TL;DR
If you want to watch the full CRMKonvo, please go ahead here (optimized for smartphones) or here (optimized for tablets/computers).
Else, be my guest and continue to read.Or do both …
The Myth of the Master Plan
Let us look at the numbers. Dr. Hill pointed out an interesting statistic from a major UK retail bank. When looking at the "manage my home finance" experience, only one-sixth of the interactions were on the actual mortgage pathway. Five-sixths of the interactions were customers trying to get help with estate agents, solicitors, or basic support. Yet the bank only cared about the mortgage pathway. They willfully ignored the vast majority of the customer's actual reality. Why? Because the bank is only interested in selling the mortgage. Everything else was viewed as an unnecessary cost instead of opportunity.
This is a fundamental flaw in modern CX strategies. We design for the happy path and happily ignore the real world. We are shocked when our highly polished onboarding process falls apart at the very first sign of customer complexity. We create rigid silos. Then we wonder why our churn rates skyrocket.
Maps vs. Reality
This brings us to the core problem of journey mapping. Mapping a journey assumes a static topographical reality. It assumes that if we just draw a line from point A to point B, the customer will obediently follow it. As a sailor, I can tell you that this is a dangerous way to navigate. You do not just draw a line on a chart and blindly sail forward. You look at the weather. You look at the currents, at the waves. You constantly adjust your route based on dynamically changing conditions.
Not only when sailing.
Customers plan their way based upon certain criteria, and they replan it every single day, every single moment even, because the circumstances change. Businesses, on the other hand, force customers onto a rigid track. When the customer inevitably encounters a storm, the business is nowhere to be found. In the ocean there is no fixed path. Why do we expect our customers to travel on rails? We provide a mapped path that is utterly disconnected from the underlying terrain.
The Requisite Variety Trap
Why do companies insist on this broken model? Dr. Hill points to Ashby’s Law of Requisite Variety. Providing a rigid pathway is cheap and manageable. If you only have one prescribed pathway to get a mortgage, you can manage the interactions and complexity. As soon as you enable customers to do what they actually need to get done, your complexity and costs increase exponentially. It becomes very expensive to try and provide everything for everybody.
So, businesses take the easy way out. They stick to the one pathway that works for them, even if it does not serve the customer well. They rely on the sad reality that the evil known to them is often better than the unknown evil. Moving to a new provider is a hassle. Customers stay where they are until it becomes so unbearable that they are forced to move. This is not loyalty. It is hostage-taking.
Dumb Automation and the AI Mirage
The push for automation is often driven by a desire to reduce costs. Companies follow the exact opposite of the Toyota Production System. Taichi Ohno taught that you make it easier for the worker, then faster for the worker, and only then cheaper for the company. Modern banks and telcos do the reverse. They implement what Hill calls "dumb automation" to make things cheaper for themselves. They make it faster for the company, but they make it infinitely harder for the customer.
When you force a customer to use a poorly designed app instead of talking to a human, you are not innovating. You are just offloading your operational friction onto the person paying you. This cost-cutting strategy is fundamentally flawed. When customers cannot get their problems solved through automated channels, they resort to other means. They call the support desk. They complain on social media. They switch providers. The cost of recovering from these failures is astronomical.
In blunt words: an automated dumb process stays a dumb process.
Now we enter the era of Artificial Intelligence – again. The hype is deafening. We are told that Generative AI and LLM technology will revolutionize customer service. We are told that chatbots (err, agents) connected to a RAG architecture will flawlessly guide customers through their issues. Let me be absolutely clear. If your underlying data architecture is a mess, an LLM will simply hallucinate solutions based on that mess. Using RAG to query a broken knowledge base will just give you highly confident, grammatically correct wrong answers. Best regards from Air Canada!
AI is not magic. It is a tool that accelerates whatever processes you have in place. If your process is designed to ignore five-sixths of the customer's reality, AI will just ignore them faster. True innovation in CX requires a solid architectural foundation. You need a unified data layer that provides a single, accurate view of the customer. You need integration across your entire technology stack. Your CRM must talk to your billing system. The billing system must talk to your support platform. Without this integration, your AI initiatives are doomed to fail.
Businesses must accept that customers do not care about their product pathways. They care about getting their jobs done. If businesses want to succeed, they must align their systems and processes to support those jobs. They must stop trying to control the journey and start trying to facilitate it. This is not a marketing problem. This is an operational and architectural challenge. It requires rigorous analysis, tough decisions, and a willingness to challenge the status quo. If you are not prepared to do that, you should probably just stick to writing press releases.
Strategic Recommendations for Enterprise AI Buyers
Here are the core learnings and recommendations for enterprise AI buyers who actually want to improve customer experience rather than just buying the latest shiny object. We are past the point of treating software like a magical incantation.
Focus on Architectural Integrity over Generative Hype
Do not be seduced by the promise of an LLM fixing your customer service overnight. Before you invest in any advanced AI, audit your data quality and integration points. If your CRM cannot communicate seamlessly with your CDP, your AI will fail. You must build a unified data architecture first. AI requires clean, structured data to function effectively. If you build on a cracked foundation, you will only automate your existing dysfunctions. Get your data house in order before inviting the AI guests into the living room.
Design for Exceptions and Keep the Human-on-the-Loop
Stop optimizing solely for the rigid product pathway. As said, the vast majority of customer interactions occur outside of your carefully mapped routes. Use technology to handle the predictable, routine transactions, but design your systems to seamlessly escalate complex issues to empowered human agents. Do not use automation to build walls between your company and your customers. Use it to provide a solution faster. A human-on-the-loop strategy is not a sign of failure; it is a recognition of reality. AI should augment your workforce. It should not isolate your customers.
Measure Customer Outcomes, Not Internal Efficiencies
Your metrics are probably lying to you. If you are only measuring handle time or deflection rates, you are incentivizing the wrong behaviors. You must measure whether the customer actually achieved their goal. Implement systems to track the entire lifecycle of an interaction, including the rework required when automation fails. Yes, that’s harder to measure. But, when you understand the true cost of bad automation, you will stop prioritizing short-term cost savings over long-term customer value. Enable the customer to achieve their goals, and the business results will follow naturally.
A customer is a consequence, not a means.
Comments
Post a Comment