Messaging Recommendations

Messaging Recommendations include email and SMS recommendations that can be used in your marketing automation provider's templates. This feature is available for the Klaviyo, Dotdigital, and Omnisend platforms. It is designed to drive a more engaging shopper experience, with the ultimate goal of enticing shoppers to return to your website and complete or expand their order.

Part of the Athos Personalization Suite

Athos Commerce Messaging Recommendations are a key component of the Athos Personalization Suite. This product extends the reach of our personalization offering so that we can truly weave personalization throughout the shopper’s journey.

Without personalization, campaigns sent to a shopper after they leave the site often use simple recommendation concepts or manual business rules such as “best selling” or “trending.” Those recommendations are not truly personalized. All shoppers receive similar recommendations regardless of how relevant those products are to the individual shopper. 

Incorporating Athos Messaging Recommendations in an email or SMS message increases the relevance of the recommended products sent in a marketing or post-purchase campaign. Our true personalization uses shopper behavior and shopper preferences based on each shopper’s on-site activity. 

How It Works

Whenever shoppers visit a site, they provide information about their interests in the form of pages they visit and the products they view. Every time a shopper views a product, adds a product to the cart, or completes an order, Athos’s personalization algorithm learns from those actions to better understand the shopper’s intent and predict what that shopper wants to buy next. Those predictions generate recommended categories, brands, and products that are used to personalize their experience and get the right product in front of them at the right time. By engaging Athos Personalization throughout the shopper's journey we are able to provide a consistent experience both on the site and in their inbox.

What Makes Our Email and SMS Recommendations Special

At Athos, we believe in setting new standards when it comes to Personalized Recommendations. Here's why our Messaging Recommendations stand out from the rest:

1. Customizable rules for merchandisers: Use our Athos console to curate and customize personalization rules effortlessly. You can set up Advanced Rules to customize a Personalized Recommendation profile, and those rules will be applied across your store and Messaging Recommendation campaigns. This allows you to maintain consistent merchandising rules and personalized recommendation behaviors. Take control of your recommendations and deliver a seamless experience to your shoppers.
2. Personalization based on email opens: We take personalization a step further by calculating it on email-open and on SMS-open, rather than just on send. By doing so, we ensure that our recommendations are hyper-relevant, aligning with the recipient's most recent activity. Stay in sync with your shoppers and provide recommendations that truly matter to them.
3. Cutting-edge machine learning and AI analysis: With every indexing of your feed, we employ advanced machine learning and AI analysis to deeply understand your site's data. Unlike others, we go the extra mile by performing frequent indexing for all sites, ensuring that our data remains synchronized and up to date with any changes made to your site. The result? Unparalleled accuracy and relevancy in generating recommendations, regardless of when they're needed.
4. Truly one-to-one personalization: We believe in delivering a personalized experience that's unique to each customer. Our Messaging Recommendations are tailor-made for every individual, reflecting their preferences and interests. Say goodbye to generic suggestions and embrace a personalized journey for each shopper.

Experience the difference with Athos's Personalized Messaging Recommendations. Unleash the power of advanced technology, individualized personalization, and easy customization to enhance your shoppers' journey like never before.

Add a profile

To generate a messaging recommendation, you'll need to create a new profile. Click "Personalization" in the main menu of the Athos console to get started.

1. Click Marketing Automation in the Personalization sub-menu. Two tabs will display: Personalized Segments and Messaging Recommendations.
2. Click the "Messaging Recommendations" tab.
3. Click the “+ Add New Profile” button in the top right.
4. A new window will appear where you will select the marketing automation provider to use.

5. Choose the type of campaign you'd like to use. In this example, we'll use "Abandoned Cart". The system will automatically open the next screen.

6. Choose the type of recommendation you'd like to use. In this example, we'll choose "Cart Cross-Sell".

Note: During the initial selection process, you can click the navigation at the top of the window to go back to a previous step and modify your choices.

Once you click "Create Profile" at the bottom right, which will open the Profile Editor, you can click the breadcrumbs at the top left to update the page placement and recommendation type, as long as you have not clicked the "Save Profile" button at the bottom right. Once the profile has been saved, you cannot modify the recommendation type or placement.

7. Once you have selected all the specifications for the profile, click the "Create Profile" button.

8. The profile editor form will appear. Here, you will customize the template, precision, and layout for your messaging.

Details tab

The Details tab will display by default. Add a profile name, then select a template from the product card drop-down menu. The template determines which custom options are available, and that setting allows you to control the layout, appearance, behavior, and content of this profile.

Email Layout tab

On the Email Layout tab, you'll set the breakpoint and how many products to show in Desktop and Mobile devices.

