In this article, we explain the image similarity score and how you can use it.
The image similarity score
Besides the image similarity overview, there are also 3 metrics that can be added to your widget visualizations:
- Image similarity score: measures how many images in your product carousel are rated as equal. The percentage is calculated as all equal images divided by the total images in the carousel. For example: if 4 out of 5 images are equal, the score is 80%.
- Image similarity score first image: measures if the first images in the product carousel is rated as equal. Secondary images are excluded from this check. For example: if the first image is equal, the score is 100%.
- Image similarity score secondary images: measures how many secondary images in the product carousel are rated as equal. The first image is excluded from the check. For example, if 3 out of 4 images are equal, the score is 75%.
To create widgets that visualize these 3 metrics, you:
- Navigate to ‘Widgets’;
- Click the ‘+’ icon on the top of the screen;
- Select KPI visualization, or the line or bar chart;
- Select the image similarity score, the image similarity score first image or the image similarity secondary images, which you can find in the content metrics;
- Select the aggregation function ‘average’.
Note: you can use these metrics in bar charts, line charts or a data table. This makes it easier for different people in your company to use the data in the most efficient way. For example, data tables give an instant overview of what images are used on what retailer, whereas line charts show improvements over time.
Combine the image similarity score with the DCC score
If you need more information on how to use the DCC score, read this article first.
Combining the image similarity score with the DCC score, gives you a complete overview. For example, if your content score includes that the product pages needs to have a minimum of 6 images, the image similarity score shows you the quality om the images on the product page.
