Everyone Knows the Problem. Nobody Owns the Fix.
AI can help CX teams find customer pain faster. But unless the business knows who owns the next move, all that intelligence just becomes a better-looking excuse for inaction
Customer experience has a dirty little secret
The biggest problems usually aren’t mysteries. They’re tolerated conditions.
The confusing bill, the broken handoff, the second call that leaves the customer with the same problem, dressed in a new case number: none of this is hidden. Even the policies usually make sense somewhere, to someone. The problem is what happens when those policies leave the conference room and land on a customer who’s already done everything the company asked them to do.
Then the CX team walks in with evidence: customer quotes, call clips, data, and customer pain points arranged like crime-scene markers. The room nods, because nobody wants to defend repeat calls or pretend that confusing a customer at the exact moment they’re trying to give you money is a growth strategy.
The meeting ends with agreement.
Yet, the customer keeps calling.
This is the expensive fraud inside a lot of CX work: the business agrees the pain is real before it agrees to fix it. Agreement can feel productive because it has all the right corporate theater: concern, notes, next steps, and the familiar promise that the issue is now “on the radar.”
But no customer experiences the radar. They experience the broken handoff, the surprise bill, the missing context, and the agent who sounds helpful but powerless.
A View Is Not an Owner
This is why I’ve been thinking about a question Bill Staikos raised in a recent webinar on the CX tech stack:
What system knows the customer best?
It sounds like a software question. It’s a question about power.
For years, the lazy answer was CRM, because CRM held the account, opportunity, contact, and owner. But the customer’s actual life with the company doesn’t live in one record. The anger may be in support, the surprise in billing, the abandonment in product analytics, and the most useful sentence buried in a transcript.
The company knows fragments and calls it a view. And we’ve spent years pretending that stitching the fragments together is the same thing as fixing the underlying condition.
AI Can Make a Bad Foundation Faster
AI makes this more interesting and more dangerous. It can stitch together tickets, transcripts, surveys, reviews, and usage data. It can find patterns that used to take teams weeks to surface. That’s useful. It’s also seductive because better pattern recognition can make it seem as though the company has become more capable of solving the problem.
It hasn’t. It’s become more capable of seeing it.
That distinction came up in last week’s DCX AI Roundtable on AI and voice-of-customer work. People are using AI to compress work that used to take weeks. But the tension kept coming back: AI is getting better at finding patterns, while organizations still struggle to turn them into owned change.
One participant said AI can classify feedback and organize themes faster, but the disconnect between that and action is still the hard part. Another put it more bluntly: “AI on top of a bad foundation just gets you to the wrong place quicker.”
AI doesn’t fix a weak operating model. It exposes it, and sometimes it accelerates it. A fragmented business doesn’t become aligned because a model found the pattern. If nobody owns the fix, AI just helps everyone admire the problem in high definition.
That’s the real platform fight: not who owns the data, but who owns the decision.
The next customer intelligence system won’t win because it can describe the customer more elegantly. It’ll win if it can answer the question the room is avoiding:
Who owns what happens next?
The Customer Gets the Whole Mess
Take a boring example, because boring examples are where the money leaks. A customer signs up under a promotion, gets a confirmation that feels official, and then receives a first bill that doesn’t match what they thought they bought.
Now the company becomes a scavenger hunt. Old emails, internal systems, offer language, billing logic, terms, product flow, and service apology all have to be reconciled after the fact. That is exactly why the customer ends up holding the whole mess together.
Internally, everyone can point to their piece of the truth. Externally, the customer has the whole problem. Inside, that contradiction becomes a tolerated condition: everyone can justify their slice, while the customer pays for the whole.
This is where companies say they need a “single view of the customer.” Maybe. But a single view of the customer isn’t worth much if it doesn’t lead to a single owner for the next decision.
The issue is not empathy. In most companies, plenty of people care. The frontline absorbs the frustration, CX sees the pattern, and product may even understand the issue, but none of that moves work across a roadmap. Owned action does, and owned action is scarce.
Owned action is scarce because it disturbs the machinery. It changes roadmaps. It forces tradeoffs. It asks someone with a scorecard and a backlog to absorb a problem the customer already paid for. This is where CX work gets political. Not office politics. Real politics: who has the authority to allocate resources, change priorities, and accept the cost of fixing what customers keep paying for?
CX Can’t Just Document Pain
That’s why CX can become the Department of Pain Documentation: a function that finds the problem, names it, quantifies it, socializes it, and builds the deck before handing the work to a business with no structural reason to move.
If CX only documents pain, we’ve accepted the wrong job.
The job is not to document pain. The job is to move CX from witness to operator.
In most companies, plenty of people care and plenty of systems can see the pattern. The frontline absorbs the frustration, CX sees the trend, product understands the issue. The real gap is cross-functional ownership: a customer issue may show up in service, but the cause may sit in product, billing, sales language, or policy design. If CX owns the evidence but another function owns the condition, the question isn’t, “Do we have the data?” The question is, “Can we get the right owner to move?”
Bill framed the older model as survey → dashboard → readout → action plan. The newer model, he said, is closer to signal → inference → workflow → decision → governance → action. In plain English: find the pattern, route it, decide who owns it, and make the work happen.
Put Ownership Before Action
I would insert one word before action:
Ownership.
Signal. Inference. Workflow. Decision. Governance. Ownership. Action.
That’s not a cosmetic edit. Governance tells the business what’s allowed. Ownership tells the business who has to move. Without ownership, governance becomes a conference room where accountability goes to die. The issue has been classified. The pain has been named. The recommendation is valid. And still, nobody changes the condition that keeps producing it.
This is where AI matters: not by generating prettier summaries, but by proving the cost of inaction.
If the business chooses not to fix a known friction point, what happens next? The answer should show the customer volume, call burden, affected revenue, and agent capacity burned on a problem the frontline didn’t create and can’t solve.
That evidence turns feedback from a moral appeal into an operating consequence.
Make Inaction Expensive
Most CX leaders I know aren’t short on evidence. They’re fighting for the authority to turn evidence into operating change. The best customer intelligence systems should help them do that by connecting the pattern to the owner, the decision, the cost of delay, and the next operating move.
Start smaller than the enterprise transformation deck. Pick one journey where the pain is visible and the owner is close enough. Map the customer signal to the internal condition, then make the operating ask explicit: who controls the condition, what decision is required, and what happens if the business does nothing. Then walk into the room with a better question:
Who owns the fix, and what will they do next?
If the room can answer it, the work has somewhere to land. Budget and priority are easier to find when ownership is clear.
If the room can’t answer it, pay attention. That means the company may know the customer well enough to diagnose the pain, but not well enough to change the experience.
The CX leader’s job is to build the mechanism that makes customer pain impossible to ignore and expensive to repeat, because once that exists, nobody needs CX to be the company’s conscience.
www.marklevy.co
Follow me on Linkedin
What Successful CX Leaders Do on Sundays
DCX Links: Six must-read picks to fuel your leadership journey delivered every Sunday morning. Dive into the latest edition now!
Thanks for being here. I’ll see you next Tuesday at 8:15 am ET.
👉 If you enjoyed this newsletter and value this work, please consider forwarding it to your friends and colleagues or sharing it on social media. New to DCX? Sign Up.
✉️ Join 1,600+ 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).