The following table shows Optimizely Email Product Recommendations strategies.
|Abandonedbasket products||Shows products that were added to the basket but not purchased. Only baskets abandoned within the last 90 days and products currently available are considered.|
|Best sellers by conversion (last 90 days)||Show best sellers based on conversion rate from the last 90 days, and the categories that the customer viewed.|
|Best sellers by revenue (last 30 days)||Show products that generated the most revenue on site in the last 30 days. This strategy also takes categories into account; that is, it gives bias to products that sold well, and are in the same category with which the customer has most interacted.|
|Best sellers by revenue per product view (last 90 days)||Show products that generated the most revenue per numbers of views in the last 90 days. This strategy also takes categories into account; that is, it gives bias to products that sold well, and are in the same category with which the customer has most interacted.|
|Best sellers by units (last 7 days)||Show best sellers based on number of products sold in the last 7 days.|
|Best trending (last 30 days)||Show products that are gaining popularity; that is, whose unit sales are growing the most in the last 30 days.|
|Best trending based on previous purchases and views||Show products that are gaining popularity; that is, whose unit sales are growing the most, based on the customer's previously viewed products and purchase history.|
|Co-purchased products based on previous purchases||Show products that other customers also ordered when purchasing the same product as the customer, and that are in any category from which the customer has purchased.|
|Co-purchased products based on wishlist||Show products bought together with products the customer has in their wish list.|
|Co-viewed products based on wishlist||Show products viewed together with products the customer has in their wish list.|
|Cross-sells from previous purchases and views||Show products bought together with products the customer bought and viewed in the past.|
|New Products (last 7/14/30 days)||Show products recently added to the site (in the last 7, 14 or 30 days).|
|New products from purchased categories (last 120 days)||Show new products from categories from which the customer previously made a purchase (in the last 120 days).|
|Newly discounted products based on viewing history||Show recently discounted products related to products that the customer previously viewed.|
|Newly discounted products from viewing history||Show recently discounted products that the customer previously viewed.|
|Popular co-purchased products||Show popular items based on the customer’s purchased products (biased towards the most recently bought product).|
|Popular products based on previous purchases and views||Show popular items from categories from which the customer recently made a purchase. If not enough are found from purchased categories, it shows popular products from categories that the customer recently viewed.|
|Popular products based on recent views||Show popular products based on the customer’s recently viewed products.|
|Popular products from categories related to previous purchases||Show popular items from categories that are related to categories from which the customer made a purchased.|
|Popular products from co-purchased and purchased categories||Show products that other customers bought together with the same product as the customer, and are in the same or related category from which the customer made a purchase.|
|Popular products from co-purchased categories||Show products that other customers bought together with the same product as the customer, and are in a related category from which the customer made a purchase.|
|Products from daily trigger campaigns||Used only with Triggers. Show products from daily trigger campaigns, such as Targeted discounts, High product interest, Post purchase, Low-in-stock abandoned-basket.|
|Products from in-session trigger campaigns||Used only with Triggers. Show products from in-session trigger campaigns, such as Abandoned basket, Abandonedbrowse, Abandoned checkout.|
|Recently purchased products||Show products the customer purchased, the most recent first.|
|Recently viewed products||Show products the customer recently viewed.|
|Recommendations based on abandoned basket||Show related products based on the customer’s abandoned basket items (biased towards the most recently abandoned product).|
|Recommendations based on the customer's popular views||Show popular products from categories that the customer browsed most.|
|Recommendations based on daily trigger campaigns||Used only with Triggers. Show related products based on the daily trigger used, such as Targeted discounts, High product interest, Post purchase, Low-in-stock abandoned-basket.|
|Recommendations based on in-session trigger campaigns||Used only with Triggers. Show related products based on the in-session trigger used, such as Abandoned basket, Abandoned browse, or Abandoned checkout for the current session.|
|Recommendations based on recent basket additions||Show popular items based on products the customer added to basket.|
|Recommendations based on specified product refCodes||Show recommendations based on what other customers purchased or viewed with the specified products. To use this strategy, a client needs to fill the value of the refCode variable in Mail recommendations links with product refCodes (for example, from a customer's order) and then the recommendations returned are based on those refCodes.|
|Recommendations from the category specified in Hints||Show products from a particular category, which needs to be specified in the Hints section using the hint category.|
|Recommendations from categories of purchased products (last 365 days)||
Show products from categories customer has purchased something within last 365 days.
