The Google Trends Agent is an AI agent designed to surface the newest Google Trends in real-time. It can identify emerging topics, analyze their velocity, and provide insights into what is currently capturing the world's attention. This is useful for content creators, marketers, and analysts who need to stay ahead of the curve. For example, a marketer could use this agent as part of their workflow to design a marketing campaign based on a trend in a specific region or city that relates to the product or service they promote.

This agent has several important limitations to be aware of:

• Dataset Constraints: The agent can only access data that exists in the public Google Trends BigQuery dataset, which contains only the top trending terms by region and time period.
• No Open-ended Searches: You cannot ask for trends on specific topics (e.g., "trending terms about AI agents") if they aren't already ranked as top terms in the dataset.
• Regional Limitations: Only regions included in the BigQuery dataset are available.

This diagram shows the detailed architecture of the agents and tools used to implement this workflow.

This agent is a sequential agent composed of two sub-agents that work together to fetch and analyze Google Trends data. The first sub-agent, TrendsQueryGeneratorAgent, generates a BigQuery SQL query from the user's request. The second sub-agent, TrendsQueryExecutorAgent, executes this query to retrieve the trends data and present it to the user. This modular design can be extended with more complex, multi-agent workflows.

• Core Logic: The agent's main logic is defined in google_trends_agent/agent.py.
• Tools: It utilizes tools to query the public Google Trends dataset on BigQuery to fetch trending data. You can explore querying the Google Trends dataset in Google Cloud Console.
• Dependencies: Key dependencies include google-cloud-aiplatform for the ADK and agent engine deployment, google-cloud-bigquery and pandas for data handling.

1. Prerequisites

   • Python 3.11+
   • Install uv
   • A Google Cloud Platform project.
   • The Google Cloud CLI.

2. Installation

   # Navigate to the agent's directory
   cd adk-samples/python/agents/google-trends-agent
   # Install the package and dependencies.
   uv sync --dev

3. Configuration

   • Set up your Google Cloud credentials. You can set these in your shell or create a .env file in the agent's root directory (google-trends-agent/).

     # Authenticate your gcloud account
     gcloud auth application-default login
     gcloud auth application-default set-quota-project <your-project-id>
     
     # Create and populate .env file
     cp .env.example .env

   • Edit the .env file with your specific configuration:

     GOOGLE_CLOUD_PROJECT="<your-project-id>"
     GOOGLE_CLOUD_LOCATION="<your-project-location>"
     GOOGLE_CLOUD_STORAGE_BUCKET="<your-storage-bucket>" # Required for deployment

   • Grant the Agent Engine service account permission to run BigQuery jobs. This is required for deployment.

     # Set your project ID
     export PROJECT_ID="<your-project-id>"
     
     # Get your project number
     export PROJECT_NUMBER=$(gcloud projects describe $PROJECT_ID --format="value(projectNumber)")
     
     # Grant the Agent Engine service account permission to run BigQuery jobs
     gcloud projects add-iam-policy-binding ${PROJECT_ID} \
         --member="serviceAccount:service-${PROJECT_NUMBER}@gcp-sa-aiplatform-re.iam.gserviceaccount.com" \
         --role="roles/bigquery.user"

You can run the agent locally using the adk command in your terminal.

1. Activate the virtual environment:

   poetry shell

2. To run the agent from the CLI:

   adk run .

3. To run the agent from the ADK web UI:

   adk web

   Then select google-trends-agent from the dropdown menu.

The agent can also be deployed to Vertex AI Agent Engine.

1. Ensure Prerequisites: Make sure your GOOGLE_CLOUD_PROJECT, GOOGLE_CLOUD_LOCATION, and GOOGLE_CLOUD_STORAGE_BUCKET environment variables are set correctly in your .env file.

2. Run the deployment script:

   poetry shell
   poetry install --with deployment
   python deployment/deploy.py

   When the deployment finishes, it will output the resource ID of the remote agent deployment and update your .env file with the AGENT_ENGINE_ID. For example:

   Created remote agent: projects/<PROJECT_NUMBER>/locations/<PROJECT_LOCATION>/reasoningEngines/<AGENT_ENGINE_ID>
   

3. Test the remote agent: Once deployed, you can interact with the remote agent by running:

   python deployment/test_deployment.py

   You can type quit at any point to exit.

User: List the top 10 terms in Canada during the past 3 weeks

[TrendsQueryGeneratorAgent]:

SELECT
  term,
  rank,
  week
FROM
  `bigquery-public-data.google_trends.international_top_terms`
WHERE
  refresh_date = DATE_SUB(CURRENT_DATE(), INTERVAL 1 DAY)
  AND country_name = 'Canada'
  AND week IN (
    SELECT DISTINCT
      week
    FROM
      `bigquery-public-data.google_trends.international_top_terms`
    WHERE
      refresh_date = DATE_SUB(CURRENT_DATE(), INTERVAL 1 DAY)
      AND country_name = 'Canada'
    ORDER BY
      week DESC
    LIMIT 3
  )
ORDER BY
  week DESC,
  rank
LIMIT 100

[TrendsQueryExecutorAgent]: The top terms in Canada for the week of 2025-07-13 are:

1. usyk vs dubois
2. man united vs leeds united
3. rashford
4. election loser nyt crossword answers
5. reading vs tottenham
6. blake wheeler
7. france vs germany
8. усик дюбуа

You can customize this agent to fit your specific needs:

• Change Data Source: Modify the agent's tools to pull trend data from different sources, such as social media APIs or other analytics platforms.
• Enhance Analysis: Add new tools to perform more in-depth analysis on the trends, deeper research (e.g. using Google Search as an additional tool) such as sentiment analysis or forecasting.
• Add Notifications: Integrate tools that send alerts via email or Slack when a new trend matching specific criteria is detected.

Use the Agent Starter Pack to create a production-ready version of this agent with additional deployment options. The easiest way is with uvx (no install needed):

uvx agent-starter-pack create my-google-trends-agent -a adk@google-trends

# Create and activate a virtual environment
python -m venv .venv && source .venv/bin/activate # On Windows: .venv\Scripts\activate

# Install the starter pack and create your project
pip install --upgrade agent-starter-pack
agent-starter-pack create my-google-trends-agent -a adk@google-trends

The starter pack will prompt you to select deployment options and provides additional production-ready features including automated CI/CD deployment scripts.