The gap between promise and reality: What Salesforce's struggle reveals about AI transformation in the tech industry

The Great AI Disillusionment: Why Salesforce Shows That Reality Looks Different

The spectacular 27 percent drop in the share price of CRM giant Salesforce since the beginning of 2025 is not an isolated phenomenon of a single company. Rather, it symbolizes a fundamental discrepancy between the high expectations of artificial intelligence and the harsh reality of its commercial exploitation. While tech companies around the world are proclaiming the revolution brought about by autonomous AI agents, Salesforce's situation reveals three central problems that could be symptomatic of the entire industry: the monetization of AI innovations, the structural maturity of the enterprise software market, and the increasing complexity of technology integration. This analysis examines what really lies behind this supposed promise of the future and what consequences it has for the tech industry.

Suitable for:

• Artificial intelligence: Why Salesforces Agentforce not (yet) started - independent alternatives are better

Basics and relevance

Salesforce's situation in October 2025 marks a turning point in the perception of artificial intelligence as a direct growth driver for established tech companies. Marc Benioff, the charismatic founder and CEO of the customer relationship management company, proclaimed the era of agent-based AI at the company's Dreamforce conference in San Francisco. His vision: Autonomous algorithms would replace human employees in companies and become Salesforce's most important revenue generator. The reality, however, paints a different picture.

The dramatic decline in Salesforce stock stands in stark contrast to the general trend in the tech industry, where technology stocks have recorded significant gains over the same period. This divergence raises fundamental questions: Has the industry overestimated the speed with which artificial intelligence can be translated into real revenue? Are expectations for autonomous AI agents realistic? And what structural problems lie behind the shiny facade of AI's promise.

The relevance of this analysis extends far beyond Salesforce. It affects all companies that rely on artificial intelligence as a key growth driver. It touches investors who are pumping billions into AI technologies. And it touches on workers whose jobs are threatened by the promised automation. The Salesforce case offers a unique insight into the mechanisms, hopes, and disappointments of an industry in transition.

This article is divided into eight sections that systematically present the historical roots, technical mechanisms, current status, practical use cases, critical issues, future developments, and a final synthesis of the lessons learned. It will become clear that Salesforce's challenges are representative of deeper industry problems that extend far beyond a single company.

From cloud pioneer to AI fighter: The strategic reorientation of an industry giant

To understand the current situation, one must trace the origins and evolution of Salesforce. Founded in 1999 by Marc Benioff, the company revolutionized the software industry with a then-radical concept: Software as a Service. Instead of selling expensive license packages that had to be installed on customers' servers, Salesforce offered its CRM solution online. Customers paid a monthly fee and could use the software simply via their browser.

This innovation made Salesforce the market leader in customer relationship management. With a market share of over 21 percent, the company still dominates the global CRM market today, far ahead of competitors such as Microsoft, Oracle, and SAP. For over two decades, Salesforce was considered a growth stock par excellence. Revenue grew by double digits year after year, its share price climbed continuously, and the company expanded through numerous acquisitions.

But already in the years leading up to 2025, the first signs of a slowdown appeared. Growth in the CRM software industry as a whole slowed as the market became increasingly saturated. Many large companies had already implemented CRM systems, and the low-hanging fruit had been picked. At the same time, new competitors emerged, gaining market share with innovative approaches and lower prices.

In this situation, Benioff increasingly focused on artificial intelligence as a new growth story starting in 2022. Salesforce first introduced Einstein, an AI platform that enabled predictive analytics and automation within its existing CRM products. Then, in September 2024, the big announcement followed: Agentforce, a platform for autonomous AI agents that would independently perform tasks in areas such as customer service, sales, and marketing.

The vision was ambitious: By the end of 2025, customers would create one billion autonomous AI agents via the platform. These agents would not only answer simple queries but would also independently plan and execute complex, multi-step tasks. They would act proactively, make decisions, and access the company's entire database.

At the same time, Salesforce invested heavily in the technological foundation for these AI agents. In May 2025, the company announced the acquisition of Informatica, a data management specialist, for $8 billion. The acquisition was intended to ensure that the AI ​​agents have access to high-quality, well-structured data. In the fall of 2024, Salesforce had already acquired Own Data, another data management company, for $1.9 billion.

But despite these massive investments and the grand vision, the hoped-for revenue surges failed to materialize. In the second quarter of the 2025/26 fiscal year, Salesforce's revenue grew by 9.8 percent to $10.24 billion. While this was slightly above expectations, it was the fifth consecutive quarter of single-digit growth. The outlook for the coming quarter was even more cautious, fueling concerns that the AI ​​offensive would not deliver the hoped-for commercial success.

