Guest article by CRXO Kathrin Michel
Why AI and CX programs provide more data than ever before, yet the customer experience across the journey still remains fragmented. And what this has to do with a structural decision most companies have not yet made.
Why AI and CX programs provide more data than ever before, yet the customer experience across the journey still remains fragmented. And what this has to do with a structural decision most companies have not yet made.
She had researched. Compared models. Configured options. Read reviews. For weeks. Her purchase decision was nearly made.
Then she visited the dealership. She wanted to take a test drive.
The dealership had to cancel. The demo vehicle did not have winter tires installed. Below eight degrees Celsius, the car was not allowed to be driven.
She left. Switched dealerships. Then switched brands entirely. And wrote about it publicly.
In how many CX systems does this moment appear?
In none. Not because the survey came too late, but because the person was never entered into the CRM system. No purchase. No test drive. No customer record. For the system, this person never existed.
The classic feedback model has a structural problem: it only captures people the company already knows and questions it actively asks. Lost leads, abandoned journeys, and unmet expectations leave no trace in the dashboard.
This moment is not an exception. It is the rule. And it happens thousands of times across dealer networks, regardless of how mature the CX program behind them may be.
Traditional automotive CX programs follow a simple process: survey, aggregate, visualize. NPS after purchase. CSAT after a service appointment. Dashboard. Quarterly report. Dealer rankings.
The problem is not the tool itself. The problem lies in the underlying data model: only customers already inside the system can be surveyed. And only surveyed customers appear in the dashboard.
What remains outside: the prospect who never bought. The dealership interaction that escalated without a ticket ever being opened. The question asked in a chat. The three star Google review. The pattern hidden across thousands of service emails. The service call recording no one ever analyzed. The feedback that was not submitted through a structured survey, but expressed publicly in a moment of frustration, unprompted and real. These signals emerge every day across dealer networks and disappear without ever being evaluated.
There is also survey fatigue. Customers in dealer networks are surveyed after nearly every interaction: after purchase, after service visits, after consultations. The result is declining response rates, more selective participation, and increasingly distorted data. Not because customers no longer want to share feedback, but because the model overuses them.
AI and CX technologies deliver more data and analysis than ever before. Yet the customer experience across the journey still remains fragmented because the most relevant signals emerge outside the measurement model.
Customer feedback was never a complete reflection of customer reality in automotive. We simply acted as if it were.
Most automotive CX programs consistently measure interactions once the customer is already inside the system: sales experience after purchase, aftersales and service, product feedback. That is valuable, but it is not where purchasing decisions are made.
The automotive industry is currently investing heavily in AI with the expectation that better analytics will close customer experience gaps. But customers behave differently today: decisions begin long before the first dealership interaction, usually in digital environments. Living room research, online reviews and communities, mobile configurators. Expectations are shaped by platforms like Amazon or Spotify: seamless, fast, intuitive.
What this phase produces is highly valuable: behavioral signals, review patterns, abandonment points. These are not edge cases. They represent the most honest picture of customer reality currently available.
What is often missing is a systematic understanding of why customers decide against a brand before they ever appear inside the system. And this is exactly where competition is won or lost.
A single negative moment, a canceled test drive, an unanswered chat inquiry, an expectation not carried from digital to physical interaction, can be enough to lose a potential customer. These moments often occur far outside structured feedback channels. And they determine whether a brand remains part of the consideration set in the next buying cycle.
The industry is investing in AI, and rightly so. But the assumption that more analytical capability automatically creates a more complete picture of customer reality falls short. Some vendors suggest AI makes surveys obsolete. Others claim unstructured data has suddenly become fully analyzable.
Neither is entirely true.
Text analytics, speech analytics, and social listening have existed for more than fifteen years. Major platform providers have offered these capabilities for a long time. What has changed is something more specific:
AI makes this analysis scalable for the first time. For dealer networks without centralized analytics teams. For importers consolidating signals across hundreds of locations. For markets without dedicated data science capabilities. What previously required six figure implementation projects and NLP specialists is now accessible to organizations of any size.
In practical terms: chat conversations can now be analyzed automatically for patterns. Service call recordings can be evaluated without manual listening. CRM comments, ratings, and app behavior, signals previously lost in the noise, become visible and actionable. Not as a replacement for surveys, but as an extension that addresses the blind spots of the traditional model.
The real shift is not the technology itself. It is the decision to rethink the capture model and use AI as a lever to uncover signals that were previously structurally invisible.
Companies operating mature CX programs today face a different challenge than organizations just getting started. The goal is not to replace what already exists. The goal is to systematically address the blind spot.
First: Expand the data foundation beyond traditional surveys. Reviews, social feedback, conversational data, and service call logs already exist across almost every dealer network. Yet they are rarely analyzed systematically. An importer with three hundred dealers generates thousands of these signals every day and still sees none of them inside the dashboard. This is not a technology problem. It is a prioritization problem.
Second: Focus on lost leads and journey abandonment points. Not only analyze what happened after purchase, but understand why customers did not buy in the first place. Why was the configurator journey abandoned? What failed before the first dealership interaction ever happened?
Third: Translate insights consistently into action. Today, the real problem is rarely data availability. It is the gap between insight and impact. Customer signals must reach the places where they can create change: inside dealerships, across networks, in the moment of customer interaction, and faster than competitors can react. Poor dealer rankings are often the result of signals arriving too late or never arriving at all. Strong rankings are the result of consistently translating insight into action.
Customer feedback is not dead.
But it was never enough. We simply pretended otherwise for a long time.
The prospective customer who switched brands because of a canceled test drive: her signal existed. It was real. It influenced a purchase decision and became public. Yet it still does not exist inside any CRM record because the data model was never designed to capture it.
Over the next years, automotive CX programs will not be judged by how well they capture what they already know. They will be judged by whether they can detect the signals that truly determine competitive advantage: before the first dealership interaction, outside structured channels, inside the moments of real customer experience.
AI is a tool. A more powerful one than ever before. But AI cannot decide which signals a company chooses to capture. That is a leadership decision. From requested data to real signals. From measuring to understanding. From understanding to steering action.
And it affects every OEM, every importer, and every dealer group, regardless of how advanced their current CX program may already be.
Kathrin Michel is CRXO at moveXM and deals with the question of how companies of all maturity levels can use customer experience as a real management tool.
If you are currently evaluating how your CX approach should evolve:
We would be happy to exchange ideas and discuss your perspective.