The Gap Between Salesforce's Vision and Reality

Attending Salesforce World Tour London this year left me with a mixture of admiration, curiosity and, perhaps unexpectedly, concern. Like many attendees, I spent the day moving between keynote sessions, product demonstrations, partner conversations and customer discussions. Regardless of where the conversation started, it almost always ended up in the same place. Artificial intelligence dominated the event. Agentforce was everywhere. Autonomous agents, digital labour, conversational interfaces and AI-driven business processes featured heavily throughout the day. Salesforce's strategic direction could not have been clearer. The company is investing heavily in an AI-first future and is determined to position itself at the centre of the next major transformation in enterprise technology. From a business perspective, this approach is entirely understandable. The technology industry is moving at extraordinary speed. Competitors are investing billions into artificial intelligence. Customers are demanding innovation. Investors expect growth. In such an environment, standing still is not an option. If Salesforce wants to remain a market leader over the next decade, it must be at the forefront of the AI revolution rather than attempting to catch up later. Yet as I listened to the presentations and watched the demonstrations, I found myself reflecting on a question that seemed largely absent from the discussion. While the technology itself was impressive, and in many cases genuinely exciting, I could not help wondering how many Salesforce customers are actually ready to embrace the future being presented to them. Perhaps that reaction comes from the fact that I spend my time in a very different part of the Salesforce ecosystem. Much of the AI conversation is understandably being driven by architects, developers, product strategists and technology leaders. My world, however, is largely populated by sales representatives, service agents, field teams, managers, administrators and product owners. These are the people who ultimately determine whether Salesforce succeeds or fails within an organisation. They are the individuals expected to use the platform every day, follow the processes that have been designed around it and generate the data upon which the entire system depends. When viewed through that lens, the gap between the future being demonstrated on stage and the reality experienced by many organisations feels larger than it has for quite some time. One area that particularly caught my attention was Headless 360. For those unfamiliar with the concept, Headless 360 represents a future in which AI systems such as Claude, ChatGPT or other intelligent agents can interact directly with Salesforce. Rather than opening Salesforce, navigating through records and manually executing actions, users can theoretically retrieve information, update records and perform business activities through conversational interfaces. It is a compelling vision, particularly for those who spend large amounts of time searching for information, analysing data or interacting with complex systems. The enthusiasm surrounding this concept is easy to understand. The demonstrations are impressive. Watching an AI retrieve customer information, analyse opportunities, generate recommendations and execute actions without ever opening Salesforce creates the impression of a significant leap forward in user experience. From a technical perspective, there is no question that the capability is remarkable. However, technology demonstrations and operational reality are often very different things. Over the years, I have worked with organisations across a wide range of industries, maturity levels and Salesforce implementations. Some have had relatively simple environments. Others have operated highly customised platforms supporting thousands of users. Despite those differences, a consistent pattern tends to emerge. The challenges organisations face are rarely caused by a lack of technology. More often than not, they stem from issues relating to adoption, confidence, capability and process execution. Users struggle to maintain data quality because they do not fully understand the importance of the information they are entering. Managers lose confidence in reports because processes are being followed inconsistently across teams. Sales representatives revert to spreadsheets because they find certain activities easier outside the system. Customer service teams create workarounds because training was rushed or reinforcement never occurred after go-live. These challenges are remarkably common and persist across organisations of all sizes. Against that backdrop, I find it difficult to believe that the majority of users are suddenly going to abandon traditional Salesforce interfaces in favour of conversational AI experiences. Consider a typical field sales representative. They are visiting a customer and discussing previous orders, current opportunities, support issues and future plans. They may be operating from a mobile device and, depending on the location, may not even have reliable internet access. What they need is immediate access to relevant information presented in a clear and structured format. They need confidence that the information they are viewing is accurate, and they need the ability to update records quickly as the conversation progresses. In that scenario, Salesforce Mobile App makes perfect sense. The information is organised. The navigation is familiar. Offline capabilities can support users when connectivity is limited. Most importantly, the user can see the context surrounding the customer relationship rather than receiving information one question at a time. Could an AI assistant provide value in that situation? Absolutely. It could summarise previous activity, highlight risks or identify cross-sell opportunities. However, those capabilities feel most valuable when embedded within the existing experience rather than replacing it entirely. And the key challenge is, most conversational AI platforms rely on an internet connection. So in this scenario, moving a business towards Headless 360 is a near impossible task. A similar observation applies to customer service environments. Imagine a contact centre agent receiving a call from a customer with a long and complicated history. The customer has multiple open cases, previous escalations and numerous interactions spread across different channels. An experienced agent wants to understand the situation as quickly as possible. They need to scan notes, review