The anatomy of autonomous AI agents: Technology between vision and feasibility

To understand why monetizing AI agents is proving so challenging, it's important to examine the technical foundations and mechanisms of these systems. Agentforce is based on several technological components that must work together to achieve the promised autonomy.

At its heart is the Atlas Reasoning Engine, which acts as the neural network or brain of the AI ​​agents. This engine is designed to mimic human thought and behavior, correctly categorize tasks, prioritize task steps, and ultimately execute them correctly. Unlike previous AI assistants like Copilot, which relied heavily on human interaction, Agentforce agents are designed to operate largely autonomously.

The second key component is the Salesforce Data Cloud, which harmonizes all relevant company data in real time and makes it available to AI agents. The quality and completeness of this data is crucial for agent performance. This also presents one of the biggest challenges: Many companies have collected their data for years in various systems without consistent standards or regular cleansing.

The third component is integration tools like MuleSoft and pre-built connectors that enable agents to interact with existing workflows and external systems. These interfaces allow agents to operate not only within the Salesforce world but also communicate with other enterprise applications.

In addition to these Salesforce-specific components, Agentforce also integrates large-scale language models from third-party providers such as OpenAI, Anthropic, and Google Gemini. These models provide the underlying natural language processing and general world knowledge upon which specific agents are built.

The functionality can be illustrated using the example of a customer service agent: A customer contacts the company with a query. The agent analyzes the query, accesses the relevant customer data from the Data Cloud, compares it with similar cases from the past, develops a multi-step resolution plan, executes these steps, and communicates the result to the customer. All of this happens without human intervention, unless the agent encounters a problem that exceeds their capabilities.

This sounds impressive in theory. In practice, however, there are numerous stumbling blocks. Agents are only as good as the data they have access to. If the data is incomplete, outdated, or inconsistent, agents make incorrect decisions. Integration into existing corporate systems is often complex and requires considerable effort. And configuring agents, although advertised as a low-code process, still requires considerable technical understanding and Salesforce-specific know-how.

Another problem is a lack of trust. Many companies are hesitant to hand over control of critical business processes to autonomous agents without robust testing procedures and security mechanisms. The risk of errors, data breaches, or undesirable behavior is real, as examples from other industries show.

The difficult path to profitability: Three fundamental challenges

Salesforce's problems can be summarized in three key challenges that are typical for the entire industry: the monetization of AI innovations, structural market readiness, and the complexity of technology adoption.

The first challenge concerns monetization

Although Salesforce has developed a technologically advanced product with Agentforce, the key question remains: How can it monetize it? Agentforce's pricing model is based on two dollars per conversation, a usage-based approach that differs from traditional licensing models. However, many potential customers are hesitant to deploy this technology on a large scale until the return on investment is clearly demonstrable.

The costs of running AI agents are significant. The underlying large language models require expensive computing resources. According to industry estimates, a single query to a generative AI model costs up to ten times more than a traditional Google search. These costs must be passed on to customers, limiting price acceptance. At the same time, customers expect AI agents to deliver clear value that justifies the higher costs.

To date, only approximately 12,000 companies use Agentforce, a tiny number considering Salesforce's vast customer base of several hundred thousand companies. Annual recurring revenue from Agentforce is less than $500 million, a fraction of the total revenue of over $40 billion. Even if this number triples or quadruples in the coming years, as Salesforce hopes, its contribution to total revenue would still be limited.

The second key challenge is the structural maturity of the CRM market

After two decades of strong growth, the customer relationship management software market has reached a saturation phase. Most large and medium-sized companies in developed markets have already implemented CRM systems. The potential for organic growth through new customer acquisition is limited.

At the same time, competition has become more intense. Microsoft with Dynamics 365, Oracle with its cloud applications, SAP with its CRM solutions, and numerous specialized providers such as HubSpot, Zendesk, and Zoho are all vying for market share. These competitors have caught up in recent years and are sometimes offering cheaper or more specialized solutions.

In this environment, it will be more difficult for Salesforce to achieve double-digit growth rates, even with innovative AI features. Customers won't simply switch their CRM system because a vendor offers new AI capabilities. Implementing a CRM system is complex, expensive, and time-consuming. Companies are reluctant to switch as long as their existing system works.

Analysts such as Karl Keirstead of UBS have pointed out that the CRM market is already relatively mature, while clients' AI investments in this area are still at a very early stage. Thus, there is a time gap between the market maturity of core products and the maturity of AI additions. This discrepancy makes it difficult for Salesforce to regain its past growth momentum.

