Last month, Salesforce introduced Agentforce 3, the latest release of its hallmark AI agent platform.
However, this announcement notably focused less on the platform and more on the ecosystem that Salesforce is building around it.
For starters, the company integrated AgentExchange directly into Agentforce and announced agent interoperability, adding open standards, like Model Context Protocol (MCP).
The open standards equip Agentforce customers to chain first- and third-party AI agents, so they collaborate and automate multi-step, cross-functional processes.
Yet, the CRM leader went further in its ecosystem shift by releasing a new Command Center, which gives businesses visibility of company-wide agent deployments.
In doing so, the addition empowers business leaders to understand existing AI agent deployments, orchestrate those cross-system workflows, and track success rates, ROI, and more.
The Significance of Salesforce’s Pivot from Platform to Ecosystem
The introduction of the Command Center opens up a new level of visibility, ensuring that workflows not only enter the Agentforce platform correctly but also exit and integrate properly with the broader ecosystem.
Moreover, the Command Center shows an evolution in Salesforce’s messaging. Since September 2024, the company has encouraged rapid agent development. Now, it’s recognizing the operational and financial complexities businesses are facing.
In other words, Salesforce now caters to the grownups in the room, asking: “Should we be automating this? And if so, how?”
Such moves are critical as Salesforce saw huge initial interest in Agentforce, which quickly plateaued once business leaders started to ask these questions.
With the competition catching up in platform capabilities, it’ll need to go further in helping brands answer them. Shiny agents alone won’t catch the eye like they did 12 months ago.
Salesforce Follows ServiceNow’s Lead
The Agentforce Command Center is comparable in functionality to the ServiceNow Control Tower, released at Knowledge 2025.
Upon that launch, Jensen Huang, CEO of NVIDIA, said: “ServiceNow is destined to be the best platform, the operating system of enterprise AI agents.
The reason for that is… AI agents are going to be curated, managed, and improved by the IT department. ServiceNow is already in every company’s IT department.
Salesforce is challenging Huang’s prediction with its Command Center and pivoting toward a similar ecosystem focus. It has also confirmed a move into IT Service Management (ITSM).
That news came after ServiceNow pushed further into Salesforce’s realm by releasing a unified CRM platform, created to disrupt an “industry built on outdated systems”.
In doing so, ServiceNow made a lot of noise and may have forced Salesforce CEO Marc Benioff to regret his earlier comparison of his company vs. ServiceNow as McDonald’s vs. Wienerschnitzel.
Nevertheless, Salesforce does have advantages in the AI agent space over ServiceNow. For instance, its position at the forefront of the business intelligence (BI) space may prove impactful.
After all, as Command-Center-type solutions evolve, they’ll eventually take enterprise data and spin up entirely new dashboards to monitor AI agent performance based on written prompts. With its agentic analytics move, Salesforce is well-placed to lead that future, ensuring a deeper level of reporting and, crucially, governance.
However, as Huang noted, ServiceNow’s dominant position in ITSM and high advocacy amongst IT teams make it a formidable opponent.
Of course, others are releasing similar Command Centers and Control Towers, from IT stalwarts like IBM to disruptors like Boomi. Yet, Salesforce vs. ServiceNow seems to be a key battle, not just in terms of AI agents but much more broadly.
What Does the Rise of Command Centers / Control Towers Mean for CX Teams?
In the age of AI agents and multi-agent workflows, customer experience becomes a team sport that includes service, sales, marketing, commerce, and more.
As Liz Miller, VP & Principal Analyst at Constellation Research, said:
If all those functions build the same agent from different data sources, we’re in trouble.
Unfortunately, that’s still the approach within many organizations. CX teams all use the same data but pull from different repositories. That’s where this idea of functional dominance: “I own this thing, so I get to control it,” gets companies into real data trouble.
As AI agents and Control Towers rise, teams need to unite, share data, and build a centralized view of customer workflows, so AI agents don’t fall down when moving across the front office and beyond.
Miller concluded: “We need to be having strategic conversations with everyone involved, asking: “Where is the handoff? Where does this all connect?” Because in the end, one system isn’t going to dominate another. Our customers don’t care who owns what, they care that it works.”
Struggling to keep up to speed with all the latest Agentforce announcements? Subscribe to the CX Today Newsletter and stay tuned for the next episode of CX Today’s Big News Update, which unpacks everything Agentforce 3 and features Miller alongside other prominent industry analysts.