What CX Leaders Need to Take Back Before It’s Too Late
DCX Links | January 4, 2026
Welcome to the DCX weekly roundup of customer experience insights!
Customer experience isn’t being redesigned in meeting rooms. It’s being shaped quietly, far from the front lines, by systems most customers—and many leaders—never see.
This week’s stories all point to the same tension. Companies are moving faster to model employees, automate decisions, and personalize interactions than they are to decide who’s responsible when those systems get it wrong. Support issues get resolved automatically. Work behavior gets scored. Customer journeys bypass the website entirely. And when something feels off, there’s often no clear owner to explain or correct it.
The risk isn’t technology itself. It’s distance.
Distance between leaders and the people doing the work. Distance between decisions and their impact. Distance between efficiency metrics and lived experience. That distance shows up in CX long before it shows up in reports.
The next phase of customer experience won’t be defined by smarter systems. It will be defined by leaders who are willing to set limits, name owners, and protect trust before it has to be repaired.
This week’s must-read links:
Digital Twins Are Coming for Employees. Leaders Need to Decide the Line
Five CX Resolutions Leaders Can’t Afford to Skip in 2026
Why AI Breaks When Ownership Fades
Real-Time Personalization Is Now Table Stakes for Engagement
DCX Stat of the Week: The “Zero-Click” Traffic Collapse
DCX Case Study of the Week: How ClassPass Used AI to Scale Support Without Sacrificing CX
Digital Twins Are Coming for Employees. Leaders Need to Decide the Line.
Digital twins were built to model machines. Now they’re modeling people. HR platforms are creating always-on replicas of employees to predict behavior, productivity, and “life quality.” The technology is advancing faster than the ethics, and CX leaders will feel the impact first.
Why it matters:
Tools from syd™ and Delve turn health, behavioral, and demographic data into persistent employee profiles that managers can interrogate anytime.
These systems promise burnout prevention and personalization, yet they quietly normalize surveillance and reduce lived experience to scores and traits.
CX outcomes follow EX design. When employees feel monitored rather than trusted, service quality erodes long before metrics catch up.
State of play:
Digital twins work best for physical systems. BMW uses them to simulate factories and avoid collisions.
When applied to people, gains are thin. Studies show productivity lifts of 1–3 percent, often paired with faster work, tighter controls, and less autonomy.
Between the lines:
This isn’t about replacing workers. It’s about managing uncertainty when growth stalls.
Digital twins create distance. Leaders see cleaner models while missing fatigue, context, and moral tradeoffs.
The CX To-Do:
Before approving workforce AI, set three guardrails: explicit consent, limits on after-hours use, and human override on any decision that affects workload, pay, or performance.
🔗 Go Deeper: Max Hancock, The Drift
Five CX Resolutions Leaders Can’t Afford to Skip in 2026
AI isn’t changing customer support at the edges. It’s reshaping the role entirely. In this piece, Adrian McDermott, CTO at Zendesk, lays out five resolutions CX leaders should commit to if they want service to remain relevant, human, and influential in 2026.
Why it matters:
AI is absorbing repetitive, rules-based work faster than most org charts are updating.
When support stays focused on ticket volume, it loses strategic ground to product, ops, and automation teams.
CX leaders now face a choice. Elevate the function or let it be optimized into the background.
What’s happening:
McDermott argues support must move from issue resolution to experience improvement across the full journey.
Humans should be redeployed toward judgment-heavy work like complex problem solving, relationship repair, and prevention.
New roles are emerging inside CX. People who configure, monitor, and improve AI, not just run queues.
Support teams need to operate like product teams. Iterative releases, clear ownership, testing, and learning.
Between the lines:
This isn’t a tooling shift. It’s an accountability shift.
AI exposes weak ownership faster than it creates efficiency.
The CX To-Do:
Review your 2026 plans. Where are you redesigning roles, assigning AI ownership, and measuring experience impact instead of volume?
🔗 Go Deeper: Zendesk
Why AI Breaks When Ownership Fades
Justin R. Greenbaum just launched his Substack, and it’s one CX leaders should take seriously. Drawing on over 20 years as a customer experience strategist at Comcast, he opens Justin’s Substack with a clear warning. AI is no longer a back-office tool. It’s becoming decision infrastructure. And the biggest risk isn’t accuracy or speed. It’s what happens to responsibility when decisions scale faster than ownership.
Why it matters:
Most AI harm doesn’t come from bad intent or broken models. It comes from diffused authority, misaligned incentives, and proxy metrics standing in for judgment.
For CX leaders, this shows up as decisions customers feel immediately but no one can fully explain or defend.
Speed without accountability erodes trust long before dashboards signal a problem.