The third fundamental challenge concerns the complexity of technology adoption

Although Salesforce promotes Agentforce as a user-friendly low-code solution, the reality for many customers is much more complicated. Successfully implementing AI agents requires a solid data foundation, well-defined processes, technical expertise, and significant investments in training and change management.

Many companies struggle with fundamental challenges such as poor data quality, isolated data silos, inadequate IT infrastructure, and a lack of AI expertise. These issues must be addressed before AI agents can realize their potential. This requires time, resources, and a long-term approach that many companies shy away from.

Added to this is the shortage of skilled workers. Demand for AI experts, data specialists, and Salesforce administrators far exceeds supply. Companies must pay high salaries to attract and retain qualified employees. This further increases the cost of implementing AI solutions and extends the time to value.

A new dimension of digital transformation with 'Managed AI' (Artificial Intelligence) – Platform & B2B Solution | Xpert Consulting - Image: Xpert.Digital

Here you will learn how your company can implement customized AI solutions quickly, securely, and without high entry barriers.

A Managed AI Platform is your all-round, worry-free package for artificial intelligence. Instead of dealing with complex technology, expensive infrastructure, and lengthy development processes, you receive a turnkey solution tailored to your needs from a specialized partner – often within a few days.

The key benefits at a glance:

⚡ Fast implementation: From idea to operational application in days, not months. We deliver practical solutions that create immediate value.

🔒 Maximum data security: Your sensitive data remains with you. We guarantee secure and compliant processing without sharing data with third parties.

💸 No financial risk: You only pay for results. High upfront investments in hardware, software, or personnel are completely eliminated.

🎯 Focus on your core business: Concentrate on what you do best. We handle the entire technical implementation, operation, and maintenance of your AI solution.

📈 Future-proof & Scalable: Your AI grows with you. We ensure ongoing optimization and scalability, and flexibly adapt the models to new requirements.

More about it here:

• The Managed AI Solution - Industrial AI Services: The key to competitiveness in the services, industrial and mechanical engineering sectors

Success stories and disillusionment: What practice teaches us about AI agents

To get a complete picture, it is worth looking at concrete use cases and practical experiences with AI agents, both at Salesforce itself and at other companies.

Salesforce itself has implemented one of the most high-profile AI agent implementations: in its own customer service. CEO Marc Benioff announced in September 2025 that the company had reduced its customer service team from 9,000 to 5,000 employees, a 45 percent reduction. The laid-off employees were replaced by AI agents, who, according to Benioff, have already handled 1.5 million customer conversations, achieving customer satisfaction levels similar to those of human agents.

On the one hand, this drastic measure demonstrates the potential of AI agents to automate repetitive tasks and reduce costs. Salesforce saves significant personnel costs through these layoffs while simultaneously being able to process more inquiries. On the other hand, it raises ethical and practical questions. The quality of customer service for more complex inquiries that require human judgment and empathy remains to be seen. Other companies, such as Klarna, that pursued similar automation strategies have had to admit that service quality suffered.

A second example is AI agents in sales. Several Salesforce customers have implemented agents who automatically qualify prospects, schedule appointments, and send follow-up emails. These agents work around the clock and can handle hundreds of leads in parallel. According to Salesforce, some customers have reported that their sales teams' productivity has increased by 20 to 30 percent as a result of using such agents.

However, there are limitations here too. Agents work best with standardized processes and clearly defined qualification criteria. They quickly reach their limits in complex B2B sales processes that require in-depth product knowledge and strategic negotiation skills. Furthermore, some users report a certain level of dissatisfaction with potential customers who prefer to speak with a human.

Beyond Salesforce, there are numerous other companies using AI agents. ServiceNow, a direct competitor to Salesforce in the IT service management space, has developed its own platform for AI agents. These agents are designed to independently diagnose and resolve IT issues, handle service requests, and orchestrate workflows.

Microsoft also relies on agent-based AI with its Copilot products, but with a slightly different approach. Microsoft agents are more deeply integrated into existing Office 365 products and focus on supporting individual productivity rather than autonomous process automation.

SAP and Oracle are pursuing similar strategies, developing AI agents embedded directly into their ERP and CRM systems. SAP has introduced Joule, an AI assistant that analyzes business processes, provides recommendations, and automates tasks. Oracle is particularly focused on AI-powered cloud infrastructure and is positioning itself as a platform for compute-intensive AI workloads.

What all these examples demonstrate is that AI agents perform best in clearly defined use cases with structured data and standardized processes. The more complex, unpredictable, and human-centric a task is, the more difficult it becomes for autonomous agents to match or exceed human performance.

