Salesforce renamed the same personalization product twice, then built a second one on Data Cloud and sold both. If you're a customer trying to plan around that, the confusion is the roadmap.
In 2020, Salesforce bought Evergage — a genuinely good real-time personalization engine with a decade of product behind it. It became Interaction Studio. Then it became Marketing Cloud Personalization. Then, instead of a third rename, Salesforce built a second personalization product on Data Cloud, called it Salesforce Personalization, and now sells both. If you've been in an evaluation and couldn't tell which product the deck was describing, that's not a you problem.
Naming churn sounds like trivia. It isn't. A product's name changes when its place in the vendor's strategy changes, and two renames plus a parallel successor is a lot of strategy changes for six years. Customers feel it concretely: documentation that references three product names, SIs quoting different implementations for "the same" capability, and renewal conversations where the safest question — which of these two products is the future? — gets an answer that depends on who's in the room.
Here's the map as of mid-2026. Marketing Cloud Personalization is the Evergage lineage: a capable on-site engine with its own implementation model, respected by the people who ran it. Salesforce Personalization — also branded Einstein Personalization — is the new build: native to Data Cloud, decisioning against unified profiles and identity resolution, wired into Agentforce, priced through consumption credits. Both are sold today. Salesforce has announced no end-of-life for the legacy product, and honesty requires saying that plainly: anyone telling you Marketing Cloud Personalization is "sunset" is ahead of anything Salesforce has published.
But you can read where the investment ships. New personalization capability lands on Data Cloud. Marketing Cloud Next consolidates the acquired marketing products onto the core platform. The legacy product's release notes have the energy of a maintained product, not a flagship. None of that is an official roadmap — which is exactly why, if you own either product, the move is the same make-them-prove-it posture we've recommended for every vendor in this category: get the roadmap in writing, with dates, from your account team. A vendor confident in a product's future puts it in the contract.
Suppose you follow the investment and evaluate the new product. The list price is not the cost. Salesforce Personalization assumes Data Cloud underneath it: licensing, data ingestion, identity resolution, consumption credits that meter usage — an implementation program before the first visitor sees a different hero. For an all-in Salesforce enterprise that wants one governed profile driving email, mobile, agents, and web, that can be a rational program. But notice what's being asked: to personalize your website, you first build a data platform.
That's the old category error wearing a new architecture. What a page shows the visitor in the next 200 milliseconds is a decision-layer problem — in-session signal, runtime choice, immediate proof. A unified profile store is a data-layer solution. It helps the known-customer, cross-channel cases and says very little to the anonymous majority of your traffic, who resolve to no profile at all. We've written about the tax you pay stitching data layers to decision layers; building the warehouse first is the same tax, invoiced up front.
Stay native if three things are true: you're genuinely all-in on Salesforce, Data Cloud is funded with a team or SI partner to run it, and your personalization strategy centers on known, logged-in customers. That buyer gets real compounding value from one governed platform, and their hard question is only which of the two products to standardize on.
Add a real-time decision layer instead if your traffic is mostly anonymous, your timeline is this quarter rather than this roadmap, or you want lift proven against a holdout before a platform-sized commitment. Critically, this isn't a rip-and-replace: the decision layer runs from a tag, and the CRM stays exactly where it is. The honest head-to-head is here, including the cases where Salesforce native is the right call.
AXO takes a position most vendors in a Salesforce comparison won't: keep Salesforce. The CRM is a system of record, and the record layer isn't the layer we contest. AXO is the experience layer on top — one tag, in-session behavioral scoring, deterministic on-page decisions, an always-on holdout — and it syncs the leads and pipeline it produces back into Salesforce. No Data Cloud prerequisite, no consumption-credit forecast, public pricing, live the same day. The maze is optional. The record isn't.
They are the same product under successive names. Salesforce acquired Evergage in February 2020, rebranded it Interaction Studio, then rebranded it again as Marketing Cloud Personalization. The product is still sold today alongside Salesforce Personalization, a separate, newer product built on Data Cloud.
Yes. Salesforce Personalization (also called Einstein Personalization) is built on Data Cloud and decisions against its unified profiles, identity resolution, and data graphs, with usage metered through Data Cloud consumption credits. The effective cost and timeline therefore include standing up Data Cloud, not just the personalization product itself.
It depends on your Salesforce commitment. Migrating makes sense for all-in Salesforce enterprises with a funded Data Cloud deployment and a known-customer personalization strategy. If your traffic is mostly anonymous or you need results this quarter, a tag-based real-time decision layer delivers on-page personalization without the Data Cloud prerequisite — while Salesforce remains the system of record. In either case, get the legacy product’s roadmap from your account team in writing before renewing.