📘 This article explains how AgenticCatalog enriches your product data, what it changes, and what it leaves alone. AgenticCatalog is part of Botify's Activation Suite, available as an option with all Botify plans.
Overview
AgenticCatalog builds its output from two sources:
Your product feed. This is the source of truth for core product data such as IDs, prices, availability, and links. AgenticCatalog never changes these attributes.
Your product pages. Botify uses the crawl data it already collects from your product detail pages (PDPs). This can include descriptions, specifications, structured data, reviews, FAQs, and breadcrumbs. AgenticCatalog does not start a separate crawl.
AgenticCatalog matches each product in your feed to its page, then uses the page content to enrich the feed.
How products are matched to pages
AgenticCatalog matches a product to a page using the product URL in your feed. Enrichment draws on page content only where that match succeeds.
If enrichment results look thinner than you expect, the product URLs in your feed and the URLs Botify crawled are the first thing to compare. Tracking parameters, canonical differences, regional URL variants, and product pages that the crawl did not reach all reduce the content available for enrichment.
What enrichment produces
For products in scope, enrichment can:
Fill empty attributes, such as material, color, or compatibility.
Improve existing values, such as titles and descriptions, so AI systems better understand what a product is.
Generate Q&A pairs, anticipating the questions shoppers ask AI assistants about fit, use case, compatibility, and care.
Which attributes AgenticCatalog can enrich depends on the destination. Google Merchant Center accepts conversational attributes such as Q&A, product highlight, and product detail. OpenAI accepts aggregated review data.
Q&A generation
AgenticCatalog generates Q&A pairs from the content on your product pages, including product details, reviews, and any FAQ content already published there. Each answer is based on information on your page.
👀 Q&A pairs are delivered as a feed attribute. They are not added to your website.
Your Botify Solutions Consultant configures Q&A generation on your behalf, as part of your enrichment configuration. Contact your SC if you have questions about Q&A generation for your feed.
What enrichment does not do
It does not invent information. Enrichment is grounded in your feed and your crawled pages. AgenticCatalog does not add a value that neither source supports.
It does not change your website. Enriched values are delivered in the feeds AgenticCatalog sends to your destinations.
It does not change core product data. Prices, availability, links, and IDs come from your feed and pass through unchanged.
How long enriched values last
AgenticCatalog stores enriched values for 60 days. If you update an enriched attribute in your feed during that period, the stored enriched value is still used until it expires.
To ensure your own value is always used when your feed provides one, set that attribute to enrich only when empty.
Enriched values are tied to the product ID. If a product leaves your feed and returns, AgenticCatalog reuses the stored enrichment while it remains valid.
Controlling enrichment
You control enrichment in two places. Managing the AgenticCatalog Enrichment covers the settings on the feed source:
Product scope: the product categories, price range, and product limit that determine which products are enriched.
Guidance: instructions for tone of voice, brand guidelines, and terms to avoid.
Configuring an AgenticCatalog Destination covers the rules that apply per destination:
Attribute rules: for each attribute, always enrich, enrich when empty, never enrich, or set a static value. Static values are available for some OpenAI attributes.
Destination differences: the same attribute can follow different rules at each destination.
Use never enrich for attributes with legal or brand constraints, or where you want to keep full control of the value.
Reviewing enrichment results
The Enrichment Report shows how many products were enriched, how many empty values were filled, and how many existing values were improved, broken down by attribute.
To review enrichment before applying it to your whole feed, use enrichment testing on a sample of products.
See also: