Connections, the Salesforce conference for marketing and commerce, is taking place in Chicago, June 11-12. The main theme running throughout is Agentic Marketing – putting autonomous AI agents to work to personalize and connect customer experiences across every touchpoint. Beyond the jargon, the question that Salesforce set out to answer is “what if every message could start a two-way conversation?” believing that (for example, email, SMS) should be the channels for customer-agent interactions, all in natural language.
The main keynote showcased the different ways that Salesforce have weaved agents into marketing use cases, arranged into the main pillars: create, engage, qualify, and optimize.
The showstopper announcement is Marketing Cloud Next – a combination of multiple agentic features that Salesforce have been developing, coming together into one ‘launch pad’ interface.
On the marketing product side, we’re seeing real-life customer implementation stories come into fruition for the newer Marketing Cloud Growth/Advanced editions (i.e. Marketing Cloud ‘on core’). “What a difference a year makes”, remarked one Salesforce Director, in light of new edition adoption, and also the feature parity (with MCE/MCAE) and innovation beyond.
What Is Marketing Cloud Next?
Marketing Cloud Next is an umbrella that collates multiple agentic offerings, as follows:
- Agentforce Campaign Creation (with unstructured data activation in Data Cloud)
- Agentforce Personalization Decisioning
- Agentforce Lead Management
- Agentforce Web Curation
- Agentforce Lead Management
- Agentforce Segment Intelligence
So, some of you may be asking “why is this new?” – especially if you are familiar with the Marketing Cloud Growth and Advanced editions. In a nutshell, these are the three areas that makes Marketing Cloud Next the new era of marketing on Salesforce:
- Combining: Salesforce have made many acquisitions in the MarTech and customer data spaces over the years. Marketing Cloud Next combines 9 of these ‘best of breed’ acquisitions.
- Rebuilding: Marketing Cloud Next is built on top of Data Cloud (which is colloquially referred to as the ‘core platform’). Additionally, Salesforce have rebuilt their personalization engine onto Data Cloud, which means that there is one shared service for all Salesforce platform products (e.g. Sales Cloud, Service Cloud, Commerce Cloud).
- Simplifying: It’s truthful to say that, in the past, there was a cocktail of offerings under the “Marketing Cloud” umbrella. Now, not only can B2B and B2C models be serviced by the Marketing Cloud Next same application, Salesforce have been striving to reduce the TCO (total cost of ownership) by minimizing integrations within their marketing stack, and between other platform products.
If you’re wondering “how am I going to access this?”, Salesforce is planning to make Marketing Cloud Next available alongside Marketing Cloud Engagement (formerly ExactTarget), so that the technology can be used in the run up to switching over. Salesforce emphasized that this is not a migration – but a convergence – as your previous investments into segments, automation, etc. can be reflected in Marketing Cloud Next.
“Keep what works and add what’s Next” can be summarized by looking at the diagram below – “Build your Next messaging and flows, and invoke existing [Journey Builder] journeys”.
Salesforce’s Agentic MarTech Stack
In the spirit of ‘shared services’ across multiple teams in the organization using their respective Salesforce offerings, the agentic marketing vision relies on Agentforce, Data Cloud, and the metadata platform working together alongside Marketing Cloud Next – which, in short, is where marketers will interface with the other pillars.
Let’s now dive into Marketing Cloud Next’s offerings, starting with the ‘create’ pillar.
Agentforce Campaign Creation
“Launch campaigns in days, not weeks.” Agentforce co-creates campaigns with marketers, from brief to launch.
- Agentforce co-creates with the human user, to generate a full campaign brief.
- AI agents help determine the audience, and prepare the segments.
- Draft the email and SMS content using brand guidelines and messaging.
- The customer journey base orchestration (i.e. Campaign Flow) is set up, ready to review and activate.
The keynote demo began as a Slack conversation, showing that agents can be effective in various mediums.
While not new, it’s worth putting an emphasis on unstructured data activation in Data Cloud at this point. Unstructured data includes documents, images – just anything that’s not arranged as a database. Right from Marketing Cloud Next, use content from new sources like Google Drive, SharePoint, Zendesk, blogs, and documents to fuel campaign creation.
Agentforce Personalization Decisioning
“You can’t build relationships with ‘do not reply’” is the tagline for the second pillar: engage. This is effective as dead-end messages rarely invite the recipients to take the desired call to action.
