Site Recommendations Overview
Enable custom, personalized experiences on your website with Site Recommendations.
As with email personalization, Site Recommendations leverages:
- custom feeds as a source for your content
- the full history of user data collected across all channels to power personalization based on interests, past behavior, and predicted behavior
- the powerful Zephyr language, allowing you to further customize which content is displayed based on context, user, and content attributes
- robust tracking and reporting features, allowing you to measure the impact of site recommendations and further augment user profiles based on user actions
Site Recommendations is fully customizable to your needs, but here are some example uses:
- Match the right content to the user based on user profile and content parameters For example, a commerce site could match the user’s gender (as defined by a var on their profile) to gender-specific products (as defined by content tags)
- Match the right content to the user using the context of that specific visit For example, a media company with an edition for each continent can automatically tailor content selections to match the location of the current visitor
- Match the right content to the user using the context of the page For example, a commerce site wants to ensure that when a shopper is looking at an infant product, the “You might also like” recommended products are also for infants
- Match the right content to the user based on past user actions For example, the client recently purchased a new pair of boots and is unlikely to want to immediately purchase another, so recommend other types of products.
- Restrict the pool of content available for recommendation to a specific feed For example, only select content to recommend for the homepage hero spot from a “hero” feed.
- Allow for updated or additional recommended content based on user actions on a page For example, a media client displays content dynamically in a list with infinite-scroll and wants to receive recommendations based on what a visitor does on the page.
- Enable total control over site aesthetics, even with a complex design and layout Have Site Recommendations return content as raw JSON, offering total control over how the recommendations are displayed.
- Use consistent messaging across email and onsite For example, a media client wants the same content (or price/discount) and a similar layout to be reflected on the landing page that the user visits.
And here’s how Site Recommendations works:
- Personalize any section of any page. Each section:
- Uses a particular content feed and set of rules to determine which content to display.
- Can be implemented on a single page or reused across multiple pages–each instance can differ based on context.
- Can produce HTML/CSS code, rendered according to your template, or a raw JSON data feed for your site to format.
- Vary the treatment of a section by creating audiences who receive alternate feeds, feed-filtering rules, and/or templates.
- Each audience corresponds to a known-user list of your choosing, whether it’s a newly created list or one that you also leverage for email campaigns.
- Audiences have an order of precedence of your choosing.
- If the user matches the first audience, it is displayed.
- Otherwise, the next audience is checked, and so on.
- If the user does not match any audience, the parent section’s default settings are used.
- Manage sections and access reporting two ways:
- By using the Site Recommendations section on my.sailthru.com
- Programmatically, using the API
- Site Recommendations supports all mobile and desktop browsers that have 1% or greater North American market share and will continue to support all newer versions of those respective browsers, regardless of market share.