AI Customer Service Fails 4X More Than Other AI tasks
PLUS: Find the Hidden Friction Killing Your Customer Onboarding
Start every workday smarter. Spot AI opportunities faster. Become the go-to person on your team for what’s next.
📅 October 8, 2025 | ⏱️ 3-min read
🎯 The big picture
While everyone’s been celebrating AI’s potential, the receipts just came in—and they’re not pretty. Today’s reality check shows AI customer service failing at 4x the rate of other tasks, but smart companies are doubling down on what actually works: human-like voice AI, proactive intelligence, and automated workflows that don’t leave customers hanging.
📊 Today’s lineup
• Why AI customer service is failing customers (and what Qualtrics found that’ll shock you)
• The voice AI that finally sounds human enough to fool your customers
• CallMiner builds AI agents that actually work together without human babysitting
• This AI predicts what customers want before they even know they want it
• Half of small businesses already use AI for customer service (and 94% are keeping their human teams)
1️⃣ Why AI customer service is failing customers (and what Qualtrics found that’ll shock you)
What’s happening:
• Qualtrics research reveals that 19% of consumers who used AI for customer service saw no benefits from the experience
• That’s a failure rate almost four times higher than other AI tasks, where only 5% see no benefit
• Half of consumers worry AI will make it harder to reach human agents when they actually need help
What’s so concerning: This isn’t about AI being imperfect—it’s about AI being deployed wrong. When customers need help, they’re not looking for clever technology; they want their problem solved. But too many companies are using AI as a cost-cutting measure instead of a customer experience improvement, creating frustrating dead ends instead of helpful solutions.
What’s next: The companies treating AI as a cheap replacement for human service will pay the price in customer satisfaction. The winners will be those using AI to make human agents better, not to make customers disappear. CX leaders need to audit their AI implementations now—before customer frustration becomes customer defection.
Go deeper: Qualtrics
2️⃣ The voice AI that finally sounds human enough to fool your customers
What’s happening:
• SharpenCX partnered with ElevenLabs to launch voice AI agents that sound genuinely human in conversations
• The AI knows when it’s out of its depth and seamlessly transfers to human agents with full context
• Built with 99.999% uptime and integrates with Salesforce, Slack, Teams, and other business tools
What’s so interesting: This addresses the uncanny valley problem that’s been killing voice AI adoption. Previous voice bots sounded robotic enough that customers immediately knew they were talking to a machine and got frustrated. These new agents are so natural that customers focus on getting help rather than fighting the technology.
What’s next: Voice AI is finally crossing the threshold where it enhances rather than hinders customer experience. Contact centers that stick with obviously artificial voice systems will create friction that customers won’t tolerate when competitors offer seamless, human-like interactions.
Go deeper: SharpenCX
3️⃣ CallMiner builds AI agents that actually work together without human babysitting
What’s happening:
• CallMiner advanced its agentic AI framework to automatically create virtual agents based on conversation patterns it discovers
• New LiveTranslate feature enables real-time translation so agents can handle customers in any language
• The AI can automatically trigger customer outreach campaigns based on conversation insights—no human setup required
What’s so great about this: This solves the “AI maintenance nightmare” problem that’s been slowing enterprise adoption. Instead of requiring data scientists to manually build and maintain AI workflows, CallMiner’s system observes what’s happening in customer conversations and builds the automation itself. It’s like having an AI that trains other AIs.
What’s next: The era of manually programming customer experience AI is ending. Organizations that still need armies of technical specialists to maintain their AI systems will be outpaced by those using self-improving agentic frameworks. CX teams can finally focus on strategy instead of AI babysitting.
Go deeper: CallMiner
4️⃣ This AI predicts what customers want before they even know they want it
What’s happening:
• InMarket launched Predictive Moments, an AI that anticipates customer purchase occasions and engagement opportunities
• The platform identifies Purchase Moments, Event-based Moments, Meal Moments, and Recreation Moments without requiring location data
• It analyzes past engagement and purchase data to reach customers before they even start searching
What’s so interesting: This flips customer engagement from reactive to proactive. Instead of waiting for customers to show purchase intent, the AI identifies when they’re likely to need something and reaches out first. It’s like having a personal shopping assistant who knows your patterns better than you do.
What’s next: Businesses that wait for customers to express interest will lose to those that anticipate needs. When AI can predict purchase occasions with precision, the competitive advantage goes to brands that solve problems customers didn’t even know they had yet.
Go deeper: InMarket
5️⃣ Half of small businesses already use AI for customer service (and 94% are keeping their human teams)
What’s happening:
• Talkdesk survey reveals 51% of small businesses have already integrated AI into customer service operations
• 94% expect to grow or maintain their customer service teams over the next two years
• Small businesses using AI report faster resolution times (65%), 24/7 support (60%), and higher customer satisfaction (40%)
What’s so great about this: Small businesses are proving that AI and humans work better together, not in competition. They’re using AI for chatbots, knowledge bases, and analytics while keeping human agents for complex issues. The result? Better customer service with happier employees who can focus on meaningful work instead of repetitive tasks.
What’s next: The myth that AI will replace customer service teams is officially dead. Small businesses—who have to make every investment count—are showing that the winning formula is AI-augmented humans, not AI-replaced humans. Larger enterprises still debating whether to adopt AI are already falling behind small competitors who figured this out.
Go deeper: Talkdesk
⚡ Quick hits
• Gryphon AI launched new compliance platform turning trust into a growth engine → 173% increase in customer reach and 51% boost in engagement effectiveness
• Unily unveiled Futures AI-native employee experience platform → creates fully-branded intranets in minutes with “Indy” AI agent and conversational interface
• NetSuite expanded partner program to accelerate AI innovation → new SuiteApp.AI marketplace and AI Elite badges for partners building secure solutions
💡 CX Prompt Tip of the Day
Find the Hidden Friction Killing Your Customer Onboarding
I want to discover exactly where new customers get stuck during onboarding and identify opportunities to eliminate friction or create wow moments.
Context: Our customer onboarding process involves [describe your key steps from signup to first value]. We see drop-offs at [mention specific points] and our current completion rate is [X]%.
Your task:
1. Map every micro-moment from initial signup to successfully using our product for the first time
2. Identify the top 3 friction points where customers hesitate, get confused, or abandon the process
3. For each friction point, determine the root cause and design a solution that eliminates the problem entirely
4. Spot 2-3 opportunities where we could surprise and delight customers instead of just getting them through
5. Create a 30-day test plan to validate which improvements actually increase completion rates
Format: Onboarding Step → Friction Points → Root Causes → Solutions → Delight Opportunities → Test Plan → Success Metrics
Focus on quick wins you can implement immediately while identifying bigger improvements that need development resources.
Quick win: Go through your own onboarding this week as if you’re a brand new customer. Record yourself doing it. You’ll spot friction points in the first 5 minutes that your team has been blind to for months.
🤔 CX reflection
Question of the day: If Qualtrics found that AI customer service fails 4x more than other AI tasks, what’s the one thing your current AI implementation might be doing that frustrates customers more than it helps them?
👋 See you tomorrow!
Mark
💡 P.S. Grab 32 Power Prompts That Will Change Your CX Strategy – Forever → Get prompts