Agentforce Marketing Summer Recap

Agentforce Marketing Summer Recap title image

Between Connections in June and the Summer ’26 release, there has been a lot to keep up with. The volume of updates this cycle is one of the biggest I can remember, and trying to cover everything would turn into a small book. So instead of listing every feature, I pulled together the highlights that stood out to me and that I think matter most for marketing teams heading into the second half of the year.

One thing worth calling out is the direction all of this points in. Most of what shipped this summer moves marketers away from manual execution and toward setting goals while agents handle the busywork. That theme is not new for Agentforce Marketing, but this release makes it a lot more concrete. Here is what made my list.

Agentforce Content Studio

The update that will change day-to-day work the most is around content creation. With Agentforce Content Studio, marketers can describe a campaign and have an agent generate content across channels: email, mobile messages, SMS, RCS conversations, and personalized promotions. Localization happens inside the same flow, so global teams stop waiting on translation handoffs before a campaign can go live.

The part I find interesting is how it changes the shape of the work. Building a multi-channel campaign used to mean weeks of coordination across production, translation, and channel owners. When an agent handles the first draft of that content, the time you save goes back into the ideas and storytelling that differentiate the brand. The agent is grounded in brand guidelines and customer data, so the output is not generic filler.

One expectation to set: as of now the Content Agent is still in pilot, so this is a preview of where things are going rather than something every org can turn on today. If you manage content operations, this is the feature to watch, because it reshapes how content gets created and managed rather than only speeding up one step.

MCP for Marketing Cloud Engagement

This is the announcement that grabbed most of the headlines, and for good reason. The Model Context Protocol (MCP) server for Marketing Cloud Engagement is now generally available. In plain terms, it lets an AI assistant like Claude, Cursor, or Gemini connect directly to your Marketing Cloud Engagement account and act on it using natural language. You can ask it to build a data extension, spin up a welcome journey, or write and test SQL, all without clicking through the UI or writing custom code.

Now for the part I want to be direct about: teams should be cautious before enabling this. The server can read your contact data and write to journeys and automations, which is a lot of power to hand to an external model. Before anyone connects, be deliberate about the permission scopes you grant, who is allowed to authenticate, and what data ends up passing through an outside LLM. Treat it like giving a new integration production access, because that is effectively what it is. Start in a non-production environment, keep the scopes tight, and make sure your security and compliance folks are in the room before it goes live.

Marketing Objects

The Summer ’26 release introduced Marketing Objects in Marketing Cloud Next, and they solve a real problem for teams that want to move quickly. Marketing Objects are data stores, similar in spirit to Data Extensions in Marketing Cloud Engagement. You import data through a CSV file, and then you can read and process it with Handlebars or AMPScript to personalize your emails. That means a team can stand up a new campaign, load reference data, and schedule sends without waiting on a full data pipeline.

The benefit that stands out is reduced dependency on Data Cloud for content personalization. Marketing Objects can act as a data source for personalization as an alternative to data graphs, so you are not blocked on a complete Data Cloud build before you can get a campaign out the door.

I want to be clear that this does not remove your need for data. You still need Data Cloud in the picture to capture engagement and manage consent properly, and where Data Cloud is well structured you can personalize directly from it. Marketing Objects are also limited to manual CSV imports for now, so think of them as a fast, lightweight option for reference data rather than a replacement for your customer data foundation. Used the right way, they take some friction out of getting campaigns live while your longer-term data work continues.

AMPScript Meets Handlebars

AMPScript earned a spot on the summer highlights too, and this one landed with a lot of the community. AMPScript is now native in Marketing Cloud Next, sitting alongside Handlebars in the message editor. For teams moving over from Marketing Cloud Engagement, this protects a decade of existing logic instead of forcing a full rewrite of every template.

The interesting part is that you can combine the two languages in a single template. A pattern that works well is to think in layers. The profile layer stays in AMPScript and handles subscriber attributes, preferences, and legacy fields. The layer that pulls in the newer Salesforce Objects sources, such as cases, opportunities, and leads, moves to Handlebars, since those sources are exposed through Handlebars and no longer require data graphs. The presentation layer stays in HTML with Handlebars helpers for conditional rendering and loops.

A couple of practical notes. Not every AMPScript function from Engagement is available in Next yet, so check the official function reference before you migrate anything non-trivial. And even though AMPScript is back, Handlebars is still the skill your team needs to reach the new data sources like Salesforce Objects, Content Variables, and Offers. My take is to make sure at least one person on the team can read and write Handlebars confidently, otherwise you stay limited to what AMPScript alone can do.

What Is Next, Marketers?

Dreamforce is around the corner, and if this summer was any indication, the core platform will bring another wave of updates. Expect more of the agentic direction to show up across the product.

My advice is to use the time before Dreamforce to think past the individual features and get clear on strategy. Where does B2B fit, where does B2C fit, and how do these new tools change your plans for each? Content Studio, MCP, Marketing Objects, and AMPScript in Next are all pieces of a larger shift, and the teams that connect them to a strategy will get more out of them than the teams chasing features one at a time.

Get your thinking in order now, watch what comes out of Dreamforce, and start preparing for the Winter release. The pace is not slowing down, so the best move is to stay a step ahead of it.