Three Companies Show How AI Actually Works in Customer Support
The Help Desk Institute announced finalists this week for its 2026 "Best Use of AI" award, and the three companies competing offer a clear picture of how artificial intelligence has moved from chatbot experiments into core business operations. Allstate, WBM Technologies, and SupportLogic each represent a different path: legacy enterprise transformation, managed services innovation, and software-native architecture.
The category matters because customer experience has become a revenue driver, not a cost center. A 2025 PwC survey found that 52 percent of consumers stopped buying from a brand after a single bad experience. Seventy percent abandon a brand after two negative interactions. The real danger is silent churn-only one in 26 unhappy customers complain. The rest leave quietly.
The financial upside of getting this right is substantial. Customers with perfect experience scores spend 140 percent more and stay loyal six times longer. Forrester research shows brands aligning customer experience with brand strategy can unlock 3.5 times more revenue growth.
Enterprises are responding with capital. Gartner projects worldwide AI spending will reach $2.52 trillion in 2026, a 44 percent increase year-over-year.
Allstate: Automating at Scale
Allstate, a 95-year-old insurance company, deployed a conversational AI system handling over 400,000 customer conversations monthly. The system achieves a 38 to 40 percent containment rate, routing simple inquiries away from agents so they can focus on complex claims.
The company went further. Almost all communications Allstate representatives send to claimants are now written by generative AI. The company's CIO said these AI-generated emails achieve higher customer satisfaction scores than emails written by human managers.
The financial results are concrete. In Q3 2025, Allstate reported net income of $3.7 billion, up from $1.2 billion a year earlier. Allstate Technology Support also made the finalist list for Best Service and Support Organization, showing the same approach works for internal IT operations.
WBM Technologies: Predicting Problems Before They Happen
WBM Technologies, a 75-year-old managed IT services firm based in Western Canada, won the HDI award for Best Service and Support Organization in 2024. This year it competes in three categories, including Best Use of AI.
The company built a patent-pending Enterprise Experience Platform that integrates AI into service delivery rather than treating it as a separate tool. The platform combines SOC 2-compliant operations with predictive technologies that resolve issues before end-users notice problems.
WBM employs over 500 IT professionals and recently appeared on CRN's 2026 Elite 150 Managed Service Provider list.
SupportLogic: Replacing the Traditional CRM
SupportLogic takes a different approach. Founded by former VMware executive Krishna Raja, the company argues that traditional ticketing systems and CRMs-passive databases requiring manual data entry-are obsolete.
SupportLogic built a "system of intelligence" instead. Ambient AI agents run in the background 24/7, analyzing support interactions without replacing existing tools like Salesforce, Zendesk, or ServiceNow. Deployment takes under 60 days.
The platform uses four engines to work. A Data Extraction Engine consolidates unstructured data across systems. A Signal Extraction Engine reads 100 percent of tickets, voice transcripts, and messages using natural language processing to extract over 140 customer sentiment signals. A Context Engine maintains memory across interactions and channels. An Orchestration Engine triggers action based on AI findings.
The result is predictive support. Because the AI constantly monitors sentiment and behavior, it flags escalation and churn risks before customers complain.
Enterprise clients report measurable results. NICE achieved a 35 percent reduction in escalation rates. Certinia saw a 30 percent reduction in escalations and a 28 percent decrease in resolution time. Nutanix reduced escalations by 40 percent while maintaining a 90 NPS score. Across its customer base, SupportLogic reports an average 80 percent reduction in escalation requests and 53 percent reduction in mean time to resolution.
SupportLogic runs its own support operation on its platform-the company uses its own ambient AI agents to manage customer interactions and predict escalations. This serves as a working example of the architecture the company sells.
What This Means for Support Teams
The three finalists show that AI for Customer Support has moved beyond automation. The focus is now on prediction and understanding customer needs deeply enough that problems are prevented.
Generative AI and LLM technology powers much of this work-from Allstate's email generation to SupportLogic's natural language processing of support interactions.
The HDI Awards will announce the winner in May in Las Vegas. But the broader trend is clear: the era of reactive support is ending. The future belongs to teams and systems that anticipate customer needs rather than respond to them.
Your membership also unlocks:
