How AI agents will unlock value in CRM systems

Insights

• CRM systems have grown in capability and scope
• AI for sales and marketing is popular, but business value is elusive
• CRM systems should be measured for efficiency, effectiveness, customer experience and enterprise scale (4E framework)
• AI agents deliver value orchestrating complex systems, and CRM is an ideal fit

At its best, a customer relationship management (CRM) system delivers timely moments of truth. Cloud and mobile computing have made it possible for CRM to deliver on more of those moments of truth. And first-wave artificial intelligence (AI) tools such as transcription and process automation have further enhanced the ability for CRM to deliver value. We are now seeing generative AI capabilities and AI agents becoming a key part of the CRM value proposition. As companies move AI from sandbox projects to full deployment, AI and AI agents are poised to be critical differentiators. AI agents are systems capable of autonomous action, creating their own workflows and interacting with other systems. This makes them particularly useful in the CRM context to automate administrative tasks, link up siloes and bring clarity to complex systems.

Sales, marketing, and CRM systems are among the top focus areas for companies seeking to deploy AI at scale, Infosys research found. The AI Business Value Radar report, released in March 2025, found that marketing, customer service and sales ranked only behind IT-focused use cases in popularity. (Figure 1.)

Figure 1. Marketing, customer service and sales trail only IT-focused use cases in popularity.

Source: Infosys Knowledge Institute

One of the first CRM platforms to embrace agentic AI as a primary focus is Salesforce, which released its Agentforce product in fall 2024. Agentforce initially focused on sales and customer service agents integrated with its CRM platform but expanded its scope and agentic use cases with the early 2025 release of Agentforce 2.0.

Businesses are poised to make the most of this expansion. For example, a US software developer that has grown by acquisitions reduced its deal cycle time from weeks to days, enabling 10,000 autonomous pricing quotes a month, by moving to a single configure-price-quote platform and introducing automation.

This is a vital step for the company: Shifting to this platform has delivered efficiency, effective sales insights and an improved customer service experience, and lays the foundation for deploying agentic AI in a subsequent phase of development.

CRM’s persistent problems

Despite their new capabilities and scope, CRM systems continue to face a number of persistent problems, including:

• Low adoption: Even state-of-the-art CRM systems can be cumbersome to use, and sales professionals would rather spend time selling than entering data.
• Data trust: CRM systems with incomplete or outdated data will be unreliable and untrustworthy.
• Silos and incomplete integration: A CRM system with only partial coverage of the enterprise will give an incomplete view to sales leaders and frustrating service to clients and partners.
• Complexity and poor fit: A system that is overly complicated and implemented without proper consideration of business strategy won’t keep pace with the sales cycle and rapid business changes.
• Unclear ROI: CRM is expensive. And it has to demonstrate its value to all stakeholders: customers, sales teams, partners, and company executives.

AI value is elusive

Despite its popularity, AI for sales, marketing, and customer service struggles to deliver value.

Infosys’s research found that AI used for personalized customer service, customer segmentation or cross-selling was not likely to generate value. AI used in marketing assets, chatbots, or sales strategy optimization showed some viability. In contrast, achieving business value from AI is very likely for software development or IT services, the survey of 3,789 executives found.

There is evidence that AI for sales and marketing will deliver more concrete value, particularly as companies invest more in AI for sales and marketing functions, and companies transform their operating models and data architecture to be ready for AI.

This is where AI agents and CRM systems converge to unlock business value. Our research finds that AI agents demonstrate strong capability to orchestrate between systems and enable richer end-to-end processes. In fact, orchestration with AI agents is the most popular use case for AI, and one of the most likely to deliver value. (Figure 2.) Orchestration with AI agents starts from the same basis as robotic process automation and goes further by using large language models, planning modules, and data from varied sources to make more intuitive and adaptive decisions. (For a deeper explanation, please see Tech Navigator: Agentic AI Systems.)

