Managing PPC campaigns for a large e-commerce company like Karaca comes with its own set of challenges, especially when it involves Performance Max (PMax) campaigns. As a leading kitchenware and cookware brand with a catalog of over 2,000 products, Karaca faced the dual challenge of increasing revenue while gaining better control over campaign performance and budget allocation, particularly for high-performing products.
To address these challenges, us as the digital marketing team decided to test an AI-driven tool, SMEC, for the first time. This was our first real experience with AI-powered campaign optimization, and we were eager to explore how it could improve our performance. Moving away from traditional category segmentation was a bold step, but we were excited to embrace automation and data-driven decision-making.
Initial Campaign Structure & Challenges
Before introducing SMEC, our PMax campaigns were segmented by product categories, such as,
- Dinnerware
- Cookware
- Small Appliances
While this approach gave us some level of control, it lacked the flexibility needed to allocate the budget dynamically across products with varying levels of performance. With such a large catalog, it was challenging to consistently direct budget to the products that were generating the highest return on ad spend (ROAS).
In the past, I remember when I was working with one of the biggest international agencies, we followed a more manual process of reviewing product ROAS biweekly and adjusting the budget for high-ROAS and mid-ROAS products. This was time-consuming and inefficient, especially with large clients. We had to ensure that top-performing products received a larger share of the budget while also maintaining stability for the rest of the catalog.
Implementing AI-Driven Optimization with SMEC
To overcome these limitations, we integrated Karaca’s Google Ads account with SMEC and completely restructured our PMax campaigns. Instead of segmenting by categories, we transitioned to a performance-based structure, dividing the products into three primary campaigns:
- High Score Campaign – For products with high ROAS and strong performance.
- Mid Score Campaign – For products with moderate ROAS and stable performance.
- Low Score Campaign – For products with lower ROAS and weaker performance.
The key innovation behind this restructuring was the dynamic scoring system powered by SMEC. We created a custom label specifically for SMEC, which continuously evaluated each product’s performance. This allowed products to automatically shift between campaigns based on real-time ROAS and sales metrics. For instance, a product that started in the High Score campaign could be moved to the Low Score campaign within days if its performance dropped.
This dynamic optimization meant we no longer had to monitor and reallocate the budget manually, a task we previously had to do bi-weekly when working with one of the biggest international agencies for very large clients. With SMEC, this process was automated, and we were able to scale our efforts with AI-based decision-making.
Additional Enhancements
In addition to restructuring the campaigns, we also leveraged several features of SMEC to further refine our strategy:
- Asset Group Segmentation: While the campaigns were organized by performance score, we still maintained asset groups by category (such as Small Appliances, Cookware, and Dinner Sets) to ensure that the creatives remained highly relevant.
- Push Strategies for Seasonal Promotions: During sales events like Black Friday, we used SMEC’s aggressive bidding strategies to prioritize discounted products. This ensured better visibility for our key strategic SKUs.
- Exclusion Strategy for Poor Performers: To minimize waste, we implemented an exclusion list to automatically filter out underperforming products. With a catalog of over 2,000 SKUs, it was crucial to eliminate non-profitable products from the campaigns.
- Close Collaboration with SMEC Team: Integrating an AI-powered system like SMEC required constant communication with the SMEC team. We worked together to troubleshoot technical issues related to custom labels, feed management, and Google Merchant Center integrations.
Results & Performance Impact
The AI-driven approach to restructuring the campaigns delivered significant improvements in performance between May 2024 and February 2025:
- 44% increase in ROAS
- 31% increase in revenue
By automatically adjusting campaign budgets based on real-time performance data, we were able to allocate more spend toward high-performing products while reducing inefficiencies and waste on underperforming items.
Conclusion
Implementing SMEC was a groundbreaking experience for us as a digital marketing team, marking our first foray into AI-driven PPC optimization. The ability to dynamically manage product prioritization and budget allocation in real-time was incredibly insightful and proved to be a game-changer for Karaca’s PMax campaigns.
The transition from traditional category-based segmentation to an AI-powered performance-based structure revolutionized the way we managed campaigns. By integrating SMEC, we were able to automate product prioritization, optimize budget allocation, and significantly increase both ROAS and revenue.
This case study illustrates the power of AI in transforming PPC campaign management, reducing manual efforts, and boosting profitability for e-commerce businesses. For brands with extensive product catalogs, adopting an AI-driven approach like SMEC can be a crucial step in scaling PPC campaigns effectively and efficiently.