Suitable for:

• Salesforce AI: Why independent AI platforms are better than Einstein and Agentforce-hybrid approach beats Vendor Lock-In!

Criticism, controversies and unresolved questions: The dark side of the AI ​​revolution

Salesforce's problems and the broader challenges of implementing AI agents have sparked an intense debate about the technology's promises and limitations. Several critical aspects deserve special attention.

The first controversial point concerns job losses. By laying off 4,000 customer service employees, Salesforce sent a clear message: AI agents are not only replacing inefficient processes, but also replacing humans. Benioff had previously asserted that AI would not lead to the disappearance of office jobs. The reality shows something different.

This trend isn't limited to Salesforce. According to data, more than 64,000 tech jobs were expected to be eliminated in the US alone by 2025, many of them related to increased automation through AI. The irony is that at the same time, many of these companies are looking to hire new employees, particularly in AI development and sales. A shift is therefore taking place, with certain roles becoming obsolete while others emerge. But the question remains whether the newly created jobs will outweigh the lost ones in both number and quality.

The second critical aspect is the discrepancy between marketing and reality. Salesforce and other tech companies have promoted AI agents with lofty promises: revolutionizing the world of work, magical productivity gains, autonomous systems replacing human employees. The reality, however, is that many implementations are still in the pilot phase, and the promised productivity gains often fail to materialize or are realized only in limited areas.

A Capgemini study found that while 90 percent of executives surveyed are convinced that agent-based AI provides a competitive advantage, only 14 percent have actually begun implementing it. The majority are still in the planning phase, and almost half lack a concrete implementation strategy. Confidence in fully autonomous AI agents has declined significantly over the past year, from 43 to 27 percent.

A third problematic issue is the dependence on individual tech giants. Salesforce Agentforce is closely integrated with the Salesforce ecosystem. Agents work best when all data and processes are located within the Salesforce world. Integrating external knowledge sources or systems requires considerable effort. This creates a vendor lock-in effect, making it difficult for customers to switch to alternative solutions.

Microsoft, SAP, and Oracle are also facing similar criticism. Each vendor is trying to create its own ecosystem in which its AI agents function best. This complicates the integration of different systems and forces customers to choose a primary provider. Initiatives such as the Model Context Protocol, which aims to enable standardized communication between AI agents from different vendors, are still in their infancy.

A fourth controversial aspect concerns data privacy and security. AI agents require access to extensive corporate data to operate effectively. This creates potential security risks, especially when this data is forwarded to external AI services such as OpenAI or Anthropic. Although Salesforce and other vendors emphasize that they have implemented strict data protection measures, concerns remain, particularly in regulated industries such as healthcare or financial services.

The fifth critical point is the environmental impact. Running large AI models requires enormous amounts of computing power and therefore energy. The data centers that power these models consume millions of kilowatt-hours of electricity and produce significant CO2 emissions. At a time when companies are increasingly under pressure to meet their sustainability goals, the environmental footprint of AI systems is becoming a growing concern.

Looking ahead: Between consolidation and the next wave

Despite all the current challenges, experts predict that AI agents will play an increasingly important role in companies in the coming years. The question is not whether, but how quickly and in what form this technology will prevail.

Gartner predicts that by 2026, approximately 40 percent of all enterprise applications will contain task-specific AI agents, a significant increase from less than 5 percent in 2025. By 2035, agent-based AI could account for approximately 30 percent of global enterprise software revenue, exceeding $450 billion. The market for autonomous AI and autonomous agents will grow from $8.62 billion in 2025 to $263.96 billion by 2035, at a compound annual growth rate of over 40 percent.

These forecasts are based on the assumption that the current challenges will be gradually overcome. Several developments could contribute to this:

First, the technology itself will evolve. The underlying large language models will become more powerful, efficient, and cost-effective. New models such as OpenAI's o1 with improved reasoning or Anthropic's Claude with longer context windows will enable more complex tasks. The cost of AI inference has already fallen dramatically, by a factor of 280 between November 2022 and October 2024. This trend is likely to continue, making AI applications more economically attractive.

Second, companies will learn how to use AI agents more effectively. Early adopters will gather experience, identify best practices, and share them with the broader community. Training programs, certifications, and consulting services will emerge to support companies in their implementation.

Third, standardization could advance. Initiatives such as the Model Context Protocol or ServiceNow's Agent-to-Agent Protocol aim to enable communication between AI agents from different vendors. If such standards become established, they would facilitate integration and reduce vendor lock-in.

