In an article published by Salesforce, the company examined how organizations can effectively deploy artificial intelligence agents in customer service to deliver tangible business results. According to the article, AI agents can perform a broad variety of tasks—from inspecting a photo of a moldy shower and helping dispatch technicians, to assisting millions of football fans with ticketing, memberships, and stadium events, to helping navigate complex financial realities. For companies willing to redefine the role of AI agents in their technology stack and underlying operational processes, these agents can serve as a significant customer service advantage.
While evaluating submissions for Salesforce's Customer Success Awards—a program honoring customers who used the Agentforce platform to drive real business results—the team identified a consistent pattern among organizations that achieved the most successful transformations. Although different organizations faced different challenges, the successful initiatives were grounded in three core principles: starting with the right problem, building the right foundation, and bringing the right people together.
Starting with the Right Operational Problem
According to the article, true transformation begins when teams identify a defined operational friction point and focus on it rather than chasing generalized AI capabilities. A prime example is London-based Tottenham Hotspur Football Club. The club fields a massive volume of inquiries from a global fanbase, particularly around match-day ticketing, stadium access, and account management. Historically, resolving fan questions required human service representatives to log into three separate systems, a process taking four to five minutes per interaction.
Club leadership focused their initiative on resolving this friction point. First, the organization used Salesforce Data 360 to consolidate disparate data sources into a single real-time context layer, unifying more than 4.6 million fans and 30 separate systems into a single "golden record." Next, they built the AI-powered Ask Spurs service agent, equipped it to handle queries in eight different languages, and grounded it directly in unified profile data. Finally, the unified data context was applied across both self-service chat channels and agent-assisted workflows.
According to the club's Chief Technology Officer, Rob Pickering, the Ask Spurs agent now handles the majority of inquiries instantaneously. Furthermore, when an interaction must be handed off to a human representative, the unified profile enables agents to resolve issues in an average of 10 seconds or less. These improvements saved 80,000 minutes in the first month and are on track to handle more than 200,000 calls per year. The article emphasizes that effective solutions prioritize clear return on investment and time savings, noting that targeting a single, repetitive interaction can save thousands of hours of human labor.
Building a Solid Operational Foundation
The second principle presented in the article is anchoring an AI project on a stable operational foundation, which includes clean data, scalable infrastructure, and adapted processes. Even sophisticated agents will encounter difficulties if built on fragmented data or outdated practices.
An example of this is The Grout Guy, a family-owned business and Australia's largest tile and grout restoration network. In the company's early days, management tracked schedules and client books using manual paper planners and basic spreadsheets. As demand expanded nationwide, managing dispatch, lead intake, and job quotes through manual methods created operational bottlenecks.
Leadership at The Grout Guy executed a comprehensive digital overhaul: they centralized workflows and replaced legacy logs with an integrated Salesforce suite including Data 360, Agentforce Marketing, Agentforce Field Service, Agentforce Service, Agentforce Builder, and Slack. In addition, lead intake and speed-to-quote workflows were automated. As part of the Agentforce implementation, the company built a multiagent system, including an agent that automatically processes incoming work orders from emails and PDF files.
Furthermore, the customer-facing service agent, named Groutie, prompts website visitors to upload photos of their bathrooms via chat and uses optical image recognition to count tiles, identify mold or discoloration, and assist in generating a quote. According to CTO Anthony Messina, quotes are now generated within 20 minutes through Groutie, compared to the three to five days the process previously took manually. The company increased its field capacity from four technicians per administrator to more than 25 technicians without significant growth in the dispatch team. Additionally, work order processing now takes less than two minutes, 24 hours a day, seven days a week.
Bringing the Right People Together and Integrating Human Empathy
The third principle stresses that sustainable AI adoption requires uniting cross-organizational teams, maintaining feedback loops, and ensuring employees view automated tools as collaborative coworkers rather than replacements for their jobs.
Sammons Financial Group, headquartered in West Des Moines, Iowa, applied this principle when exploring how agentic AI could support customers and distribution partners while expanding service capacity. To establish rapport with customers, technology teams worked with Agentforce to develop an agent with a friendly, empathetic tone and clear conversational guardrails. As agent coverage expanded to nights, weekends, and Friday afternoons, positive customer responses were recorded. Andrew Walling, the company's Assistant Vice President (AVP) of Capability Planning & Delivery, noted that customers engaged in meaningful, safe conversations with the agent, reflecting genuine trust.
Concurrently, company leadership worked to ease employee concerns by creating an internal team SharePoint site that framed the agent's development as an evolving persona—progressing from a learning "apprentice" to a full-fledged service colleague. Combining empathetic design with human oversight allowed the system to handle more than 16,000 policy calls in its first six months of launch, providing service responses for most inquiries while freeing human advisors to handle complex and emotionally sensitive requests. This approach enabled the company to absorb seasonal peak call spikes without hiring temporary staff and without increasing hold times. The article concludes that focusing on the right problem, laying a solid foundation, and engaging people throughout the process allows service organizations to turn visions into daily operational reality.