When you create a new recommendation, you can define the page type, the recommendation type, the inclusion and exclusion filtering rules and the visual settings. However, you have access to a wide range of advanced settings. You can access the advanced setting by clicking Advanced Settings at the Final Touches stage.

Alternatively, once a new recommendation has been created, you can click the recommendation and you will automatically land on the menu including the advanced settings.

General settings

Recommendation title is the publicly visible title of the recommendation and not to be mixed with DivID.

Enabled-switch enables and on the contrary, disables individual recommendation and allows editing each recommendation slot independently in preview-mode when a recommendation is disabled. Displaying recommendations publicly requires also enabling all recommendations under Account-menu, which works as a global on/off switch for all on-site recommendations.

Div ID is the unique technical ID of the recommendation slot. This is not visible for customers and is used as an id for the recommendation placeholder. ID needs to be unique as it is used to tie a recommendation placeholder on the site to settings as configured on the Nosto admin panel. DivID is editable, but if reconfigured, make sure you have an equivalent recommendation placeholder code on your site as otherwise Nosto can’t match recommendation configurations to a placeholder.

Min amount is the Minimum amount of products required to recommendation to display itself. If Nosto has less than the minimum amount of products to display in its index, based on the set recommendation rulesets and filters, the recommendation will automatically hide itself if Min amount target is not met. Essentially Min amount is a design setting that allows better control over visibility of recommendations. For example, if a physical space reserved for a recommendation is designed for three products and if Min amount is set as 3, the recommendation will hide itself unless Nosto for any reason can’t recommend 3 or more products.

Max amount defines the number of products displayed in a recommendation. Essentially it’s a similar design setting to min amount, but it limits how many products Nosto will populate in a recommendation. Using the same example, if a physical space reserved for a recommendation is designed for three products and if Max amount is set as 3, Nosto will only display maximum of three products in a recommendation. Note that if a recommendation design is structured as a carousel, it sets how many products are displayed in a carousel.

Recommendation settings

Recommendation settings differs depending on the algorithm selected. You can find a glossary of the algorithms available here.

Best Sellers

Three scoring modes can be selected:

  • Most Views: Ranks products according to the number of views gained usually displayed as: “Popular items” or Trending products”. Works well for highlighting recently added or otherwise often viewed products, which are not necessarily yet best sellers
  • Most Buys: Ranks products according to the quantity of items sold usually displayed as: “Bestsellers” by putting the best converting items to the limelight.
  • Most Buyers: Ranks products according to the number of check-outs, ignoring the number of sold quantity. Use most buyers instead of buys in case you sell products which are often bought in large quantities per order.

Time period can be adjusted according to your needs. Read more about the Time Periods here.

Product based Cross Selling, Cart Based recommendations, Browsing History Related and Order Related Products

  • Relationship Score Based: Intelligent point-based logic, giving more weight to items other visitors have bought or added to their cart with the current item, but also tracks views together albeit giving a lower score for views. Usually displayed as: “Customers who viewed this also liked”. Suits well for a recommendation suggesting supplementary products for all stores despite the traffic and sales volume.
  • Viewed Together (smart): Displays items other visitors have viewed close to currently viewed item. Usually displayed as: “Customers who viewed this also viewed”. Works well for a recommendation displaying alternatives.
  • Viewed Together (plain): displays items generally viewed together during the visit o Usually displayed as: “Customers who viewed this also viewed”.
  • After Viewing This Finally Bought: displays items visitors have gone on to buy after viewing the item. Usually displayed as: “Customers who viewed this finally bought”
  • Bought Together: items repeatedly bought together with the current item. Usually displayed as: “Frequently bought together”. Great for recommendation suggesting supplementary products, but suitable only for big shops with significant amount of conversions across the whole selection as the scoring mode requires a lot of data.

Free Shipping Recommendations

Similarly to the algorithms above, the Relation Type can be defined.

Visibility settings defines the cart value and when the recommendation is displayed.

  • From: Minimum cart value. Recommendation is displayed when the overall cart value exceeds the given value
  • To: Maximum cart value. Recommendation is hidden when the overall cart value exceeds the given value.

In the screencap above, recommendation would be only visible when cart value is between 60 to 100.

Personalized recommendations

While displaying personalized recommendations, you can include or exclude products already in cart.

Browsing History

It defines the number of products related to the browsing history to be displayed in the recommendation.

Cherry-pick recommendations

In the input field can be defined the products to be displayed in the recommendation.

Note: make sure the products are all in stock.

Filter settings

Inclusion and Exclusion products filtering

Filters featuring include and exclude rules are settings which effectively limit displayed products to a more narrow subset based on product attributes and use-cases. Settings are slightly different between different recommendation types and covered as separate chapters in this article. Read more about the inclusion and exclusion filtering rules here.

Advanced Dynamic Filtering

Dynamic filtering is an advanced feature and enables a possibility to create hand-picked and exact recommendation pairing or merchandising based on product attributes. Read more about the Advanced Dynamic Filtering here.

Fallbacks

Fallbacks produce product recommendations if no results has been found in a given context for the primarily set recommendation type. Read more about fallbacks here.

Set-up

Fallbacks can be applied to any recommendation. Select a target recommendation and browse down to the Fallbacks section.

You can create a falllback recommendation similarly to a regular Nosto on-site product recommendation slot. You will observe a notification informing that all changes are saved before creating a fallback recommendation. You can pursue and click “OK”.

Define the settings according to your needs.

The newly created fallback recommendation has been created and you can view it via the on-site product recommendations menu.

Multiple fallback recommendations can be applied to a recommendation slot and you can create a new or additional fallback recommendation by clicking “Create a new Fallback Recommendation” or via the general settings of a selected slot. There is no limitation in the number of fallback recommendations.

You can change the priority order of the different fallback recommendations via the primary recommendation slot general settings menu by dragging and dropping.

Note:

  • If the target recommendation is disabled, all of the possible fallback recommendations will be by default disabled.
  • On the contrary, if a fallback recommendation is applied to a primary recommendation slot which is disabled, when the primary recommendation is re-enabled, all fo the fallback recommendations will be enabled simultaneously.
  • Performance of fallbacks will be displayed separately in the analytics and will not include primary recommendation’s statistics.
Did this answer your question?