Note: You may recommend up to 20 products in an email, but SMS recommendations will contain only one product recommendation.

Desktop/Mobile Breakpoint

Enter the number of pixels to represent the width of the shopper's screen, at which point the responsive design will adapt content and layout to that size.

Note: The maximum number of items that can be recommended is 20.

Max items = (Recommendations per Row) times the (Number of Rows)

Desktop

• Recommendations per Row = This value represents how many product recommendations the shopper will see in one row on the screen.
• Number of Rows = This value represents how many rows of products will be shown to the shopper.
• Cell Padding (in px) = This value provides spacing around each product shown in the layout preview.

Mobile

• Recommendations per Row = This value represents how many product recommendations the shopper will see in one row on the screen.
• Number of Rows = This value represents how many rows of products will be shown to the shopper.
• Cell Padding (in px) = This value provides spacing around each product shown in the layout preview.

Note: You may recommend up to 20 products in an email, but SMS recommendations will contain only one product recommendation.

Layout Preview

Before you integrate this code into your marketing automation provider's site, you can test how the layout looks. Click the "Layout Preview" button in the top right of the window.

A new window will display tabs for Desktop and Mobile, so you can compare how each layout looks. The window will open on the Desktop tab.

Click the Mobile tab to see the mobile layout you have configured.

Click the "x" at the top right to exit the layout preview.

Integration instructions

Click the "Save" button in the upper right-hand corner. Once you save the profile, a new button will appear at the bottom left of the Profile Editor window, "Show Integration Instructions". This button will appear on every tab within the Profile Editor.

Click the "Show Integration Instructions" button to see how to integrate these layout settings with your marketing automation platform. A new window will appear, with separate tabs for Email and SMS.

Click the "Copy Code" button at the bottom right to copy the HTML code that you will paste into your marketing automation provider's template. Click here to see how to set up that template for each provider:

• Klaviyo
• Dotdigital
• Omnisend

Click the SMS tab to see a similar form; this window will contain separate code sections for the SMS link and the SMS image. As with the HTML code for email, you will click the "Copy Code" button at the bottom right of each section to copy the HTML code that you will paste into your marketing automation provider's template.

Precision tab

On the Precision tab, drag the slider to configure how precise recommendations are. This value dictates how many conditions are used to find recommended products. Broad precision expands the recommendations algorithm to ensure products can be recommended even when shopper data is sparse. Narrow precision restricts recommendations to prevent the algorithm from defaulting to related products or top sellers. This could result in recommendations not being displayed for certain products.

• Curated = Curated recommendations will be limited to the products set in Product Rules. If the seed product does not have one or more products associated with it in a rule, no recommendations will be returned. For example, if merchandisers want to create a "Complete the Look" experience, they would set the precision filter to "Curated" so the recommendations carousel would only show the pinned products.
• Cross-sell = Cross-sell recommendations are used on a page to recommend products that go well with the product that the shopper is viewing. 
• Similar = Similar recommendations are products that a shopper should consider buying as an alternative to the product they’re currently viewing.
• Trending = Trending recommendations are popular product bundles across a rolling window of time. This recommendation type supports either site-wide trending or trending within a category or brand.

Note: The Precision tab is available only on Similar, Cross-sell, Cart Cross-sell, and Search Cross-sell profiles. You can learn more about precision here.

Filters tab

Filters can be used to limit which products can possibly show up in a profile. This selection varies for different types of profiles. Note that recently viewed profiles cannot be filtered.

For custom profiles, the filters are applied in advance to help determine the set of products to recommend. A common use would be to only display products that have a certain tag. For example, if you only want to recommend products from a specific season, you could set up a filter requiring products to match a seasonal tag such as "ss_season is Winter". Any field that is available in the search index can be used for custom profile filters.

For other types of profiles (including similar, cross-sell, and trending), filters are applied at the time that the recommendations are requested by the website. That means the filters will reduce the number of results returned and may cause the profile to be hidden if no appropriate results are available. It is best to apply these filters only in cases where it is better to show nothing instead of the wrong thing.

Tip: See Personalization Settings for more details on how you can use business rules to influence recommendations instead of using filters.

Presets tab

The Presets tab allows you to control the default behavior for the selected profile. Note that the preset options will differ for each type of recommendation. Merchants can specify default brands, categories, or search terms for a recommendation profile, rather than having their front-end team or the Athos support team code these values on the front-end page.

Presets provide you with a great amount of leeway to "set it and forget it". For example, a preset recommendation profile for the home page could be updated once a month: swimwear in July, back-to-school in August, and autumn/Halloween for September.

In contrast, a regular (not preset) "category trending" recommendation profile for the category results page would display different products for each category page the shopper visits, based on the category sent from the front end.

