74% of CX AI programs fail. Here’s why.
PLUS: Prompts to test where AI should (and shouldn’t) play in your CX
Start every workday smarter. Spot AI opportunities faster. Become the go-to person on your team for what’s next.
🗓️ August 19, 2025 ⏱️ Read Time: ~5 minutes
👋 Welcome
CX leaders are racing to plug AI into support, journeys, and retail. The real story is not about speed, it is about scope. Without strategy and human guardrails most programs fail. The winners are not doing more AI, they are doing AI smarter.
🔎 Signal in the Noise
The numbers tell the truth. Seventy four percent of CX AI programs flop. Leaders skip the basics like clean data, mapped journeys, and clear use cases. The companies that ground AI in human strengths are the ones creating value.
🎯 Executive Lens
AI is not a silver bullet. It is a scalpel. Use it to cut wasted effort, sharpen empathy, and support your people. Your edge will not come from adopting AI everywhere. It will come from knowing exactly where not to.
📌 Stories That Matter
🧠 AI support works best when it boosts humans
Customers do not mind AI in support. What they mind is when it replaces people entirely. A study of 5,000 consumers found 88 percent satisfaction when a human led the interaction with AI assist, compared with only 60 percent when AI went solo. Executives keep talking about AI-first service, but customers want AI behind the scenes, not out front.
Why this matters: Customers remember empathy, not scripts.
Try this: Give your agents AI dashboards that surface context and suggestions so they can focus on tone and care.
Source: CustomerExperienceDive
📉 Why 74 percent of CX AI programs flop
Most rollouts fail because the foundations are not ready. Companies deploy bots without clear use cases, data that is incomplete, or systems that do not integrate. The result is expensive experiments that never scale. Only 13 percent of firms say they are confident about AI readiness today.
Why this matters: You do not have an AI problem. You have a readiness problem.
Try this: Before you spend more money, check if your data is clean, your customer journey is mapped, and your use case is tied to ROI.
Source: CustomerThink
🗺️ Conversational AI only works if you map the journey
Too many leaders treat conversational AI as a channel solution instead of a journey fix. You cannot just drop a bot into WhatsApp or voice and expect customers to be happy. Gartner says 85 percent of CX leaders will pilot conversational AI this year, but success only comes when it is tied to real journey pain points.
Why this matters: Bots without journey maps add friction instead of solving it.
Try this: Choose one journey. Highlight where customers drop off or repeat calls. Test conversational AI only at those points and measure impact.
Source: No Jitter
🎧 Agents, not algorithms, drive future support
AI is in almost every contact center. Ninety eight percent use it, mostly for quality assurance and analytics. The real win is not in replacing humans but in sharpening their focus. With AI handling scoring, transcripts, and trends, agents spend their energy on the conversations that build or break trust.
Why this matters: If AI is not making agents sharper it is wasted.
Try this: Review every AI tool in your center. Kill anything that does not clearly free reps for complex or emotional problems.
Source: contact-centres.com
🛒 Walmart is betting big on AI “super-agents”
Walmart is not dabbling. It is rolling out multi-role AI agents for customers, employees, suppliers, and developers. Customers will meet Sparky, a digital assistant for orders and recommendations. The company’s target is to have half of its sales online in five years. Leaders insist these agents are here to assist people, not replace them, but the balance will be tested.
Why this matters: Retail’s future is multi-role AI agents. Customers will expect the same from you.
Try this: Prototype one small-purpose agent, such as a returns assistant, and measure CSAT and containment before expanding.
Source: Reuters
✍️ Prompt of the Day
Enhance, don’t replace:
In our contact center, identify one low-complexity task and one empathy-driven interaction, and define how AI can support—without replacing—the agent in each.
What this uncovers: Where AI adds value vs where humans must lead.
Apply it quickly: Use in your next pilot or training session.
🛠️ Try This Prompt
List customer touchpoints from discovery to support, flagging where emotions run highest. Then ideate how AI could gently lighten agent workload at those points.
Immediate use case: Pinpoint where AI can ease pressure in emotional moments.
Tactical benefit: Stops you from automating empathy.
📝 CX Note to Self
If AI isn’t lifting your people, it’s dragging your CX down.
👀 See You Tomorrow
Today’s stories show a clear pattern: most AI in CX fails when it is rushed or misplaced, and it succeeds when it frees people to do their best work. The question is not how fast you adopt AI, it is where you put it first.
Where are you testing AI in your CX right now—admin shortcuts, customer journeys, or new digital assistants? Hit reply and let me know, I want to hear what is working (and what is not).
Have an AI-mazing day!
—Mark
💡 P.S. Want more prompts? Grab the FREE 32 Power Prompts That Will Change Your CX Strategy – Forever to start transforming your team, now. 👉 FREE 32 Power Prompts That Will Change Your CX Strategy – Forever
Special offer for DCX Readers:
The Complete AI Bundle from God of Prompt
Get 10% off your first purchase with Discount code: DI6W6FCD