Increasingly, and understandably, registered investment advisors (RIAs) are asking important questions about the role of artificial intelligence (AI) in wealth management and Agentforce: Are we behind? How are other wealth firms using Agentforce for their business? How are individual advisors using AI with their clients? What’s the current adoption of Agentforce?
As we kick off 2026, I thought it would be beneficial to discuss the current state of Agentforce (agentic AI from Salesforce) with some real statistics and in the context of wealth management.
Agentforce Adoption Thus Far
Straight from Salesforce’s F3Q26 (Oct-Q) results (i.e., Dec 2025), Salesforce has closed 18,500 Agentforce deals since its launch in September 2024. Out of those deals, 6,000 came in 3Q vs. 4,500 in 2Q (note Salesforce’s fiscal year ends in January, not December). What might be surprising is that only 9,500, just 51%, were paid deals. Salesforce either gave Agentforce away to entice customers to try the product, or the “deal” was really a free trial.
How important is Agentforce to Salesforce’s bottom line? Today, Agentforce represents $540 million, around 1%, of the annual recurring revenue (ARR) for Salesforce. I will say, as a Salesforce Partner for 15 years, I’ve never seen a product evolve so quickly — nor have I seen Salesforce put so much effort into helping Salesforce Partners ramp up their ability to deliver Agentforce. It’s already a very capable product in my opinion, but we must remember we are still early in this journey.
Is Agentforce being deployed in the real world?
How is the Agentforce adoption going so far? As of October 2025, per Cleveland Research, less than 10% of Salesforce customers with Agentforce have moved past “proof of concept” (POC) to broader production deployments. My take on this is that customers don’t want to fall behind on the technology, but they are also unwilling to risk their customers having a bad experience.
A recent IBM study titled The State of Salesforce 2025–2026 stated:
“For 1,200+ Salesforce customers surveyed, only 33% of AI initiatives are meeting ROI targets. Even more concerning: 72% have failed to scale across business units, and 20% have stalled, failed outright, or been abandoned.”
That same report goes on to say:
“62% of Salesforce customers express concern about unpredictable AI-related costs, and 64% report that unclear total cost of ownership makes it difficult to determine whether AI agents will save money or quietly inflate budgets.“
Salesforce has tried to tackle the cost transparency problem head-on by releasing consumption calculators and more flexible pricing models.
Despite the hype, we’re still in the early days of AI, and firms are responding accordingly. As stated in the IBM study, “executive confidence in using AI to transform customer and employee experiences remains frozen at last year’s level, 16%,” while 17% in 2025 don’t know where to start, and 20% are interested but concerned about the implications of using AI in their business.
How is Agentforce being used? Customer self-service and call centers lead the way.
The main AI use case for AI deployment continues to be call centers, as reported by Cleveland Research:
“Customers [are] primarily focusing on simple/few use cases with minimal examples of end-to-end process automation via agentic; top use cases are customer support (FAQ, chat, and now voice), internal employee support, and sales development/lead-gen.”
The IBM study correlates this “in areas where AI can directly impact both the top and bottom line. Customer self-service leads the way [at] 70%.”
Agentforce Is Only an Option for the Largest RIAs
While customer service and call centers may have the broadest business potential for AI, the most probable use cases for wealth management include internal activities such as client meeting prep, creating client summaries, personalizing client communications, and predicting client attrition risk. This all sounds great, but is it currently achievable?
Unfortunately, though increasingly capable, Agentforce is too onerous for the vast majority of RIAs. To deploy Agentforce, wealth firms require either a large, skilled information technology (IT) department or help from a Salesforce consulting partner. According to the Investment Adviser Association (IAA), 92.7% of RIAs have 100 or fewer employees, 88% have 50 or fewer, and 59% have 10 or fewer. If you are one of 9,316 RIAs with 10 or fewer employees, you likely don’t even have a full-time employee dedicated to IT, so Agentforce wouldn’t be an option.
