⏳ Duration: 5 minutes
🤖 Complexity: medium
You can adjust how several recommendation types are calculated, directly from your Aqurate account. No code or API calls needed.
👉 Go to Personalize > Settings, or open Advanced settings directly.
⚠️ Some of these settings can impact performance. Change one setting at a time and check the results in your Dashboard before adjusting anything else.
Each block has its own Save button. Changes apply after the next model refresh, which can take up to 24 hours.
👉 Recommended for you
Endpoint: /user
Personalized recommendations based on a user's browsing and purchase history. Two settings control when a product is taken out of rotation:
Maximum number of orders: how many times a user can buy a product before it stops being recommended to them. Set a higher value for products people buy repeatedly, like groceries or consumables.
Maximum number of views: how many times a user can view a product before it stops being recommended to them. Set a higher value for considered purchases, like furniture or electronics, where people come back to the same product several times.
⭐️ Bestsellers
Three recommendation types show your best-selling products. Each one has the same two settings.
Recommendation | Endpoint | What it shows |
Popular now |
| Best-selling products across your whole store |
Popular in category |
| Best-selling products in a category (category pages) |
Popular in same category |
| Best-selling products from the same category as the product being viewed (product pages) |
Method: decides how products are ranked.
volume: number of units sold
value: revenue generated
balanced: a mix of units sold and revenue
distinct buyers: number of different customers who bought the product
most viewed: number of product views
Lookback window: how many days of data are used to calculate the ranking, from 0 to 365. A short window reacts faster to trends and seasonality. A long window gives more stable results, which suits smaller stores or slower-moving catalogs.
💡 Example: A fashion store running a seasonal collection might use volume with a 14-day window on the homepage. A store selling high-value items might prefer value or distinct buyers with a longer window, so a few large orders don't distort the ranking.
🎁 Smart Bundle
🛍️ Smart Bundle is available only for stores on Shopify.
Endpoint: /item/bundle
Endpoint: choose which recommendation logic builds the bundle: cross-sell (products frequently bought together), substitutes (alternatives from the same category) or similar (closest matches from any category).
Discount: set the discount type and value for the bundle.
⬆️ Upsell to
Endpoint: /item/up-sell
Fallback to cheaper items: when there aren't enough more expensive alternatives for a product, turn this on to fill the widget with cheaper items instead of leaving it short or empty.
🖥️ Display settings
At the top of the Advanced settings page, you can also set the currency in which to display the revenue: RON, EUR, USD or HUF.
Pick the option you want and click Save.
🎯 Want to prioritize or hide specific products?
These settings change how recommendations are calculated. To pin, boost, weaken or hide specific products, use Merchandising instead.