Figure 2. AI used for marketing has been less likely to generate business value, and AI agent orchestration is highly likely to generate value.

Source: Infosys Knowledge Institute

As described in case studies below, AI agents for CRM have the potential to accelerate conventional value and solve the most persistent problems in these systems.

The 4E framework

How can AI agents make CRM systems more efficient, effective, and deliver better customer experience at enterprise scale? Infosys and Simplus have developed a framework gauging impact across four measures: Efficiency, effectiveness, experience, and enterprise scale.

Figure 3. CRM should deliver efficiency, effectiveness, improved customer experience and better enterprise integration. The 4E framework measures current performance and the impact of AI agents on CRM.

Our 4E framework can measure CRM performance and help formulate a roadmap to where AI agents can have the greatest impact for all stakeholders. Here are our definitions:

For CRM to be efficient, it gets to the point.

1. Customers get prompt responses
2. Employees get relevant data quickly
3. Partners get a streamlined and consistent relationship

AI agents are proving their worth in orchestrating processes between systems. They also have value in observing and improving processes within systems.

A European packaged food business deployed AI agents to identify similar customer service inquiries coming into its CRM system. The agent was able to identify similarities among the requests and resolve queries more rapidly.

For a CRM to be effective, it must:

1. Anticipate and deliver on customer need
2. Deliver useful and relevant information for sales professionals
3. Enable win-wins and value-add opportunities for partners

A global telecommunications equipment maker used agentic AI to arm its sales reps with a digital assistant. The assistant provides contextual summaries of customer service matters. It prepared them to anticipate customer needs and be more effective in sales processes. This led directly to increased staff productivity, improved customer experience and more precise customer service.

When a CRM delivers top-tier experience, companies see

1. Customer satisfaction and retention
2. Increasing usage and adoption by sales professionals
3. Deeper partner relationships and new collaboration avenues

That same telecom equipment company added a digital assistant to interact with customer queries and an AI agent designed to resolve atypical customer problems. These enhancements led to more accurate customer service responses and delivered overall better customer experience.

But that is not the complete story. CRM systems and AI agents achieve their greatest value operating at enterprise scale.

For a CRM to achieve enterprise scale, it must

1. Deliver on all of the above
2. In as fully integrated ways as possible, and
3. Where silos exist or are necessary, deliver awareness and proper linkage

A European advanced manufacturing business deployed an AI agent to give its sales team a more complete view of prospects. The agent developed sales summaries and contextual intelligence drawn from unstructured data and a range of data sources. The summaries and intelligence have aided the sales team in closing more sales opportunities.

A new approach

Rather than more complexity and cost, AI agents can make CRM systems more useful, trusted, customizable and responsive. Those improvements aligned will bring ROI into focus.

As described in the software company experience above, AI agents can integrate and manage data intake at the wide end of the sales funnel. This leads to better adoption and data trust and ultimately makes sales teams and customers more efficient.

Data trust and integration can benefit from agents that develop their own workflows to test data for accuracy and organize information from systems. Similarly, an AI agent can manage complex processes in the background and deliver only relevant details to sales teams, customers, and business partners.

For example, a telecommunications provider in Singapore used an AI agent to add messaging capability for channel partners and customers. This improved customer experience. Related upgrades also enabled the company to deliver consolidated summaries of customer support cases and interactions.

Specific to its business, the company and business partners had previously struggled with quickly communicating what partners and prospects were eligible for special channel partner programs. With AI enabling improved interaction modes, the company has improved channel management and driven stronger engagement from partners and customers.

In another instance, a US technology services and equipment business used AI agents to integrate its internal sales team with channel partners. This streamlined sales activities and commission payouts when they achieved success.

With proper implementation and governance, AI agents in CRM systems can unlock more value and resolve persistent problems. In the end, companies with an agent-empowered CRM system will use it more, trust it more deeply, carry it to the whole enterprise, configure it for new fits and never doubt the ROI.