Marketing Cloud Personalization
Marketing Cloud Personalization is the Salesforce product that lets you visualize, track, and manage customer experiences in real time.
Definition
Marketing Cloud Personalization is the Salesforce product that lets you visualize, track, and manage customer experiences in real time. It captures behavior across web, mobile app, email, and server-side channels, builds a unified visitor profile, and decides what content, product, or offer to surface to each individual at the moment they interact. The product was acquired as Evergage in 2020, renamed Interaction Studio, then rebranded to Marketing Cloud Personalization in the Winter '23 release.
The product sits in the Marketing Cloud family as the real-time activation layer on top of profile data. Marketing Cloud Engagement runs email and journey orchestration, while Personalization makes the moment-of-truth decision when a visitor lands on a page or opens the app. Its recommendation and decisioning engines run on Einstein, and Data Cloud increasingly supplies the unified profile behind both products.
How Personalization decides what to show in real time
Evergage roots and the rename to Personalization
Salesforce acquired Evergage in 2020 and folded the product into Marketing Cloud as Interaction Studio. In the Winter '23 release the product was rebranded again to Marketing Cloud Personalization, with Salesforce stating the new name better reflected its role inside Marketing Cloud Engagement. The underlying platform stayed the same across all three names, so the rename was cosmetic rather than a re-architecture. This history matters when you read documentation or community posts. Anything written before late 2022 calls the product Interaction Studio, and the very early SDK references still carry the Evergage class names. The iOS SDK, for example, ships an Evergage object that predates the rename. When you search Help or Trailhead today, both Interaction Studio and Personalization point at the same feature set, so treat the two names as interchangeable. Knowing the lineage also helps with skills hiring and certification. Practitioners who list Evergage or Interaction Studio experience are describing the same tool a newer resume would call Marketing Cloud Personalization, and the configuration concepts carry straight across.
Behavioral tracking and the affinity wheel
Every visitor action feeds a detailed profile. A JavaScript tag on the website and SDKs on mobile apps capture page views, clicks, scrolls, product views, cart adds, form submissions, and video plays as they happen. Server-side events can be sent through the API for channels a tag cannot reach, such as a call center or a point-of-sale system. Personalization turns that raw stream into structured profile data. The interaction history logs time spent on pages, items viewed, and transactions completed. User attributes capture things like loyalty tier and location. The affinity wheel summarizes what each visitor cares about, inferring interests from the categories and attributes they engage with most. Anonymous and known-user profiles unify when a visitor authenticates, so the pre-login behavior is preserved against the identified person. That unification is the difference between a generic recommendation and one informed by everything the visitor did before they logged in. The profile is the single foundation for every downstream decision, which is why gaps in tag coverage degrade personalization quality everywhere at once.
Segments built from live behavior
Once profiles exist, you group them into segments. Personalization supports segments of visitors, customers, and accounts, defined from behavioral signals, profile attributes, affinities, and engagement history. Because the profile updates in real time, a visitor can enter or leave a segment within the same session, which is what makes behavioral segmentation different from a static list export. A segment of high-affinity shoppers who viewed a category three times this week is recomputed continuously rather than refreshed on a batch schedule. Segments do double duty. They power audience analysis, so you can see how a group behaves across the catalog and across campaigns, and they target campaigns, so a banner or recommendation only fires for the right people. You can also feed segments outward. Through the Data Cloud connector, a Personalization segment can be activated back into the unified profile and used by other Marketing Cloud tools, which keeps the audience definition consistent across email, journeys, and on-site experiences instead of redefining it in each channel.
Campaigns across web, mobile, email, and server-side
A campaign is the unit of work that delivers a personalized experience. Personalization runs campaigns across digital and physical channels, and the documentation calls out web, mobile, open-time email, and server-side delivery. Web campaigns inject content into defined content zones on the page. Open-time email campaigns resolve the personalized content at the moment the email is opened, so the recommendation reflects the latest behavior rather than the state at send time. Server-side campaigns return decisions through the API for any system that can make a call, which covers contact centers, kiosks, and custom apps. Campaign types include testing and targeting experiences such as banners, recommendations, and exit-intent popups. The same profile and segment data drives all of them, so a cart abandonment on web can inform an email reminder, a push notification, or a banner on the next visit. Click and view tracking is handled automatically in web campaigns. In server-side Einstein Decisions campaigns only view logs are captured automatically, and clicks must be tracked manually through the API, which is an easy detail to miss during implementation.
