Data-Driven Decisions: How GIGA (Gemini Insights Generation & Analysis) Supports Google Ads Campaign Strategy and Business Development

Google tools have long provided us with a wealth of data regarding search queries, their popularity, and seasonality. To obtain this information, solutions such as the Google Ads Keyword Planner or Google Trends are most commonly utilized. These allow specialists to check search volumes, compare the popularity of various topics, and identify seasonal spikes in interest.

In practice, however, analyzing larger sets of keywords can be time-consuming. Furthermore, manual work often leads to focusing on obvious queries, which may result in overlooking phrases that could be equally valuable from a business perspective. This is particularly true for emerging trends or less obvious queries that possess real marketing potential.

This is precisely where AI-driven tools come in, capable of automatically processing large data volumes and transforming them into actionable business insights — GIGA (Gemini Insights Generation & Analysis). This solution automatically analyzes large search engine datasets and enables faster trend detection. It also assists in thematic grouping and finding keywords to be used in marketing strategies.

It is important to note that GIGA is not an official Google product. It is an open-source project shared by the Google Marketing Solutions team on GitHub, demonstrating how Gemini models can support marketing data analysis.

How to Run GIGA? How to use Gemini for keyword analysis?

GIGA was developed as an application operating within the Google Apps Script environment, deployed as a web app. It utilizes Vertex AI to generate insights and connect keywords into clusters using Gemini models. Meanwhile, the Google Ads API (Keyword Planner) fetches keyword ideas, search volumes, and historical data. Consequently, the application combines data retrieval, analysis, and result presentation into a single workflow.

In practice, using GIGA in Apps Script can be described as a four-step process:

1. First, the user launches the web application.
2. Then, they configure the connection to Google Ads and the cloud environment.
3. Next, the system retrieves keyword data and trends.
4. Finally, Gemini analyzes the data and returns organized results.

The repository emphasizes that the outcomes include keyword clusters, insight summaries, and recommendations designed to help brands react more quickly to search trends.

Analysis capabilities and practical application of GIGA

1. Search Trend Analysis Over Time

One of the primary views in GIGA is a chart showing changes in search volume across successive months for selected keywords. This allows you to:

• Observe the seasonality of interest in a product or category.
• Identify growth periods throughout the year.
• Compare the popularity of different search queries.
• Determine the optimal moment to launch a marketing campaign.
  

Practical Example:
If interest in a specific product begins to rise in the spring, the advertising campaign should start earlier to capitalize on the growing demand.

2. Automatic Topic Grouping

Instead of manually organizing hundreds or thousands of keywords, GIGA enables automatic grouping into thematic areas. The tool can:

• Group search queries into thematic clusters.
• Identify the primary areas of user interest.
  

This approach makes it easier to view data from a broader perspective. In practice, this helps with:

• Building a logical structure for advertising campaigns.
• Planning SEO content architecture.
• Creating content marketing strategies.

3. Search Trend Detection

GIGA also allows for identifying shifts in user behavior based on search data. By analyzing large query sets, the tool indicates:

• Rising topics and phrases gaining popularity.
• New user queries appearing in a specific thematic area.
• Areas with business potential that may represent new marketing opportunities.
  

Such insights allow for faster detection of market changes and earlier adjustments to marketing strategies or ad campaigns. 

The tool generates a very broad set of keyword suggestions, allowing for a much more comprehensive view than a standard analysis in the Google Ads Keyword Planner. By entering a few “seed” phrases, we can receive queries that might not appear during manual analysis.

While not every suggestion will directly fit a specific business, valuable phrases with dynamic growth or high search potential are often found among the suggestions. These can reveal interest in products or categories not yet in a company’s portfolio. Such data is a valuable hint for e-commerce stores deciding to expand their assortment or for manufacturers developing new products to meet real demand.

Furthermore, GIGA helps identify keywords whose popularity is beginning to decline. Analyzing these is equally valuable as it helps:

• React faster to market shifts.
• Limit investment in phrases with decreasing potential.
• Gradually shift budgets toward topics with higher growth potential.

4. Analysis and Insight Generation

A key feature of GIGA is the automatic processing of data to generate conclusions. The tool uses AI to analyze large information sets, providing users with not just raw data, but an initial interpretation. Although some suggestions may require additional verification or selection to fit a specific business context, these conclusions significantly speed up the data analysis process. Users receive:

• A summary of major search trends.
• An interpretation of the analyzed data.
• Actionable suggestions for marketing strategies.

5. Advanced Settings in GIGA

The configuration panel allows users to control several parameters that influence data analysis and filtering:

• Gemini Model: Users can choose between Gemini 2.5 Pro (more accurate analysis, higher cost) or Gemini 2.5 Flash (faster analysis, lower cost).
• Temperature: Controls the model’s creativity.
  • Lower: More consistent, predictable keyword grouping.
  • Medium: A balance between stability and discovering new connections.
  • Higher: More exploratory insights, though potentially less organized clusters.
• Top-p: Influences the diversity of query interpretations and the scope of generated insights.
• Min. Search Volume: Sets the minimum number of searches for a phrase to be included, allowing a focus on more popular keywords.

The temperature parameter has the greatest impact on results, as it determines whether the model remains conservative or discovers less obvious dependencies.

6. Google Ads Campaign Support

The Campaigns section is designed to assist in creating advertising communication based on analyzed keyword data. While other modules focus on data analysis, this part focuses on generating ad suggestions. Users can input a brand name, keyword list, and additional instructions. The system then generates multiple ad variants (headlines and descriptions) that comply with Google Ads standards. This allows for the rapid preparation of communication versions for testing or deployment. These suggestions should be treated as a starting point, with final versions tailored to brand strategy and target audiences.

Case Study: GIGA in Action

When planning a Google Ads strategy for a footwear client, GIGA was used to analyze sports shoe keywords. The tool automatically clustered queries, making it easier to identify high-potential areas.

We discovered significant potential in the phrase “gym shoes” (buty na siłownię) and related terms regarding fitness training—a segment previously not considered in the campaign structure. Analysis showed this group combined growing search volume with clear purchase intent, making it a valuable target. Based on this, we adjusted the campaign structure and schedule to match the seasonality of these queries, increasing intensity during peak fitness activity periods.

About author

Renata Jędrasiak Senior SEM Specialist

Senior SEM Specialist at data.rocks with extensive experience in digital marketing. Expert in advanced Google Ads campaigns and analytics (GA4, GTM). Renata specializes in scaling e-commerce and B2B performance across global markets. She builds full-funnel strategies that turn hard data into measurable ROAS growth. 🤘