What’s happening:
Responsibility has to be designed into defaults, permissions, escalation paths, and stop mechanisms. Policies and principles don’t survive scale on their own.
AI should surface uncertainty and risk, not quietly replace human judgment. Prediction can inform decisions. It cannot legitimize them.
Deployment is a responsibility event. The moment AI enters a workflow, power, risk, and accountability shift.
The bottom line:
Greenbaum’s message is blunt. Prediction is not judgment. Automation is not responsibility. Efficiency is not legitimacy. CX leaders need named owners, clear escalation, and the ability to slow or stop systems when trust is at stake. Better models won’t solve this. Better system design will.
🔗 Go Deeper: Justin R. Greenbaum, Substack
Real-Time Personalization Is Now Table Stakes for Engagement
Most website experiences still assume customers are static. They’re not. In this piece, Epsilon Marketing argues that real-time personalization is no longer a growth tactic. It’s the difference between engagement and abandonment.
Why it matters:
65% of consumers say companies don’t use their data well enough to personalize experiences.
Nearly 40% of shoppers have abandoned over $100 worth of items due to poor website experiences.
Customers share data when they trust brands to use it responsibly and visibly in their favor.
What’s happening:
Static websites fail because they don’t respond to intent signals like search behavior, scrolling, exits, or repeat visits.
Real-time personalization turns those signals into dynamic messaging, behavioral segmentation, and ongoing experimentation.
First-party data becomes more valuable when it’s activated immediately, not stored for later analysis.
Between the lines:
Personalization isn’t about pushing more offers. It’s about reducing friction at moments of intent.
Trust grows when customers feel recognized without feeling tracked.
The CX To-Do:
Audit your site for moments where intent appears but nothing responds. If behavior changes and the experience doesn’t, personalization is broken.
🔗 Go Deeper: Epsilon Marketing
DCX Stat of the Week
The “Zero-Click” Traffic Collapse
DCX Stat: According to Michael Mallett, VP, Digital Center of Excellence at Medallia, by 2026, brands are projected to lose upwards of 40% of their organic website traffic as generative AI platforms (ChatGPT, Gemini, Perplexity) resolve customer queries and complete transactions directly within their own interfaces, bypassing company websites entirely.
Takeaway: The traditional “website-as-destination” model is dying. CX leaders must pivot from optimizing web funnels to ensuring their brand’s data and “buy-buttons” are natively integrated into the AI ecosystems where customers now start and end their journeys.
🔗 Source URL: 15 Predictions That Will Redefine Customer Experience in 2026
🔗 MORE STATS: Daily Stats on Substack Notes
DCX Case Study of the Week
How ClassPass Used AI to Scale Support Without Sacrificing CX
CX Challenge:
As ClassPass grew globally, customer support demand spiked fast. The team needed to scale without ballooning headcount — but traditional chatbots risked frustrating members who expected fast, human‑quality support.
Action Taken:
ClassPass partnered with Decagon to deploy AI agents trained on real support conversations. These agents handled common member issues end‑to‑end, escalated complex cases seamlessly, and continuously improved using feedback from resolved tickets.
Result:
65% of support tickets fully resolved by AI without human intervention
Instant response times, reducing wait frustration for members
Support agents freed up to focus on high‑value, complex member needs
Lesson for CX Pros:
AI works best when it’s treated like a teammate, not a gatekeeper. Train AI on real conversations, let it fully resolve simple issues, and design clean handoffs — that’s how you scale CX without eroding trust.
Quote:
“Though we already had a robust Voice of the Customer program and an understanding of customer inquiries we thought we could deflect, we saw 10x higher deflection at launch than we anticipated.” — Sarah Vanden Broek, former ClassPass Director, CX Operations Strategy and AI Director
🔗 Further Reading: Decagon
Have a case study to share? Reply and let me know!
🔚 Final Thought
CX doesn’t fail all at once. It thins out—through small decisions made faster than they’re owned.
As AI reshapes work, support, and discovery, the real question isn’t what these systems can do. It’s who is accountable when they act, and where you’re willing to draw the line.
Before the next rollout, ask yourself:
What decisions have we automated without clear ownership?
Where has efficiency replaced judgment?
The next move is yours.
Thank you!
If this edition sparked ideas, share it with a colleague or team member. Let’s grow the DCX community together!
✉️ Join 1,460+ CX leaders who stay ahead of the next customer curve.
Human-centered insights. Plug-and-play frameworks. Smart tools that actually work.
All designed for CX pros who want to build with purpose—and deliver with impact.
👉 Subscribe today and get the tools to elevate your strategy (and your sanity).
Thanks for the thoughtful synthesis.
What I keep seeing is that most AI failures aren’t technical. They’re ownership failures that surface as CX issues because that’s where consequences land first.
I’m trying to make those patterns visible before trust is spent.