure than expected

Cause: The output schema depends on the Type parameter. A News search returns a news array, not organic.

Solution: Check the Type in the node settings and adjust your downstream JSON Path Extractor to match ($.news[*], $.images[*], $.places[*], etc.).

Auto-corrected results do not match my exact query

Cause: Google rewrote the query because Auto correct is enabled.

Solution: Disable Auto correct in the node settings to force the literal query, useful for brand names, rare terms, or product SKUs.

Invalid JSON format error when passing multiple queries

Cause: The query input starts with [ or { but is not parseable JSON.

Solution: Either pass a plain string (single query) or a valid JSON array of strings (e.g. ['seo', 'sea'] rendered as JSON). Validate the upstream node output before connecting it.

Best practices

How does it fit into a workflow?

The Google Search node typically sits at the start of a SEO or research pipeline, feeding a list of URLs or snippets into downstream extraction and analysis nodes.

graph LR
    Input[Text Input keyword] --> Search[Google Search]
    Search --> Extract[JSON Path Extractor organic links]
    Extract --> Loop[Loop over URLs]
    Loop --> Scraper[Web Scraper]
    Scraper --> LLM[LLM summarises content]

Related nodes