timelines, identify patterns and determine the appropriate next step. Human beings are remarkably effective at processing visual information when it is presented well. A service console that displays relevant information in a structured and accessible format allows an agent to develop situational awareness within seconds. While AI-generated summaries could undoubtedly enhance that experience, replacing the interface with a conversation may not necessarily improve it. In fact, in many situations it could introduce additional complexity. This is perhaps the aspect of the current AI narrative that concerns me most. There appears to be an assumption in some quarters that conversational interfaces are inherently superior to traditional applications. While that may be true for certain use cases, it is far from universally true. The best interface is not always a conversation. Sometimes it is a dashboard. Sometimes it is a report. Sometimes it is a list view. Sometimes it is a carefully designed screen that allows users to absorb large amounts of information quickly and efficiently. The reality is that many organisations are still trying to master fundamental Salesforce adoption challenges. Throughout my work, I continue to encounter businesses that struggle with user engagement, data quality, process compliance and reporting accuracy. These are not niche issues affecting a small minority of organisations. They are widespread challenges that have existed within the Salesforce ecosystem for many years. This is why I often find myself questioning the practical readiness of many businesses for technologies such as Agentforce and Headless 360. While the vision is compelling, the operational foundations required to support that vision are frequently missing. Artificial intelligence depends upon reliable data, clearly defined processes and strong governance. Without those foundations, organisations risk amplifying existing problems rather than solving them. A poorly designed process does not become a good process simply because AI is involved. Poor quality data does not become trustworthy because an AI assistant can access it. Users who lack confidence in Salesforce today are unlikely to become more confident simply because another layer of technology has been introduced. In many cases, the opposite may occur. Additional complexity can create additional confusion, particularly for users who are already struggling to understand the platform. This does not mean that Salesforce is moving in the wrong direction. In fact, I believe some of the most compelling use cases for these technologies lie outside the frontline user scenarios that often dominate marketing demonstrations. Developers, administrators and operations teams stand to gain significant value from conversational access to Salesforce environments. The ability to analyse metadata, investigate dependencies, review configurations and interact with complex datasets through natural language could dramatically improve productivity for technical users. Similarly, there is considerable potential in allowing AI agents to perform routine background activities. Data cleansing, categorisation, monitoring, enrichment and process automation are all areas where intelligent agents could deliver meaningful efficiency improvements without fundamentally changing how employees perform their day-to-day work. These are practical use cases that align closely with genuine business challenges and offer clear return on investment. However, even within these scenarios, organisations must still confront an important reality. Every new capability introduces new costs. Those costs are not limited to licensing or technology investment. They also include training, governance, support, change management and ongoing administration. As organisations evaluate AI initiatives, they must consider not only what the technology can do but also what it will take for employees to use it effectively and responsibly. This is an area where I believe the industry conversation remains surprisingly immature. There is enormous focus on capability and relatively little focus on readiness. Businesses are being encouraged to imagine autonomous agents and AI-driven workflows while many continue to struggle with adoption rates, inconsistent processes and underutilised platform investments. Before organisations can successfully embrace the future, they must first ensure they are extracting value from the systems they already own. As I reflected on the event during my journey home from London, I found myself increasingly convinced that the future Salesforce is building and the reality many customers are living currently exist on different timelines. Salesforce is understandably focused on where the market will be in five or ten years. Most customers, however, are focused on solving the operational challenges they face today. They are trying to improve adoption, increase productivity, strengthen reporting and build confidence in their existing systems. For that reason, I do not believe the future is one in which users abandon Salesforce in favour of conversational AI interfaces. Nor do I believe that technologies such as Headless 360 will become mainstream overnight. What seems far more likely is a gradual evolution in which AI becomes an increasingly valuable layer sitting alongside existing experiences. Salesforce remains the system of record. The interfaces remain familiar. Users continue working within established processes. Meanwhile, AI assists by reducing administrative effort, identifying insights, highlighting risks and recommending actions. That future feels both powerful and realistic because it enhances human capability rather than attempting to replace it. Salesforce deserves credit for investing boldly in innovation and pushing the boundaries of what enterprise software can become. The technologies being developed today may eventually transform how organisations operate. Yet for many customers, the immediate challenge is not reaching the future. It is mastering the present. Until organisations improve adoption, strengthen capability and establish reliable operational foundations, technologies such as Headless 360 may remain more relevant to developers and operations teams than they are to the average salesperson, service agent or manager. The future being presented is undoubtedly exciting. The question is whether most customers are ready for it. Based on what I continue to see across the organisations I work with, that remains a far more open question than many in the industry appear willing to acknowledge.