Access to Quality Data Is Key for AI Success
To perform well, AI needs quality data. The IBM study points out that only 26% of the customers surveyed had most of their data in Salesforce, which means three out of four customers are not getting a 360-degree view of their customers using only their customer relationship management (CRM) platform. The study also observes that, at 53%, “poor data availability and quality is the leading adoption barrier.” This is a key point because, if your users are not consistent in how and where they store data in Salesforce, Agentforce can’t leverage it, potentially contributing to hallucinations caused by poor data.
To that point, when RBC demonstrated an AI solution at Dreamforce in 2024 to generate household summaries, one of their big takeaways from working with advisors was the necessity to keep their data clean and up-to-date. Generative AI only works if the data is where it’s expected to be.
Ideally, a wealth firm would have a data strategy in place to unify all of its customer data. That further pushes likely adopters to probably the top 500 RIAs — those that have the revenue and scale to have a data warehouse or equivalent in place. The top three tech platforms used by RIAs are CRM, financial planning, and portfolio management, in that order. Data needs to be unified, accessible, and accurate for Agentforce to provide value and insights. Though there are multiple options to get custodial financial account data into Salesforce, options for financial planning (i.e. MoneyGuide Pro and eMoney) continue to be do-it-yourself solutions, which are simply not an option for smaller firms that lack sufficient IT expertise and resources.
Agentforce Templates Included in Financial Services Cloud — It’s a Start
Didn’t we see spiffy demos at Dreamforce last year with pre-configured agents tailored for wealth management? Yes, there are a couple: “Wealth Advisor Client Meeting Preparation” and “Wealth Advisor Post Meeting Follow-Up.” Note that these are currently only available in the Financial Services Cloud (FSC) CORE version (i.e., not the widely currently deployed managed package version of FSC), and firms must store information where the templates are programmed to search for it — not in a custom object you created. These templates can serve as a starting point, but they must be configured for your business and require ongoing maintenance and fine-tuning.
The Rise of AI Off-the-Shelf Point Solutions
A recent F2 Strategy survey asked RIAs, “Are you working on any AI projects currently (or recently completed)?” Surprisingly, 78% (RIAs, hybrid RIAs, and multifamily offices) answered yes, marking a 23% increase in two years. What is unclear from that statistic is whether this was an IT-driven integrated solution (i.e., with compliance aware of what data is being passed to a large language model [LLM]) or something more rogue, like advisors using individual licenses of ChatGPT, Claude, or the like.
As an anecdotal data point, I attended a wealth management conference on mergers and acquisitions (M&A) at the end of January, and the audience was asked who thought AI could disrupt their business in the next three to five years. The show of hands had to be close to 100%. AI unequivocally was on everyone’s minds.
If Agentforce seems to be out of reach for the vast majority of RIAs today, what are wealth firms using for an integrated AI solution? We are seeing traction with off-the-shelf point solutions such as Jump.ai, Zocks, and Zelpyn because they are easier to implement and use, can meet some level of compliance, and have a predictable cost. The downside is that you can’t create a custom AI solution like you can with Agentforce, but the upside is its strength in enabling firms to tackle very defined use cases without the need for heavy IT assistance.
The AI Great Divide
As we kick off 2026, we find ourselves with a great divide in wealth management — those that have the scale and resources to utilize Agentforce and those that are limited to off-the-shelf AI solutions. Though very capable, Agentforce is an enterprise-level solution suited for customers with a data strategy in place and a forward-focused commitment to unifying their customer data.
No one wants their business to fall behind or seem out of touch with AI, potentially losing a competitive advantage. My take, based on the data presented above, is that companies want to have access to AI tools to understand current capabilities, but few are willing to risk the client experience. It will be interesting to see whether, a year from now, the numbers start to swing toward true deployments rather than just POC. Watch this space!
We Can Help with Your Salesforce Journey
Want to understand what AI for wealth management looks like for your firm? Connect with ShellBlack to evaluate your readiness for AI in Salesforce and define a roadmap that matches your business — not the hype.
Note: AI was not used to write this blog post. 😉
Author Credit: Shell Black, president and founder, ShellBlack