Presets give you the option to set the following:

• Items blocked from being shown in the selected profile
• Specific categories to use in a category trending profile
• Specific brands to use in a brand trending profile
• Search terms to use in a dynamic custom profile

Note: Front-end settings can override presets configured here. If you configure a limit of 10 in the presets, but the front end has a limit of five, the system will show five to a shopper. If you wish to configure presets but there are already preset configurations in place on the front end, these will need to be removed either by your team or by Athos via a support ticket request.

Blocked Items

Click the drop-down list to select products to block. You can also use this field as a search box to find a particular item. If you change your mind after an item is added, simply click the "x" at the far right to un-block the selected product.

Note: Blocked items are not supported in "Recently Viewed" profiles.

Pins tab

Use the pins tab to assign specific products to fixed positions in the recommendation profile. Note that items pinned here are associated with the overall profile, not the seed product/category/brand entered in the search bar (at the top right side of the page). If you prefer to pin only for a specific seed product, category, or brand, modify the personalization setting's similar or cross-sell rules instead.

Note: Pins are not supported for the "Recently Viewed" profile type.

Note: Pinned items will be ignored if they also match an Excluded Products rule (for example, if they go out of stock).

Items can be pinned to any position, and after a specific slot (position) is selected, you can determine if the pin should only show when the slot is available. For example, if there are only four recommendations but a product is pinned to slot #8, setting the pin to "float" would have the product display at the end of the four recommendations. If the pin is set to "fixed", then it will not display if there are less than eight recommended items.

To pin an item, search for it in the search box at the top of the panel, then click the pin icon at the top left. Once you have selected an item, you have the option to accept the pinning position suggested, or enter another number. Click the green checkmark icon to confirm the pinned position.

You can also pin an item in the Recommendation Preview on the right side of the page from any tab, and you can drag-and-drop an item to pin it in a different position.

Editing a profile

Existing profiles will be listed under Messaging Recommendations. To edit a profile, click the "..." button under the "Options" column on the far right, then choose "Edit".

The profile editor will display, and you can make changes on this form, then click the "Save" button to save your changes.

Boost Rules (for Custom Recommendations)

If you've chosen the Custom recommendations type, you can select the products to be recommended by applying boost rules. You can read about the options for custom recommendations profiles here.

Just like Boosting Rules in Visual Merchandising, this function allows you to create rules that will dynamically boost products with certain attributes to the top of your recommended products. These rules will continue to stay in effect and re-group your products in real time as their information changes.

Although we show only the first 20 items in a recommendations profile, you can use boost rules to elevate specific products to the top of the list. Each boost rule separates products into an elevated group that meets the rule and a deprecated group that does not. Additional rules create sub-groups within the first set.

The higher the rule is positioned in the list, the greater impact it will have on the results. Boost rules are not required.

Previewing Recommendations

For custom profiles, the product recommendations preview (at the right side of the window in the Profile editor) will show you the actual product results matching your boost and filter selections. Similar, Cross-sell, Cart Cross-sell, and Search cross-sell profiles require a product seed for the preview. Category and Brand Trending profiles require a category or brand to be selected; you can search for the appropriate input to see the results for that selection.

Note: These are only previews. When the profile is placed on your site, the seed product, category, or brand will be captured from variables on that page and sent in the request.

Note: Previews are not supported for Personal, Session, and Recently Viewed recommendations, as they are entirely dependent on an individual shopper’s behavior.

When you click on a recommended product, it will show the details of that product and provide context on why that product has been recommended. For example, on a similar or cross-sell profile you will see:

• Total Sales and Total Views for the recommended product
• Common Sales or Common Views between the recommended product and the seed product
  • Common Sales will be shown for Cross-Sell recommendations while Common Views are used for Similar recommendations.
  • Common Views are calculated over 26 weeks. 
  • Common Sales are calculated over 18 months and measure how often a product is repurchased by the same shopper. The repurchase does not necessarily need to be on the same order to qualify as a common sale. 
• The Brand and the Categories associated with the recommended product
  • The preview will highlight the categories that the recommended product shares with the seed product.

Note: This preview will not exactly match what shoppers will see once the profile is implemented on your site. Depending on the individual shopper’s behavior, the order of recommended products will change in order to provide a personalized experience.

Ready to Install? 

Check out Installing a Recommendations Profile to continue.

Create Klaviyo Recommendations

Click here to learn how to create messaging recommendation intelligence in Athos and inject it into Klaviyo.

Create Dotdigital Recommendations

Click here to learn how to create messaging recommendation intelligence in Athos and inject it into Dotdigital.

Create Omnisend Recommendations

Click here to learn how to create messaging recommendation intelligence in Athos and inject it into Omnisend.