We’ve seen Agentforce embedded on websites before, going beyond chatbots to surface contextual replies from your organization’s customer and product data. What’s new here, is that Agentforce adapts website content in the moment, and can understand sentiment at an even more impressive level (thanks to Salesforce’s top-notch RAG engine).
In the back-end, the engagement feed (right-hand panel) shows the real-time signals (customer behavior, preferences, and intent) that Agentforce Personalization Decisioning has utilized.
Agentforce Web Curation
With the temptation to ‘bounce’ from your website and use another LLM (e.g. ChatGPT) to get their product questions answered, not only have you lost the captive audience, but LLMs could output recommendations for competitors.
Similar to Personalization Decisioning, Agentforce Web Curation makes changes to your website, such as rendering recommendations, changing images, and even the tone of the messaging.
In short, changes in web pages are influenced by real-time user behaviour signals that are packed into an SDK (software development kit) and combined with other organizational data stored in Salesforce.
Agentforce Lead Management/Qualification
This is perhaps the most interesting one, in terms of driving revenue – not to mention taking a step in the right direction to solve the age-old marketing and sales alignment challenge.
Imagine having a BDR working 24/7 (and not complaining about the lead quality!). This objective lead assessment means that prospects are less likely to fall through the gap between marketing and sales.
In the image below, we see the campaign view in Marketing Cloud Next. This is an attractive summary, with:
- A breakdown of the number of contacts per status.
- A timeline (right-hand panel) of how Agentforce has interacted with/had response from the selected contact.
Guided Agent Setup, a platform-wide feature that will be GA July ‘25, is accessible from Marketing Cloud Next. The 5-step path is outlined on the left-hand panel. Highlights here include:
- Agentforce Lead Management set up starts with assigning the settings to an AI agent, which acts like a user.
- Email settings include writing the primary value proposition (what the agent should know), followed by nudge settings, as in, the maximum number of outreach attempts the agent is allowed to make.
- Product Knowledge involves defining the data types and files that the AI agent can utilize during its outreach.
- Prospect Assignment is the logic that assigns leads/contacts to an agent. Several conditions can be added.
- Throughout, you have an opportunity to test out the scenarios by picking a specific lead or contact record, seen on the right-hand panel.
Agentforce Segment Intelligence
Segment Intelligence has been around for some time, and continues to be a key, now as a part of Marketing Cloud Next.
The way that the Marketing Intelligence app has been designed is that analytics can be converted into action smoothly. In the image below, we see a performance overview chart, with the Agentforce panel open on the right-hand side. Before this point, the human user has asked the agent about underperforming ads. Note the hyperlinked actions the user has been recommended within the agent’s response.
By taking the agent’s recommendation (reviewing the Segment Intelligence dashboard), the user can continue to interrogate the performance data – in this example “show me which city has a high conversation rate and low CPC for these segments”. A new visual of the drilled-down data is presented.
If the human user gives the green light, the agent will prepare the desired segment, presented in a clear interface to be reviewed and/or amended. Each segment also has a Segment Performance page, with metrics and visualizations relating to engagement and demographic data.
Summary
Marketing Cloud Next has multiple underlying factors that makes this offering not just a rehash of existing product features in the Salesforce marketing space – how it caters to both B2B and B2C models, its rebuilt personalization engine, and the move to minimize integrations. Plus, with the same view of the customer across all areas of the business, all departments are looking at the same analytics – in other words, no more doubt internally of “are we looking at the same data?”.
Marketing Cloud Next is certainly one to watch. Now that Salesforce has showcased their vision for the future of Marketing Cloud, in time, adoption will grow and innovation of agent-first features will accelerate with this solid underlying infrastructure.
Availability:
- Marketing Cloud Next is now generally available
- Agentforce Campaign Creation is now generally available
- Unstructured data in Data Cloud is now generally available
- Agentforce Personalization Decisioning is now generally available
- Agentforce Lead Management is now generally available
- SDR Agent is now generally available
- Agentforce Paid Media Optimization is now generally available
- Segment Intelligence is now generally available
- Marketing Cloud+ SKUs are planned for July 2025
- Agentforce Web Curation is planned for October 2025
- Expanded Collaboration with LinkedIn is planned for November 2025
