We’ve seen the Salesforce ecosystem debating the adoption of Agentforce over the last year or so regarding product readiness, pricing, technical capability, security, and perhaps most importantly, customer appetite. It’s no secret that not everyone has warmed to Salesforce’s flagship AI product. Despite the mothership’s extensive attempts to tell and show us otherwise, a lot of people are still unconvinced, not yet seeing a successful Agentforce implementation that convinces the masses they should adopt.
But recently, a conversation with a Salesforce industry expert got me thinking about the issue from another angle. During a discussion around AI adoption hurdles, they suggested that one of the biggest barriers may be less about the technology than the people making decisions about it.
Now, whether you agree with that statement or not, it raises an interesting thought that I wanted to look at more closely. Is Agentforce adoption really being held back by technical limitations, pricing concerns, security fears, or immature use cases, or are those simply the reasons we talk about because they’re easier than confronting a potentially deeper issue?
After speaking with more Salesforce pros, a common theme emerged on the ground. Many of the challenges being blamed on AI today existed long before Agentforce even arrived – unclear processes, fragmented ownership, low trust between teams, and businesses that can’t agree on what success even really looks like. So, is the real challenge actually what AI could expose?
The Usual Agentforce Blockers Don’t Tell the Full Story
If you asked a reluctant Salesforce leader why they’re moving cautiously on AI, it’s very likely they’d point to security, compliance, budget, limited resources, or unclear ROI as key issues. They could even mention how the technology is moving too quickly or that the market needs more time to mature before they make a move.
These are all perfectly valid concerns, and in many cases, valid, as we’ve discussed at length on SF Ben. AI (especially Agentforce) is evolving faster than many can keep up with, and few organizations want to make a significant investment only to discover they’ve backed the wrong technology or strategy.
But these objections may only be telling one part of the story. For Alex Borland, an independent Salesforce Consultant, the challenge is less about whether the technology can theoretically work, and more about whether businesses know what they want it to really do for them.
“I don’t hear from my clients that the technology isn’t capable,” he told SF Ben. “I think it’s what they’re trying to do is understand how it fits within their business, especially large-scale businesses that you’re dealing with. How do you put AI within there? What kind of processes should AI own? How do you govern it? And who’s overall accountable for that actual process?”
The last point in particular is important – once AI moves from a demo environment to a live business process, the conversation changes to who’s accountable. Who owns the agent, or trusts the output? Who takes responsibility when the AI gets it wrong?
Josh Grace, Executive Director of Enrollment Technology at IWU National & Global, has seen a similar pattern emerge on his side. In his view, the stated blockers – security, budget, staff capacity, etc – are often real, but not always the true conversation being had.
“Sit through enough of these, and you notice the AI conversation is rarely the real one,” he said. “The tool gets blamed. It’s the easiest thing in the room to point at. But AI didn’t create those questions. It exposed them.”
In other words, AI and Agentforce may actually be forcing businesses to confront poor processes and trust issues that already existed before this conversation arrived.
AI Exposes the Processes Business Never Fixed
Now this is where this conversation may start to get uncomfortable. But some businesses may not yet understand the process it really wants to automate, and that is not attached to the usual reasons behind not adopting AI or Agentforce.
As we know, Agentforce agents are supposed to act, recommend, or make a decision within a fixed framework that a business has (hopefully) already created. If that framework isn’t clear, then you’ve got a very confused and useless agent on your hands.
Josh mentioned how he saw this pattern long before the conversation reached AI. “I’ve watched teams ask for predictive analytics before agreeing on which student outcomes actually matter,” he said. “I’ve watched leaders ask for AI-powered communications before anyone mapped the journey a student is already on. I’ve watched a dashboard get requested while every team pulls numbers from a different system and defines success in a different way.”
Many organizations already have different departments using Salesforce in different ways, different definitions for the same metric, and different assumptions about who owns each stage of a customer or student journey. In a traditional CRM, those gaps can survive for a long time through manual workarounds, Slack messages, and institutional memory, but Agentforce obviously makes that much harder to hide or put a band-aid over.
As Josh put it: “When the advising workflow lives in spreadsheets, three different systems, a stack of reports, and a few people’s memory, AI doesn’t clear up the confusion. It speeds it up.”
Alex made a similar point, stating that one of the biggest mistakes a business can make is treating AI as something that can simply be added on top of existing ways of working. “AI encourages businesses to rethink the processes themselves,” he said. “How should work be done? Can we get an AI agent to do this work, or do we need a human element?”
That is a very different question from “Can Agentforce do this?” It asks whether the business knows enough about the process, the data, the risk, and the desired outcome to let AI work safely. This, in itself, may be the real leadership challenge right now.
Final Thoughts: Agentforce May Be a Mirror, Not a Magic Wand
There is still a version of this story where Agentforce adoption accelerates significantly (soon enough). The company’s recent quarterly earnings showed that Agentforce’s business, on paper, is showing success – surpassing $1B in revenue.
Salesforce can simplify pricing further, publish more customer success stories, improve confidence around security, and continue maturing the product. Many customers may simply need more time, more evidence, and more practical examples before moving from interest to investment. But even if all of that happens, the leadership question doesn’t just disappear.
For Alex, history is unlikely to say the biggest problem was that the technology simply was not ready. Instead, he believes the slower-moving part will be the organization itself. “Technology changes quickly within weeks, months or years, and we’ve seen that throughout the last few years,” he said. “Some organizations, they can take years to change certain processes.”
That gap between product speed and organizational speed may define the next phase of Salesforce AI adoption. Salesforce can release new Agentforce capabilities quickly, but businesses still need to work out what AI should own, what humans should own, what governance is required, and where trust can realistically be placed.
Josh is also skeptical that time alone will solve the problem and does not believe a new generation of leaders will automatically make AI adoption easier. “The problem isn’t whether someone knows how to use ChatGPT, Copilot, Claude, or whatever comes next,” he said. “The problem is organizational clarity.”
That is why the generational argument only gets us so far. Younger leaders may be more comfortable experimenting with AI, and older leaders may bring more experience in other areas of the business. But neither group can shortcut the harder work of defining outcomes, clarifying ownership, and building trust.
Josh’s long-term view is perhaps the most useful one: “Ten years from now, the institutions making progress won’t be the ones that waited for better technology. They’ll be the ones that did the slower work first. They mapped the journey, clarified ownership, agreed on outcomes, and built trust. Then they applied AI.”
That may be the real lesson for Salesforce customers from this discussion. Agentforce is asking whether leaders are ready to look clearly at how their business actually works, and not just whether they’re ready to adopt AI, which could prove to be the biggest implementation challenge of all.