Einstein Recipes and Einstein Decisions
The machine learning sits in two related features. Einstein Recipes generate product and content recommendations. Ingredients form the core algorithm of a recipe, selecting which catalog items qualify, and Personalization then weights those items by each visitor's behavior and affinities. You refine a recipe with inclusions and exclusions to filter by category, boosters to elevate preferred items, variations to add diversity, and fallback strategies so empty slots still fill. Einstein Decisions handles next-best-offer and next-best-action choices, scoring options in real time to pick the experience most likely to engage a given visitor. Decisions can run in a web template or a server-side template, and the two are configured slightly differently. The web template uses the content zone selector defined in the sitemap. The server-side template uses ContentZoneLookup instead, which is not limited to sitemap zones and lets a business user choose any content zone or tag value stored on a promotion asset. Recipes answer what to recommend, Decisions answer which offer or action to make, and a single campaign often combines both.
Where Data Cloud fits
Salesforce's strategic direction routes Personalization profile data through Data Cloud rather than a standalone silo. The Marketing Cloud Personalization Connector creates a data stream that ingests Personalization engagement and behavioral data into Data Cloud, maps it to the standard data model through preconfigured bundle mappings, and improves identity resolution by enhancing user mappings. Activation runs the other way too. A unified segment built in Data Cloud can be sent back into Personalization and used to target on-site and in-app experiences, so the audience defined once in the lake reaches the real-time decisioning engine without being rebuilt. For new deployments this is the recommended architecture, because it keeps the profile that powers a website banner consistent with the profile that powers an email or a service interaction. Existing standalone deployments can continue to run on the older internal profile and migrate over time. The practical takeaway is that Personalization is increasingly the real-time activation surface for Data Cloud, not a separate database, and planning the connector early avoids reconciling two competing profiles later.
Licensing and where it sits in the stack
Marketing Cloud Personalization is licensed separately from Marketing Cloud Engagement, Account Engagement, and Data Cloud, and pricing typically scales with profile volume. High-traffic sites need to size that volume carefully, because anonymous visitors count toward profile usage and a popular catalog can generate far more profiles than the marketing team expects. In the wider Customer 360 picture, Personalization is the layer that turns unified data into a decision a customer actually sees. Data Cloud assembles the profile, Einstein supplies the models, Engagement orchestrates the outbound journeys, and Personalization makes the in-the-moment choice on web, app, email open, and server-side surfaces. Because it touches the live website, implementation is as much an engineering exercise as a marketing one. It needs a correctly deployed sitemap, reliable tag coverage on every template, and a catalog feed that matches the items you want to recommend. Teams that treat it as a pure marketing tool and skip the technical groundwork usually end up with thin profiles and recommendations that miss, which is why a staged rollout beats a big-bang launch.
Implementing Marketing Cloud Personalization
Standing up Personalization is a configuration and integration project, not a single record you create. The high-level path takes you from account access through tag deployment, catalog and sitemap setup, and a first live campaign. The steps below outline that path; the detailed field-level work lives in the Salesforce Help implementation guide.
- Get account access and plan the data
Confirm your Personalization dataset is provisioned and decide which channels you will track first. Map the visitor attributes, catalog categories, and key actions you care about before touching code, so the implementation has a clear target.
- Deploy the tag and SDKs
Add the Personalization JavaScript tag to every web template and integrate the mobile SDKs for app coverage. Verify the tag fires on each page type, because a missed template creates a permanent blank spot in the profile.
- Build the sitemap and catalog
Configure the sitemap so Personalization understands your page structure and content zones, and load the product or content catalog that recipes will recommend from. Keep the catalog feed in sync with the live site.
- Define segments and a first campaign
Create a starter segment from real behavior, then build one campaign, such as a recommendation strip or an exit-intent banner, targeted at that segment. Launch small, measure, and expand once the data looks clean.
Web content zones, mobile app SDK, open-time email, or server-side API; pick the surfaces you will personalize first.
The recommendation algorithm, configured with ingredients, inclusions and exclusions, boosters, variations, and fallbacks.
Web template using sitemap content zones, or server-side template using ContentZoneLookup for promotion assets.
Optional data stream that ingests engagement data into Data Cloud and activates unified segments back into Personalization.
- In server-side Einstein Decisions campaigns only view logs are tracked automatically; clicks must be tracked manually through the API.
- Anonymous visitors count toward profile-based licensing, so high-traffic catalogs can exceed expected profile volume.
- Documentation and community posts written before late 2022 call the product Interaction Studio; it is the same tool.
- Missing tags on any template leave gaps in the visitor profile that degrade every downstream recommendation and decision.
Trust & references
Cross-checked against the following references.
Straight from the source - Salesforce's reference material on Marketing Cloud Personalization.
Hands-on resources to go deeper on Marketing Cloud Personalization.
About the Author
Dipojjal Chakrabarti is a B2C Solution Architect with 29 Salesforce certifications and over 13 years in the Salesforce ecosystem. He runs salesforcedictionary.com to help admins, developers, architects, and cert/interview candidates sharpen their fundamentals. More about Dipojjal.
Test your knowledge
Q1. What is Marketing Cloud Personalization?
Q2. What channels does it work across?
Q3. What's the original name of the product?
Discussion
Loading discussion…