Adding Parameters to Products#

Adding Parameters to Products is a Feed Image Editor feature that uses AI to analyze products’ main images and automatically generates data from them. The resulting parameters are written directly into the feed. This improves product positions in search and filtering, especially in the era of agentic commerce, where detailed product context is key to success.

How it works#

If your feed is missing important data, Feed Image Editor can extract visual product properties (e.g., color, material, texture) from the main images and convert them into text parameters in new AI elements.

The feature is located in the settings of a specific template, on the AI Feed Enrichment tab. You turn it on simply with a toggle button. Parameters are generated for the specific product query selected in the template. If your template is applied to a product query for, say, a specific type of footwear, parameters will be generated exclusively for those products. The whole process runs dynamically in the background, and processing time depends on the number of product images in the query.

Structure of the generated data

The generated data is written into Mergado in two main elements: AI_ATTRIBUTES and AI_VISUAL. These are further split into specific subelements.

  • AI_ATTRIBUTES – contains subelements for product properties. For example, product color, material (e.g., plastic), surface texture, or style.

  • AI_VISUAL – contains subelements of the image’s own visual parameters. It records, for example, the background type (white), image type (photograph), viewing angle (three-quarter view), number of products in the photo (one product), or overall image quality.

For each subelement you will also see a confidence value in the range of 0 to 1 (e.g., 0.8). It expresses how confident the model is about that particular parameter.

Importantly, you always have full control over the data and can intervene and edit it manually at any time in the Mergado interface. The result is an enriched feed for more effective advertising without manual work.

What it’s useful for#

  • Quick feed enrichment – automatic generation of parameters from images replaces hours of manually looking up and rewriting product properties.

  • Product listing – on selected price comparison sites and marketplaces, certain product categories require parameters without which you cannot get listed at all.

  • Agentic commerce – a richly enriched feed gives shopping AI assistants the precise context they need to recommend your product.

What you need#

  • Products must have an element in the feed filled with the URL of the main image.

  • The feature works with images that are available online; local paths are not supported.

FAQ#

Where exactly can I find the generated parameters?#

The analysis results are saved as new subelements under the main AI_ATTRIBUTES and AI_VISUAL branches directly in Products in Mergado.

How can I use the confidence score?#

The numeric value from 0 to 1 helps you automate quality control. In Mergado’s rules, you can set parameters with high confidence (e.g., above 0.85) to be approved for advertising right away, while filtering out those with a lower score for quick manual review.

Will the feed update when new products are added?#

Yes. The feature is fully dynamic. As soon as new items are added to your product query in the template, the extension analyzes their main images in the background and adds the parameters.

Can I edit the generated AI data before sending it to advertising?#

Yes. AI data works like regular elements in Mergado. You can apply further rules to it, rewrite it, or edit the values manually as needed.

Backward AI Feed Enrichment Visual Similarity Forward