Fourth, vendor consolidation is to be expected. The market for AI agents is currently fragmented, with dozens of startups and established players vying for market share. The coming years are likely to see acquisitions and market shakeouts, similar to those seen in other technology segments in the past. Large companies like Salesforce, Microsoft, Google, SAP, and Oracle will acquire smaller vendors to expand their AI capabilities.

For Salesforce specifically, it will be crucial whether the company can successfully integrate the Informatica acquisition and generate real value for Agentforce. The acquisition is the largest in the company's history since the Slack purchase in 2021. It carries risks, as demonstrated by RBC's downgrade, which drastically reduced the price target. But it also offers opportunities if it enables Salesforce to create a more comprehensive data management platform that makes AI agents more effective.

In the medium term, by 2030, Salesforce aims to achieve revenue of over $60 billion, corresponding to an organic growth rate of over 10 percent per year. This would mark a return to double-digit growth after falling below this mark since mid-2024. Whether this goal is realistic depends largely on whether Agentforce and other AI products deliver the hoped-for success.

In the long term, Gartner predicts, the trend could move toward complex multi-agent ecosystems. In such systems, specialized agents work together, coordinating their actions and sharing information. One agent could analyze customer inquiries, another develop proposed solutions, a third coordinate implementation, and a fourth monitor quality. This orchestrated collaboration could automate even more complex business processes.

But there's still a long way to go. The next two to three years will be crucial to see whether the current problems can be overcome and whether the promised productivity gains and revenue increases actually materialize.

Lessons from the Salesforce crisis for the tech industry

The analysis of the Salesforce problem reveals fundamental truths about the state of artificial intelligence and its commercial exploitation. The key finding is that there is a significant discrepancy between the technological feasibility of AI agents and their commercial profitability in the current market environment.

Salesforce is a prime example of an industry that entered the AI ​​era with high expectations but is now confronted with the harsh realities of monetization. The three main challenges identified—monetization difficulties, market saturation, and adoption complexity—are not specific to Salesforce, but affect the entire enterprise software industry.

Experience shows that technological innovation alone is not enough. Companies must also develop a compelling business model, demonstrate clear customer benefits, and lower barriers to adoption. Salesforce has created a technologically impressive product with Agentforce, but translating it into sustainable revenue growth remains a challenge.

For investors, this means they must distinguish between short-term hype and long-term value. The high valuations of many AI companies are based on expectations of future profits that may not materialize or may be significantly delayed. A sober analysis of actual adoption rates, revenue contributions, and profitability is essential.

For companies looking to deploy AI agents, the recommendation is: Start with clearly defined use cases, invest in data quality and change management, and don't expect overnight miracles. The most successful implementations focus on a few well-executed projects rather than launching numerous superficial experiments.

For workers, this development means that certain tasks will be automated by AI, while new roles will emerge. Investing in AI-relevant skills—whether in the development, management, or strategic application of AI—is becoming increasingly important.

The Salesforce case is thus far more than the story of a single company in trouble. It is a lesson about the challenges of technological transformation, the gap between vision and reality, and the need to maintain a clear view of economic realities despite all the enthusiasm for new technologies. The AI ​​revolution will come, but it will proceed gradually, bumpily, and selectively—not as the often-invoked Big Bang, but as a continuous process with ups and downs.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your national language!

Konrad Wolfenstein

I would be happy to serve you and my team as a personal advisor.

You can contact me by filling out the contact form or simply call me on +49 89 89 674 804 (Munich) . My email address is: wolfenstein ∂ xpert.digital

I'm looking forward to our joint project.

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitalization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

Our US expertise in business development, sales and marketing - Image: Xpert.Digital

Industry focus: B2B, digitalization (from AI to XR), mechanical engineering, logistics, renewable energies and industry

More about it here:

• Xpert Business Hub

A topic hub with insights and expertise:

• Knowledge platform on the global and regional economy, innovation and industry-specific trends
• Collection of analyses, impulses and background information from our focus areas
• A place for expertise and information on current developments in business and technology
• Topic hub for companies that want to learn about markets, digitalization and industry innovations

Benefit from Xpert.Digital's extensive, fivefold expertise in a comprehensive service package | R&D, XR, PR & Digital Visibility Optimization - Image: Xpert.Digital

Xpert.Digital has in-depth knowledge of various industries. This allows us to develop tailor-made strategies that are tailored precisely to the requirements and challenges of your specific market segment. By continually analyzing market trends and following industry developments, we can act with foresight and offer innovative solutions. Through the combination of experience and knowledge, we generate added value and give our customers a decisive competitive advantage.

More about it here:

• Use the 5x expertise of Xpert.Digital in one package - starting at just €500/month