Skip to main content

Enriching Product Data with Google Books and the AI Agent

A thin, half-empty product listing is one of the fastest ways to lose a sale. StoreLinkr's Enrich product data tool gives every product in your PIM a one-click path to a fuller, better-written listing — pulling in bibliographic data for books and, wherever it's configured, handing everything to your AI Agent to turn into an optimized title, description and attribute set.

This tutorial walks through exactly what happens when you click that button, using two scenarios: a standard product and a book.

How it works, in short​

  1. Open any product in the PIM Editor and click Enrich product data (the star icon, top right).
  2. StoreLinkr checks the product's EAN. If it looks like an ISBN (a 13-digit code starting with 978 or 979), it automatically queries the Google Books API for that title.
  3. If your AI Agent is enabled, StoreLinkr hands everything it has gathered — your existing product data, plus any book data it just found — to your configured AI engine, and asks it to write an optimized name, short and long description, and a full attribute list.
  4. You get an Enriched product data preview: a side-by-side table of every field, old value next to new, with changes highlighted so you can see exactly what would happen.
  5. Nothing is saved until you click Accept and apply. Click Close instead and the product is untouched.

The same button and the same review screen handle both scenarios below — StoreLinkr decides which data sources apply automatically, based on the product's EAN.

Before you start

Enrichment needs a product with an EAN filled in — without one, StoreLinkr has nothing to look up and shows "No EAN found to enrich." Use the Data Quality dashboard to find products that are missing one.

Reading the preview screen​

Whichever scenario you're in, the review table works the same way:

ColumnWhat it shows
FieldName, Short Description, Long Description, EAN
Current ValueWhat's on the product right now
New ValueWhat enrichment found — highlighted green if it differs from the current value

Below the main table, a nested Attributes table shows every technical attribute (facet) side by side:

  • Green — a new attribute, or an existing one whose value changed.
  • Yellow, "Will be removed" — an attribute that exists on the product today but wasn't part of the new data.
  • Unchanged attributes are listed too, greyed out, so you can see the full picture.

If enrichment found new images (a book cover, for example), they appear at the bottom for comparison against the product's current image.

Your EAN itself is never overwritten by enrichment — it's only ever used as the lookup key.

Scenario 1: Enriching a standard product​

Most of your catalog won't have an ISBN, so there's no external bibliographic database to pull fresh facts from. For these products, the AI Agent isn't optional — it's the entire value of the enrichment step. Without it, clicking Enrich product data on a standard product mostly just reflects back what's already there.

The setup: imagine a product imported from a supplier feed with a generic name, a one-line description, and only a couple of technical specs — the kind of listing the Data Quality dashboard would flag as "short description."

  1. Configure the AI Agent first. Go to Settings > AI and:

    • Toggle AI Agent Enabled to ON.
    • Choose an engine (Gemini, OpenAI or Claude) and enter your API key.
    • Fill in your Tone of Voice, Preferred Language, Target Audience and Keywords — this is the context the AI uses to write in your brand's voice, not a generic one.

    See How to configure AI settings for the full walkthrough of this page.

  2. Open the product in the PIM Editor and click Enrich product data.

  3. StoreLinkr builds a prompt from everything it knows about the product — current name, descriptions, brand, and existing attributes — and sends it to your configured AI engine, asking for an SEO-optimized name, descriptions and a complete attribute list.

  4. Review the preview. For a thin listing, expect to see:

    • Name rewritten into something more descriptive and search-friendly.
    • Short and Long Description filled in with persuasive, on-brand copy — highlighted green since they were previously empty or minimal.
    • New rows in the Attributes table for specs the AI was able to infer from context, alongside your existing attributes carried through unchanged.

  5. Click Accept and apply to save the new data to the product, or Close to discard it and keep the product exactly as it was.

Scenario 2: Enriching a book​

Books are the one product type StoreLinkr can enrich with hard bibliographic facts before the AI ever gets involved. If a product's EAN is a valid ISBN-13 (starts with 978 or 979), enrichment automatically looks it up.

The setup: a physical book imported with just a title and an EAN, e.g. 9781234567897.

  1. Open the product and click Enrich product data — there's no separate "book mode" to switch on; StoreLinkr recognizes the ISBN prefix on its own.

  2. Behind the scenes, StoreLinkr queries the Google Books API with that ISBN. Google Books itself aggregates data contributed by publishers, libraries and booksellers worldwide, so a single lookup typically returns:

    • Title and description
    • Author, publisher, publish date and page count (added as new attributes)
    • Cover art

  3. If your AI Agent is also configured, StoreLinkr doesn't stop there — it feeds this freshly gathered book data, together with your existing product record, into your AI engine and asks it to write an SEO-optimized title and description in your configured tone of voice and language, while keeping the factual attributes (author, publisher, page count) intact.

    If the AI Agent isn't configured, you still get the raw Google Books data merged onto the product — just without the rewritten, SEO-optimized copy.

  4. Review the preview. For a book, expect to see:

    • Name matched (and, if AI is enabled, polished) against the title Google Books returned.
    • Long Description filled with the book's synopsis — rewritten by the AI if configured.
    • New attribute rows for Author, Publisher, Publish date and Page count.
    • A cover image in the Images row, ready to compare against whatever image (if any) the product currently has.

  5. Click Accept and apply to save, or Close to walk away without changing anything.

Good to know​

  • Enrichment always shows you the preview first — it never silently overwrites a product. Accepting applies every change in the table at once; there's no per-field accept/reject, so review the table before confirming.
  • The same underlying action is available via the API (POST /api/portal/products/enrich-product/{uuid}) if you want to trigger enrichment as part of a larger automated import workflow — see API Authentication to get started.
  • Run enrichment from the Data Quality dashboard's filtered product lists to work through your catalog's weakest listings first, rather than enriching products at random.