AI Just Got Better at Actually Helping Your Customers
PLUS: Design AI Escalation Rules That Actually Work
Start every workday smarter. Spot AI opportunities faster. Become the go-to person on your team for what’s next.
📅 September 29, 2025 | ⏱️ 3-min read
🎯 The big picture
While everyone debates AI’s future, smart CX teams are quietly solving real customer problems right now. This week’s developments show AI moving from experimental to essential for customer experience professionals.
📊 Today’s lineup
• Adobe releases AI agents that actually handle customer experience tasks automatically
• Zendesk’s AI Copilot saves 45 seconds per ticket with smarter routing and triage
• Asana solves the “our AI agents keep failing” problem with collaborative teammates
• Microsoft’s Real-Time Intelligence detects customer experience anomalies as they happen
• Medallia and Adobe team up to connect feedback analysis with automated action
1️⃣ Adobe releases AI agents that actually handle customer experience tasks automatically
What’s happening:
• Adobe launched six AI agents in September 2024 that automatically handle audience segmentation, journey building, and experimentation without human intervention.
• The agents integrate with third-party systems to execute complex customer experience workflows across multiple platforms and touchpoints.
• Agent Orchestrator coordinates these AI agents to work together, creating seamless automation of previously manual CX processes.
What’s so great about this: This solves the bottleneck problem that plagues CX teams: having great ideas but lacking the resources to execute them quickly. These AI agents don’t just analyze—they actually do the work. They can segment audiences, build journeys, and run experiments while CX professionals focus on strategy and creativity.
What’s next: CX teams that still manually build every customer journey and segment will struggle to keep pace with those using autonomous AI agents. The competitive advantage goes to teams that can execute ideas at machine speed while maintaining human strategic oversight.
Go deeper: CX Today
2️⃣ Zendesk’s AI Copilot saves 45 seconds per ticket with smarter routing
What’s happening:
• Zendesk’s AI Copilot now automatically routes tickets to the right agents on first contact, eliminating the back-and-forth that frustrates customers and wastes agent time.
• The system analyzes customer intent, sentiment, and language in real-time to predict exactly which agent can solve each problem most efficiently.
• Intelligent triage processes tickets 45 seconds faster than manual routing, allowing agents to focus on actual problem-solving instead of administrative tasks.
What’s so interesting: This addresses the hidden inefficiency that drives both customer frustration and agent burnout: tickets bouncing between the wrong people. When a billing question goes to a technical agent, everyone loses time. Zendesk’s AI eliminates this friction by understanding what customers actually need before humans even read the ticket.
What’s next: Customer service teams that rely on manual ticket routing will create unnecessary friction for both customers and agents. Smart routing becomes table stakes for competitive customer support operations, not a nice-to-have feature.
Go deeper: Zendesk Support
3️⃣ Asana solves the “our AI agents keep failing” problem with collaborative teammates
![[AI Studio] Hero Image](https://substackcdn.com/image/fetch/$s_!9uod!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8b0c8af-534b-4a38-abe3-5d2a8f08d92e_1400x1400.webp "[AI Studio] Hero Image")
What’s happening:
• Asana launched AI Teammates on September 29, addressing the 70% failure rate of autonomous AI agents by building collaborative agents that work with teams, not alone.
• These agents have context across your entire organization, built-in checkpoints for transparency, and enterprise controls that prevent rogue automation.
• AI Teammates can serve multiple teams simultaneously as Campaign Strategists, IT Ticketing Specialists, Bug Investigators, or Launch Navigators—learning and adapting from human feedback.
What’s so great about this: This tackles the biggest problem with workplace AI: agents that work in isolation and fail at complex tasks. Asana’s approach recognizes that real work happens between teams, not individuals. Their AI Teammates understand your company’s “blueprint”—who does what, when, how, and why—making them actually useful for collaborative work.
What’s next: CX teams that embrace collaborative AI will outpace those chasing autonomous solutions. When AI can work effectively across marketing, IT, and operations teams simultaneously, customer experience becomes more coordinated and responsive. The future belongs to human-AI collaboration, not AI replacement.
Go deeper: Antaranews
4️⃣ Microsoft’s Real-Time Intelligence detects customer experience anomalies instantly
What’s happening:
• Microsoft Fabric’s Real-Time Intelligence now includes AI-powered anomaly detection that identifies unusual patterns in customer behavior and experience metrics as they happen.
• The system monitors streaming data from support tickets, product usage, and customer interactions to detect problems before they impact large numbers of customers.
• Automated alerts notify CX teams immediately when satisfaction scores drop, support volume spikes, or customer behavior changes unexpectedly.
What’s so interesting: This solves the “boiling frog” problem in customer experience—issues that develop gradually and go unnoticed until they become major crises. Real-time anomaly detection acts like an early warning system, helping CX teams spot problems when they’re still small and fixable.
What’s next: CX teams will split between those operating with real-time intelligence and those still discovering problems after the fact. When you can detect and address experience issues within minutes rather than weeks, customer satisfaction becomes dramatically more manageable.
Go deeper: Microsoft Fabric Blog
5️⃣ Medallia and Adobe team up to connect feedback analysis with automated action
What’s happening:
• Medallia’s expanded partnership with Adobe connects voice-of-customer analytics directly to Adobe’s AI agents for automated experience optimization.
• When Medallia detects negative feedback patterns, Adobe’s AI agents automatically adjust customer journeys, messaging, and touchpoints to address the issues.
• The integration closes the loop between listening and acting, ensuring customer feedback drives immediate improvements rather than sitting in reports.
What’s so great about this: This eliminates the biggest gap in customer experience management: the time between hearing customer feedback and actually doing something about it. Most companies collect feedback but struggle to turn insights into action quickly enough to matter. This integration makes feedback actionable in real-time.
What’s next: The advantage goes to companies that can act on customer feedback immediately, not just analyze it thoroughly. When your competition can adjust customer experiences based on feedback within hours instead of quarters, listening without acting becomes a liability.
Go deeper: CMSWire
⚡ Quick hits
• Zendesk enhances intelligent triage → automatically detects customer intent, sentiment, and language to improve first-contact resolution
• Microsoft Fabric expands real-time intelligence → nearly half of Fabric’s 24,000+ customers now use real-time data processing for operational insights
• Adobe Experience Platform launches Agent Orchestrator → coordinates multiple AI agents to handle complex customer experience workflows automatically
💡 CX Prompt Tip of the Day
Design AI Escalation Rules That Actually Work
I want to create smart escalation rules that route customers to humans at exactly the right moment.
Context: Our AI chatbot currently escalates too late (frustrated customers) or too early (unnecessary handoffs). Our main customer issues are [insert top 3 support categories].
Your task:
1. For each issue category, identify 3 specific conversation signals that indicate AI isn’t helping (e.g., customer uses “frustrated” language, asks the same question 3 different ways, requests to “speak to a manager”).
2. Create escalation triggers that activate BEFORE customers get angry, not after.
3. Write a smooth handoff script that explains what the AI has already tried and what the customer needs.
4. Design a fallback plan for when no human agents are available.
Format: Issue Category → Early Warning Signals → Escalation Trigger → Handoff Script → Backup Plan. Focus on catching problems before they become complaints.
Quick win: Review your AI chatbot’s escalation logs from last week and identify one conversation that escalated too late. Use this prompt to design a better trigger for that scenario.
🤔 CX reflection
Question of the day: If your AI could predict which customers are about to have a bad experience before it happens, what would you want it to do automatically to prevent that outcome?
See you tomorrow!
Mark
💡 P.S. Grab the FREE 32 Power Prompts That Will Change Your CX Strategy – Forever → Get prompts