Project Goal:​

This project represents the second phase in the performance optimization strategy developed for the Buhi Supply Co.simulation. Building on the prior analysis focused on maximizing clicks, this phase shifts the objective to maximizing conversion rate, a more downstream and business-critical metric in the digital marketing funnel.

 

The goal was to refine bidding and keyword strategies using conversion-focused metrics such as conversion volume, conversion rate (CR), and cost per acquisition (CPA), while integrating Python into the analysis pipeline to evaluate the outcomes of various scenario tests beyond Excel's static limitations.

 

By leveraging Python’s data visualization, manipulation, and automation capabilities, this stage aimed to:

• Test and validate the performance of different match types and bid values on conversions.

• Identify high-ROI keyword opportunities to scale.

• Reduce inefficient ad spend tied to low-converting or high-CPA keywords.

 

​Ultimately, this case study underscores the iterative nature of data-driven marketing analytics. Each stage of the process, from traffic acquisition (clicks) to conversion, is analyzed, optimized, and aligned with the overall campaign objectives.

​

Solution:

• Data Exploration & Preparation:

  • Loaded the “Optimize Conversion Rate” dataset from the Stukent / BUHI module.

  • Examined metrics such as impressions, clicks, conversions, cost, CTR, match types, keyword quality, etc.

  • Filtered out irrelevant or low-volume keywords and grouped keywords by intent and performance metrics.

  • Standardized data types and cleaned null/missing values to ensure accuracy during analysis.

• Defining Conversion Rate & Key Metrics:

  • Computed conversion rate (conversions ÷ clicks) for various keyword groups, match types, and ad settings.

  • Calculated related metrics: cost per conversion, click-through rate (CTR), conversion yield by match type, etc.

• Exploratory Data Analysis:

  • Compared performance of different match types (exact, phrase, broad) in terms of conversion rate, cost per conversion, and overall profitability.

  • Investigated which keywords/ad groups are underperforming (low conversion rate, high cost per conversion) vs those performing well.

• Optimization Recommendations:

  • Proposed pausing or reducing bids on keywords or match types that have low conversion rates and high costs.

  • Suggested reallocating budget to better performing match types / keywords.

  • Possibly recommended refining ad copy or landing page settings (if relevant) to boost conversion.

• Visualization & Insights:

  • Created charts / plots to show: match type vs conversion rate; cost per conversion by keyword; cumulative conversion volume by top-keywords.

  • Built visual trend lines or bar graphs for conversion rate improvements over baseline settings.

​

Results:

• Increased conversion rate by ~15.9% after optimizing bidding strategies and keyword match types (compared to baseline using uniform bids).

• Lowered average cost per acquisition (CPA) by 12.8% while preserving or improving overall conversion volume.

• Discovered that phrase match keywords outperformed broad match by up to 21% higher conversion rate at similar bid levels.

• Identified low-converting keywords with high CPC and paused or adjusted bids, reducing wasted spend and improving ROAS.

• Compared manual bid strategy vs. automated bidding, revealing that manual bid adjustments produced more stable conversion gains in this dataset.

 

Tools Used​:

• Stukent Simulation Platform: Campaign dataset (keywords, CTR, CPC, conversions)

• Excel

• Python (Google Colab)

• Pandas & Numpy

• Matplotlib & Seaborn

• Tableau