Similar Products

Similar Products powers “you may also like” recommendations: given a product a shopper is viewing, it finds other products in your catalog that are close to it — typically substitutes within the same or a related category.

How it works

flowchart TD
    A[Shopper views a product] --> B[Match campaigns\nby product category]
    B --> C[Search catalog for\nclose matches]
    C --> D[Category & price\nweighting applied]
    D --> E[Filters, boosting]
    E --> F[Recommendations shown]

What it does

🔁 Finds substitutes, not complements

Results are drawn from the same neighborhood of the catalog as the source product — the recommendation is “similar to this,” not “goes with this” (that’s Cross-Selling).
⚖️ Category and price awareness

Beyond pure visual/text similarity, you can bias results toward the same category and a similar price range.
🎚️ Sensitivity controls

You can require results to match the source product on specific attributes — e.g. always recommend the same brand, or never mix age groups or genders.
📈 Boosting

Results can be re-ranked by your own business signals — see Boosting.

Configuring in Panel

In Panel, open Site → Campaigns, and open or create a Similar Products campaign.

General settings — Name, Active, Default Campaign toggles.

Methods

Semantic Search — turn on and weight the AI models used for similarity (e.g. Berry Punch Net, Volcanic Water Net).
Color — turn on color-aware matching if a color-capable Net is activated for the site.

Influencers

Price impact — set Price sensitivity (how strongly price proximity affects ranking), Reasonable price difference (the price gap where the penalty starts to bite), and a Scaling factor for how sharply it grows beyond that.
Category impact — choose whether category matching is judged by product type or Google product category, and how strongly it should influence ranking (Category sensitivity).
Only Buyable Products — restrict results to in-stock products.
Age Groups / Gender — toggle to keep recommendations within the same age group or gender as the source product.

Interface controls

Preferred Image, maximum results — same as other services.
Filters — configure which attributes are filterable on the results panel.

Products — restrict the campaign to specific categories via Assigned Products.

Boosting — link one or more Boosting campaigns and set how strongly they influence ranking.

With item grouping on, the source product’s own variants are excluded from its own recommendations, and result variants are grouped.

MCP Tool

Agents can call this service directly as the similar_products MCP tool, which wraps the Similar Products REST operation — its request and response shapes match that operation’s schema.