pip install google-ads pandas matplotlib

from google.ads.google_ads.client import GoogleAdsClient
import pandas as pd

# Load Google Ads client using OAuth2 credentials
client = GoogleAdsClient.load_from_storage("path_to_google_ads.yaml")

# Define the customer ID (replace with your Google Ads account ID)
customer_id = "INSERT_YOUR_CUSTOMER_ID"

# Initialize Google Ads service
ga_service = client.get_service("GoogleAdsService")

# Query to get campaign performance data (impressions, clicks, conversions)
query = """
    SELECT
        campaign.id,
        campaign.name,
        metrics.impressions,
        metrics.clicks,
        metrics.average_cpc,
        metrics.conversions,
        metrics.cost_micros
    FROM
        campaign
    WHERE
        segments.date DURING LAST_7_DAYS
"""

response = ga_service.search_stream(customer_id=customer_id, query=query)

# Process the response into a Pandas DataFrame
data = []
for batch in response:
    for row in batch.results:
        data.append({
            'Campaign ID': row.campaign.id,
            'Campaign Name': row.campaign.name,
            'Impressions': row.metrics.impressions,
            'Clicks': row.metrics.clicks,
            'Avg CPC': row.metrics.average_cpc / 1e6,  # Convert micros to currency
            'Conversions': row.metrics.conversions,
            'Cost': row.metrics.cost_micros / 1e6  # Convert micros to currency
        })

# Create DataFrame
df = pd.DataFrame(data)
print(df)

import matplotlib.pyplot as plt

# Calculate additional metrics (e.g., Conversion Rate and Return on Ad Spend)
df['Conversion Rate'] = df['Conversions'] / df['Clicks']
df['ROAS'] = df['Conversions'] * 10 / df['Cost']  # Assume $10 per conversion

# Plot the performance metrics
plt.figure(figsize=(10, 6))

# Plot CTR vs Conversion Rate
plt.subplot(1, 2, 1)
plt.scatter(df['Impressions'], df['Conversion Rate'], color='blue')
plt.title('CTR vs Conversion Rate')
plt.xlabel('Impressions')
plt.ylabel('Conversion Rate')

# Plot ROAS vs Cost
plt.subplot(1, 2, 2)
plt.scatter(df['Cost'], df['ROAS'], color='green')
plt.title('ROAS vs Cost')
plt.xlabel('Cost')
plt.ylabel('ROAS')

plt.tight_layout()
plt.show()

from google.ads.google_ads.errors import GoogleAdsException

# Function to adjust campaign bids based on performance
def adjust_campaign_bids(campaign_id, target_cpc):
    campaign_service = client.get_service("CampaignService")
    
    campaign = campaign_service.get_campaign(campaign_id)
    campaign.cpc_bid_micros = target_cpc * 1e6  # Convert to micros
    
    try:
        campaign_service.mutate_campaigns(
            customer_id=customer_id,
            operations=[{
                'update': campaign,
                'update_mask': client.get_type('FieldMask').paths.append('cpc_bid_micros')
            }]
        )
        print(f"Bid for Campaign {campaign_id} adjusted to {target_cpc}")
    except GoogleAdsException as ex:
        print(f"Error adjusting bid for Campaign {campaign_id}: {ex}")

# Automatically adjust bids for campaigns with high ROAS
for _, row in df.iterrows():
    if row['ROAS'] > 3:  # Example threshold: ROAS > 3
        adjust_campaign_bids(row['Campaign ID'], target_cpc=1.50)  # Adjust to new CPC