Your Next Customer May be an AI
Plus: CX teams now need to design for software that can act, not just answer
Your daily signal on AI and CX — minus the hype.
📌 DCX Stat of the day: 78% of Americans now use AI-enabled tools in daily life, but only 18% say they would trust AI to make financial recommendations on its own. TD Stories
In this issue:
→ Visa says AI is entering commerce directly
→ Gartner shifts the focus to workflow results
→ Customers still want humans accountable
→ Service tech spend is rising fast
→ Sierra launches a tool for tuning live AI
CONTEXT
Visa just gave the shift a name
Visa says commerce is moving toward a model in which AI not only assists with shopping but also participates in decision-making and execution. It's April 2, research says more than half of surveyed business leaders are open to AI-to-AI negotiation, while consumers still treat trust and override capability as non-negotiable.
WHY IT MATTERS
That changes the CX job. A year ago, the practical AI question was whether a bot could answer well enough to reduce effort. Now the harder question is whether a system can take action without breaking trust. Once software starts comparing options, negotiating terms, or moving a transaction forward, CX inherits a new set of responsibilities around permissions, reversibility, proof, and recovery.
EXEC SUMMARY
🎯 Exec Briefing: Why this should be on your agenda
The shift here is simple. AI is moving from interface layer to action layer. That means the most important design work no longer sits in prompts and summaries. It sits in rules, approvals, event logs, and the ability to step in fast when a customer says the system made the wrong call. Visa is already describing the new rules of commerce around trusted autonomous agent payments.
The brands that handle this well will make AI feel useful without making the customer feel replaced. That sounds obvious. It is also where plenty of teams will get into trouble.
📬 Copy-Paste Take: Send this to your COO
Hey, check this out. AI is moving past answering questions and starting to act inside the customer journey. That shifts the risk. The problem is less about whether the model sounds smart and more about whether the business has set clear limits, approvals, and recovery paths before the system makes a real decision on a customer’s behalf.
🔎 Deep dive
Assistive AI is already becoming workflow AI
Gartner said on April 2 that by 2028, over half of enterprises will stop paying for assistive intelligence, such as copilots and smart advisors, and instead favor platforms that commit to delivering workflow results. It also says the first disruption will hit approval-heavy, timing-sensitive workflows where AI collapses decision latency and reallocates authority to policy-bound agents.
That matters for CX because most teams are still treating AI like a thin layer on top of existing service design. The market is moving toward systems that complete work inside real operations. If your journeys are messy, your policies are fuzzy, or your data is split across silos, AI will expose that faster than it fixes it.
OPERATOR PLAYBOOK
Audit one delegated journey now
Start with a flow where AI could influence a real outcome. Returns. Refunds. Billing disputes. Plan changes. Identity checks.
Audit every delegated-action journey for four things:
What the AI is allowed to do without approval
Where the customer can review, stop, or reverse the action
What evidence exists if the outcome is disputed
How quickly a human can take over and fix the issue
Then test whether a frontline leader can explain that flow in plain English to legal, operations, and a customer.
Ask your team: If an AI-driven action goes wrong, can we prove what happened in a way the customer will accept?
Signal: The next CX failures will come from unclear authority more than weak model output.
📈 Market Reality Check
Customers are comfortable with AI support, not AI control
TD Bank’s new U.S. consumer survey provides a useful benchmark.
Two-thirds of consumers are most comfortable when AI handles behind-the-scenes functions such as fraud detection, tracking spending, and calculating credit scores. Comfort falls when AI becomes the autonomous decision-maker for complex or high-stakes choices. Only 18% say they would trust AI to make financial recommendations on its own.
That is a better rule of thumb for CX than broad adoption headlines. People are not rejecting AI. They are setting terms. They want speed, simplicity, and support. They still want humans accountable for the outcome. In agentic CX, that difference is the whole game.
High automation + visible human accountability = adoption path
🧰 Tool Worth Knowing
Sierra Explorer
What it does: Sierra launched Explorer on April 1 as an “agent-optimizing agent” that continuously analyzes conversations in the background, surfaces what is happening, explains why performance is moving, and recommends actions teams can implement. Sierra says it also delivers weekly briefings on what changed and what to do next.
CX use case: This is useful for teams that are past the launch phase and now need to improve live AI performance without spending their lives buried in dashboards. Sierra positions it around questions like what customers are frustrated about, what changed in the last week, and which fixes matter most.
Worth watching because: Most AI tooling still focuses on building agents. Explorer focuses on improving them after launch, where much of the real CX work lives. That makes it more relevant than another shiny bot release.
Bottom line: If your team already has AI in production, tools that diagnose and tune live customer interactions may matter more than tools that just help you launch faster.
⚡ 90-Second CX Radar
Service tech spend is climbing faster than labor is falling
By 2028, over 50% of customer service organizations will double their technology spend without an equivalent reduction in talent. Useful reality check for anyone still pitching AI as a quick labor story. The bill is going up before the org chart gets smaller.
Tata Play Fiber is fixing the data layer before the AI layer
Tata Play Fiber is building an AI-ready lakehouse that unifies 25 disparate data sources into one environment to strengthen retention and unlock cross-sell and upsell. Good. Most AI roadmaps still run into the same old problem: fragmented customer data wearing a new badge.
Experience orchestration is where enterprise budgets are moving
Genesys reported nearly $2.6 billion in ARR for Genesys Cloud in its fourth quarter and tied that momentum to demand for AI-powered experience orchestration. That does not prove quality. It does show where enterprise budgets are moving. Orchestration is no longer a side theme.
🧭 Your Move
Do not ask whether AI belongs in the customer journey. It already does. Ask where you are willing to let it act, what proof you need when it does, and how quickly a person can step in when trust slips. Visa is talking about machine payments. Gartner is talking about workflow outcomes. Customers are telling you that they still want a human in charge. The pattern is not hard to read. The next CX gap will not come from AI sounding strange. It will come from AI doing something the customer did not feel in control of.
Until tomorrow,
👥 Share This Issue
Think of one person who’s wrestling with AI in CX right now and forward this to them.
I’m obsessed with Wispr Flow Pro! Get a Free Month on me.
