Service Lifecycle
Last updated
Last updated
To fully understand how Karsa works in the background and how it contributes to optimizing your Google Shopping campaigns, it's important to familiarize yourself with its typical lifecycle. This cycle describes a sequence of automated and user-controlled steps that regularly repeat.
The following diagram and description illustrate the key phases of the service lifecycle:
What happens: Every day, Karsa automatically downloads current performance data for your products directly from your Google Ads account.
Importance: Ensures that all analyses and clusterization decisions are based on the most up-to-date information available.
Conversion backfill: The system also takes into account any retrospective changes in conversion values that Google may make (e.g., due to attribution models or additional conversion assignments).
What happens: You, as a user, define your requirements for product segmentation in the Cluster Designer tool. Here you set:
Parameters for clusterization (e.g., ROAS, Conversion value, Number of conversions).
Limits for the number of clusters and minimum number of conversions in a cluster.
Settings for dynamic product movements and for new products.
Importance: This step gives you full control over the segmentation strategy.
What happens: After saving the configuration in Cluster Designer, you launch the clusterization creation operation. Karsa's algorithms begin analyzing your product data according to the specified parameters and creating optimal clusters.
Duration: This process usually takes 2 to 15 minutes depending on the total number of products, their data distribution, and the range of cluster combinations being explored.
Monitoring: You can monitor the progress of this operation in the Operation Manager.
What happens: After successful completion of the process, Karsa Labelizer automatically creates a new clusterization category (visible in the left menu). This category contains the individual generated clusters as its subcategories.
Significance: Each such cluster represents a proposal for a future Google Shopping campaign.
What happens: Now it's up to you to review and analyze the proposed clusterization and its clusters using the reports that Karsa Labelizer provides (e.g., Clusterizations Overview
, Cluster Detail
).
Decision: If you are satisfied with the proposal and consider it suitable for deployment:
In the detail of the given clusterization category, you switch its state to Run
(Production mode).
A clusterization category in production mode is visually distinguished in the left menu by a green color.
Importance: This step ensures that only structures approved and verified by you go into your real campaigns.
What happens: Once a clusterization is in ProductionRun
mode, Karsa Labelizer automatically exports the relevant Custom Labels
(indicating a product's affiliation to a given cluster) to your Google Merchant Center every day.
Importance: This ensures that your Google Ads campaigns can segment products according to their current cluster assignment.
What happens: For clusterizations in Run
mode, Karsa Labelizer not only exports data but also continuously monitors product performance. Based on current data and set rules for dynamics:
It automatically moves products between individual clusters if their performance profile changes and better matches another cluster.
This process respects your defined limits (e.g., maximum percentage of moved products) to prevent campaign destabilization.
Movement history: All these movements are recorded and available to you in the Product Movement History
report for analysis.
Importance: Ensures that your segmentation remains current and optimized relative to the real performance of products, without you having to make constant manual adjustments.
This cyclical process – from data collection, through design and approval, to automated daily maintenance and optimization – is the heart of Karsa Labelizer and allows you to achieve better long-term results in your Google Shopping campaigns.
More information:
More information:
More information: and