Frequently Asked Questions (FAQ)
Last updated
English
Last updated
In this section, you will find answers to some of the most common questions regarding the use and functionality of Karsa Labelizer.
A: Karsa Labelizer automatically downloads new performance data from your connected Google Ads account every day. This ensures that all analyses and clusterization decisions are based on the most up-to-date information. The system also takes into account retrospective changes in conversions that Google may make. You can find the operation launch time in the Operation Manager.
A: Currently, it is not allowed to have multiple clusterizations simultaneously in production mode (ProductionRun) within one project. Only one clusterization can be active (in production) at a time. When switching a new clusterization from test state (Test) to production state (ProductionRun), the existing production clusterization (if any) is automatically switched to test state (Test). This ensures that data is always exported to Google Merchant Center only from one, currently selected active strategy.
If you need to implement multiple different production clusterization strategies simultaneously (for example, for completely different product assortments with different data sources or for different markets, such as "Pharmacy" and "Electronics" departments in a large e-commerce store), the solution is to create a new, separate project in Karsa Labelizer for each such strategy. You can then switch between these projects. Each project will have its own active production clusterization.
A: Karsa Labelizer offers several strategies for handling new products that do not yet have performance history. You can configure this in the in the "New Products Strategy" section. The main options are:
Automatic value derivation: The system estimates the performance of a new product based on similar products and immediately places it in a potentially relevant cluster.
Addition to a special category/campaign: New products are placed in an "incubation" campaign for a defined period (e.g., 30 days), where they collect data, and are then categorized according to actual performance.
Placement with zero values: The product is categorized with zero values, which usually places it in a lower-performing cluster until it gains its own history.
More details and recommendations can be found on the [New Products Strategy](../iii.-cluster-designer-creating-clusterization/product-dynamics-management/new-products-strategy.md) page.
A: The process has several steps:
Karsa Labelizer: After switching the clusterization to ProductionRun, Karsa starts exporting Custom Labels to your Google Merchant Center (GMC) daily. Alternatively, it is possible to immediately start generating the XML feed in the Operation Manager.
Google Merchant Center: GMC needs some time to process this data (typically from a supplementary feed). It can take from a few minutes to several hours.
Google Ads: Once the Custom Labels are updated for products in GMC, Google Ads gradually loads this data and starts using it for segmenting your campaigns. There may also be a delay here.
Overall: Expect that from activation in Karsa Labelizer to full propagation and functionality in Google Ads campaigns, it may take several hours to approximately one day. We recommend checking in GMC and Google Ads the day after activation.
A: "Consistency" is a Karsa Labelizer metric that indicates the degree of stability and similarity of product performance metrics within individual clusters of a given clusterization over time.
Higher value means better result: Clusters are more internally homogeneous, their performance is more stable and more predictable for Google algorithms.
Why it's important: It helps you compare the quality of different clusterization strategies. When comparing, it's appropriate to consider clusterizations created with the same number of parameters.
More information can be found in the description of the [Clusterizations Overview](../iv.-analysis-and-optimization/clusterizations-overview.md) report.
A: While the goal is to have products in the right clusters, too frequent or massive movements can destabilize your campaigns in Google Ads. Google algorithms (especially Smart Bidding) need time and consistent data to "learn" how to work with a given segment of products. Constant changes can disrupt this process and lead to a temporary decrease in performance.
Karsa Labelizer therefore offers detailed [settings for product movement limits](../iii.-cluster-designer-creating-clusterization/product-dynamics-management/movement-limits.md), which help find a balance between dynamic optimization and the necessary stability of campaigns.
Do you have another question? Don't hesitate to contact us at support@karsa.ai.