Control Beats Clever In The AI Journey
Plus: The brands that earn trust this year will not have the flashiest bot. They will have the cleanest handoff.
To opt-out of receiving DCX AI Today go here and select Decoding Customer Experience in your subscriptions list.
I’m obsessed with Wispr! Get a Free Month of Wispr PRO.
📅 March 9, 2026 | ⏱️ 5 min
Good Morning!
Hope you had a great spring forward!
Today’s edition is all about journey control: clearer disclosure, cleaner transfers, lower ordering friction, and better audit trails. So, grab your coffee, and read on for the one thing CX leaders need to tighten this week: the escape hatch.
The Executive Hook:
Here is the tension. AI keeps getting better at moving customers forward, but the real CX risk shows up at the moment a customer wants out, wants clarity, or wants a human. The FCC’s latest proposal puts that pressure in writing by focusing on disclosure, transfer rights, data handling, and possible expansion beyond voice into chat and email. That makes this a control story, not a chatbot story. The operator edge is simple: design every AI step so a customer can understand it, interrupt it, and recover from it fast.
🧠 THE DEEP DIVE: The Handoff Just Became The Product
This week’s biggest CX story is not a new bot. It is the FCC’s March 5 proposal on customer service, onshoring, and consumer protection. Read it like an operator, and the message is blunt: the “where” and “how” of service is moving from a cost decision to a trust decision. And trust lives or dies at the handoff.
1) Disclosure is now part of the journey, not a fine-print task.
The proposal would require covered providers to tell customers when a call is handled outside the United States. It also asks whether similar expectations should apply to chat, text, and email. Translation: customers will judge you on transparency the same way they judge you on speed. If the customer thinks you hid the ball, the rest of the interaction starts in a hole.
2) “Transfer to U.S.-based support” is a routing feature, not a policy statement.
If customers can request a transfer, your routing and staffing model needs to behave like a product. That means: clear choice, predictable wait, and no lost context. Otherwise you built a compliance switch that creates customer effort. Watch this like a hawk: transfer rate, repeat explanation rate, and post-transfer resolution.
3) Sensitive moments need a tighter lane, with stricter controls.
The proposal would require sensitive customer transactions to be handled only at contact centers located within the United States. Even if your industry is not directly covered, the pattern matters. You need intent-based lanes: low-risk automation, assisted resolution, and high-control handling for anything involving identity, payments, account changes, or fraud. Put your AI in the lane it can safely drive in.
4) Digital channels are next, and most teams are underprepared.
The FCC is explicitly asking whether these rules should extend beyond voice. Many organizations pushed the riskiest automation into messaging and chat because it felt safer than voice. It is not safer. It is just less visible. If your chatbot cannot explain itself, escalate cleanly, and preserve context, it will create the same trust damage, just quieter.
My take:
Most AI programs are still built like a maze. They optimize for containment, then act surprised when customers feel trapped. Customers do not hate automation. They hate being cornered. The brands that scale AI responsibly will do three boring things well: disclose, preserve context, and make exits obvious.
What CX leaders should do this week:
Pick your top five AI-supported journeys. For each one, test three moments end-to-end: (1) disclosure and clarity, (2) fast escape to a human, (3) context transfer that prevents repeat explanations. If you cannot pass all three, do not scale. Fix the handoff first. Then automate more.
Source: FCC Fact Sheet: Improving Customer Service and Protecting Consumers through Onshoring
📊 CX BY THE NUMBERS: AI Is Not Shortening The Journey, It Is Stretching It
Data Source: Net Conversion “The New Rules for Consumer Engagement”
64% of consumers use AI-powered tools to research (up from 45% six months earlier).
98% check additional sources after an AI query. AI is the starting gun, not the finish line.
84% have switched brands for value-related reasons. “Value” now includes shipping, availability, and returns, not just price.
The Insight:
If AI is expanding research, your job is not only to rank. Your job is to show up everywhere the customer sanity-checks. That means consistent answers across search, your site, your help center, and your agents. If any channel contradicts another, customers will keep shopping, or they will contact you to confirm what they just read.
🧰 THE AI TOOLBOX: DoorDash’s Guided Ordering Flow
The Tool: DoorDash is using AI to turn messy pizza menus into a guided, visual ordering flow.
Problem: Complex menus create hesitation, modifier chaos, abandoned carts, and bad orders.
Solution: Picture a customer trying to build a half-and-half pizza for four picky people. Instead of throwing a wall of modifiers at them, the experience reorganizes the menu into a step-by-step flow with visual size, crust, and topping choices. It also keeps customizations in place when the customer changes size or crust, which removes one of the dumb little friction points in digital ordering.
Benefits:
Time: Faster decision-making on complex orders.
Quality: Fewer configuration mistakes and fewer abandoned carts.
Experience: More confidence at the exact moment people usually give up.
Where it sits: Front stage.
Best Fit:
Works best when the product has lots of options, modifiers, or bundles.
Not a great fit when the offer is simple and the real problem is fulfillment, not ordering.
Key Takeaway:
Use it to reduce decision friction before checkout, not just to mop up service contacts after checkout. One caution: this is a vendor announcement, so the promise is clear, but no hard conversion or abandonment metrics were shared.
Source: DoorDash Uses AI to Make Pizza Ordering Easier and Help Merchants Drive Sales
⚡ SPEED ROUND: Quick Hits
ServiceNow introduces Autonomous Workforce and EmployeeWorks - The useful bit is not “AI specialists.” It is that the L1 service desk agent runs inside approvals, role-based access control, and audit trails, which is how customer-facing AI should be built too.
Zendesk March update expands AI agent ticket visibility - Turning AI agent tickets on by default for some customers signals where QA is headed: teams want the bot transcript, not just the bot outcome.
Home Depot and Google Cloud expand agentic AI tools across the project journey - This is journey thinking in practice: discovery, guidance, build lists, delivery, and support connected end to end. The caveat is familiar: strong vision, no metrics shared.
📡 THE SIGNAL: Build AI Journeys Customers Can Exit
The pattern is getting obvious. AI is moving beyond search bars and support macros into real journey moments: order building, service triage, routing, and execution. That is great news until a customer hits uncertainty, policy friction, or a sensitive task. Then the whole game becomes recoverability.
This week’s execution choice is simple: invest one sprint in better handoffs before you invest one more dollar in wider automation. The question to ask your team is this: where does our AI still trap customers instead of guiding them?
See you tomorrow.
👥 Share This Issue
If this issue sharpened your thinking about AI in CX, share it with a colleague in customer service, digital operations, or transformation. Alignment builds advantage.
📬 Feedback & Ideas
What’s the biggest AI friction point inside your CX organization right now? Reply in one sentence — I’ll pull real-world examples into future issues.