In Germany, you can send email marketing advertisements to recent customers without explicit consent according to §7 Abs. 3 UWG, as long as the advertised products are "similar," relating to products the customer has already purchased within the past 365 days. Check your local laws.
Personalized and merchandised recommendation strategies
You can use over 30 different strategies in Email Product Recommendations, singularly or in combination.
- Personalized strategies. Depend on the association of personal web behavior to an individual email address. For example, recently viewed and abandoned basket are linked to actions that an visitor has completed on the website.
- Merchandised strategies. Relate specifically and only to products. For example, best sellers, best trending, and new products do not depend on whether the email recipient saw the products.
The following table shows personalized and merchandised strategy examples.
|Abandoned basket products||Best trending (last 30 days)|
|Cross-sells from previous purchases and views||Best sellers by units (last 7 days)|
|Newly discounted products from viewing history||Best sellers by revenue (last 30 days)|
|Popular co-purchased products||New products (last 7/14/30 days)|
|Recently viewed products|
You should adopt a combination of personalized and merchandised strategies to ensure that emails are optimized to generate the best sales returns.
There is no single best answer for which stack of strategies generates the most engagement because this varies on a client-by-client basis. The nature of what is being sold or offered on a site also is an important consideration because this has a direct bearing on which strategies it is best to deploy. For example, a fashion site will have completely different buying behavior to a site that sells building supplies. A fashion site retailer may want to focus strategies on as much personalization as possible, while a building supplies retailer may want to focus more on the merchandised strategies available that are devised based on the crowd behavior captured. Over time, more emails are personalized as Optimizely associates web behavior to further email addresses.
To see a full list of strategies, expressions, and hints, click Help in any product position to open the Help section.
Filtering strategy results using expressions and hints
To narrow and restrict the recommended products in your Email Product Recommendations campaigns, add rules in the expression and hints sections for each product position. These rules work in conjunction with the configured recommendation strategies by filtering the strategy output using the specified parameters.
- Expressions – Explicit rules that the recommended products must match to appear to the email recipient. Expressions ensure that certain product attributes are included in the selection. For example, if you were to use the Recently viewed strategy, you could specify an expression to show only those products that were recently viewed and are from a specific brand or category you want to push. To use a certain product attribute in expressions, pass the attribute and value to Optimizely in the product catalog feeds imported for your site.
- Hints – While expressions let you filter by a specific product attribute and its value, and are configured for each product position individually, use hints as a broader rule to take into account what other products are recommended in the same campaign. For example, if you use the Recently viewed strategy, you may want to exclude items that were previously purchased (recent purchase), or show products from the same category. You can do this with a hint (but not with an expression because recent purchase is not a product attribute provided in product catalog feeds).
Adding an expression
To add an expression to an individual product position:
- Select the Expression tab for the desired position.
- Open the first drop-down list and select the attribute (such as Category) that you want in the rule (based on the product catalog feed, plus a few attributes Optimizely created by default). If you do not see the attribute that you want, you can add it to your product catalog feed.
- Select the condition you want in your rule.
- For attributes that have a text value, such as Color or Brand, select equals to or NOT equals to.
- For attributes with a numeric value, such as Sale price or Unit price, you also can use comparison operators like greater than or less than.
- In the text field on the right, enter the value of the attribute that is evaluated in this rule. You can start typing in the box to filter the results.
Most attributes have a drop-down list with possible values when you select the text field. For numeric values (such as prices) you will not see a pre-populated list and you will need to enter the value manually, using the decimal point separator (.) and not the comma (,).
- You can add multiple rules for the same position by clicking Add group or plus +. When using multiple rules, you can choose to apply one or more of them to the product position, by setting the AND or OR operator.
To remove a rule, click minus - next to it.
Adding a hint
To add a hint to an individual product position:
- Select the Hints tab for the desired position.
- In the drop-down lists select the hint you want to apply. You can apply up to two hints for each product.
You can use hints and expressions in combination, but be aware that the more you add, the more you restrict the range of products from which to choose to generate useful recommendations.
When you have configured the products in the campaign, click Save campaign.
Select the HTML and preview step next.