AI Discovery on Google

AI discovery is an add-on for a Google channel. It adds four product fields Google's AI shopping surfaces read: highlights, an enriched description, structured product details, and questions and answers.

Updated November 6, 2026

Google's AI shopping surfaces read product data from the same Merchant Center feed you already send, and Google's specification has fields built for them: short benefit lines, structured specifications, and questions with answers.

AI discovery is Renable's name for switching those fields on. It's an add-on chip on a Google channel's Used for row, and turning it on adds four recommended fields to what the channel's Fields tab asks for.

What it can and can't promise#

Google says these fields help customers discover information about your products across AI-driven surfaces. It publishes no figure for how filling them affects visibility, ranking or traffic, and neither does Renable. These are the fields Google's AI shopping surfaces read. Google decides what it actually shows.

  • All four fields are optional, and there's no deadline.
  • Renable has no "AI share" number to show you. Google has announced AI performance insights in Merchant Center, rolling out to the US, Canada, Australia, India and New Zealand.

The four fields#

Feed fieldWhat it carriesWhere the content comes from
product_highlightShort selling-benefit linesThe Write product highlights prompt (AI)
descriptionYour product description, already required for a Google feed. AI discovery asks for a fuller, factual oneThe Write enriched descriptions prompt (AI)
product_detailYour product's specifications, as attribute name / attribute value pairsYour catalog's own product attributes, with no AI and no credits
question_and_answerQuestions with their answersText content you write yourself on the project's Creative tab

Google's documented limits:

  • Highlights are 1–150 characters each, with a minimum of 2 and a maximum of 100 per product. Google recommends 4–6. Renable exports up to 10 per product, its limit for a field with Allow multiple values on. (Google's product_highlight reference)
  • Product details are an optional section name plus a required attribute name and value. Google documents no limit on how many a product may carry. Renable caps each field at 100 blocks per product by default, and you can change that on the field. (Google's product_detail reference)
  • Questions and answers allow up to 30 pairs per product, up to 1,000 characters per question and per answer, and 10,000 characters combined. (Google's question_and_answer reference)

Google also asks you not to repeat the same information across these fields.

Renable doesn't offer related products, popularity rank, document links or variant options, because your product data has no reliable source for them.

Turning it on#

The fastest route is to ask Aimée:

"I want my products to be found when people ask AI what to buy."

She'll turn on the add-on, create the prompts and add the feed fields, telling you first that an active prompt spends credits on every product in scope. You still activate and publish the prompts yourself, and you still write your own questions and answers.

To do it by hand:

1. Turn on the add-on#

Open the channel, go to the Settings tab and find the Used for card. On the Google row, switch on the AI discovery chip. This applies at once, with no publish step.

Your feed doesn't change yet. The Fields tab now counts the AI discovery fields among its recommended ones, each with a note saying where its content comes from.

2. Create the content prompts#

On the project's Creative tab, click Create prompt and pick:

  • Write product highlights (AI discovery)
  • Write enriched descriptions (AI discovery)
  • Write use cases (AI discovery), if you want it. This one writes short "good for…" phrases, which go into the same product_highlight field as the highlights.

These are ordinary text prompts. Their output is tagged ai-highlights, ai-description or ai-use-cases, so your existing ad copy doesn't change. An enriched description reaches more than this channel, though: the description column that Google, ChatGPT, Microsoft, Pinterest, Reddit, Perplexity, Prisjakt, PriceRunner and Criteo channels start with prefers it, with or without the add-on.

Like any prompt, they don't run until you activate and publish them. See Working with prompts.

3. Write your questions and answers#

Questions and answers are yours to write, because an invented answer about a product's warranty or fit is the last thing you want an AI shopping surface repeating.

  1. Open the project, go to the Creative tab and find the Project content section.
  2. Click Add text content.
  3. Add a headline, which is the question, and an excerpt, which is the answer. Add one pair per item.
  4. In the same dialog, add the tag faq in the Tags field. Without it, the item never reaches the question_and_answer field.

Every product in the project carries the same set, so keep to questions that are true of your whole range. A good start is what your support inbox actually gets: delivery times, sizing, materials, returns, care. Google asks you to keep out pricing and anything time-sensitive.

4. Add the fields to the feed#

Back on the channel's Fields tab, a notice at the top counts the recommended fields that aren't set up yet. Click Show on it, and each appears as a suggestion row with an Add button. Adding one creates a working field with its content filled in. Then Publish the channel.

product_detail and question_and_answer are nested fields, which write a block per item. See Channel fields and placeholders.

What this costs#

The prompt-written fields cost credits, charged per product each time a prompt runs (see What uses my credits). product_detail and your questions and answers cost nothing.

If your catalog is large, activate one prompt first and look at the results on the Content page before you activate the next. Creating a prompt is free; running it across a whole catalog is what costs.

Checking it worked#

  • Fields tab: each AI discovery field shows a fill rate from the channel's last run. A low one usually means the prompt hasn't been activated or published, or hasn't reached every product yet.
  • Preview tab, Source view: check one product's entry for repeated highlight lines, and nested blocks for details and questions.
  • Aimée: ask "how AI-ready is my catalog?" on the channel for the gaps per field, with a real product as evidence.

Unapproved AI discovery content ships to your feed like all other AI content. Approval is an editorial lock, not a gate. To read it before it goes out, review it on the Content page, filtered to unapproved content.

FAQ#

Does this work on Meta or TikTok?

No. These are Google Merchant Center attributes, so the chip only appears on the Google row.

Will turning this on break my existing feed?

No. Every field it adds is recommended, so it can't block a publish, and your existing ad copy stays as it is.

My products have no attributes, so product_detail is empty. Is that a problem?

No. A product with no attributes just doesn't get that field. To fill it, map more of your feed's columns as attributes on the catalog's Mapping tab.

Related articles#