1. The Morning of the GCOO Team Was Still Intense: The Beginning of Change

A Summer Morning in 2024

I recall the time heading to the office under the scorching midsummer sun. As soon as the laptop opens, countless notifications pour in. As always, the day began alongside thick Excel files.

"How many came in yesterday..."

When opening the files, it left me breathless.

• 1,978 entries: More feedback than usual accumulated over the weekend.

A Typical (Or Not?) Day for the Product Team

09:00 AM - Starting Work

"Over 1,500 pieces of feedback came in over the weekend."
"Today might not be enough to read them all..."

10:30 AM - Urgent Meeting

"There's feedback about a payment error due to an app glitch near Gangnam Station!"
"When did it come in?"
"...Two days ago."

A Turning Point: A Single Piece of Feedback

"The app froze and was resolved after 5 minutes, during which an extra amount was chargedㅠㅠ As a student, even this amount is burdensome."

By the time this feedback was found, three days had already passed. It wasn’t just a simple refund issue. It marked the moment when a student’s trust in GCOO was shattered. On a swelteringly hot summer day, any inconvenience would have felt greater.

Changes That Can No Longer Be Delayed

We had our own 'system':

• Regular CSV file extraction

• Data analysis with pivot tables

• Color coding and notes...

But this wasn’t enough. Fatigue increased, and valuable feedback continued to get buried.

2. User Experience Told by Data: Discovering Truth in Feedback

"This doesn’t seem like just a simple complaint"

After reading hundreds of feedbacks each night, distinct patterns began to emerge through the User Journey Map. Like fitting pieces of a puzzle, users' stories started connecting into one.

Reality Through Numbers

Upon beginning feedback analysis, a clear pattern emerged:

• App usability-related: 28%

• UI/UX-related: 25%

• Location-based services: 22%

• Customer support: 8%

• New feature requests: 17%

But behind these numbers lay deeper stories. Understanding users' latent needs and the context of use was crucial.

The Truth in the Feedback

"The app frequently crashes. It needs to be more stable."

Is this merely a report of a technical glitch? There was a deeper context: The urgency to move quickly, the frustration from repeated failures, and the decline in trust toward the service.

Hidden Gems of Insight

Strikingly noticeable were users' proactive suggestions:

"It'd be nice to have a voice alert when entering a no-parking zone for bikes."
"In the morning, a bike is convenient, but I'm tired in the evening, so I prefer to ride a scooter."

This feedback wasn't just complaints, but seeds for service development.

3. First Encounter with AI: A Story of Mistakes and Growth

"What if we introduce AI?"

A sudden thought, followed by immediate doubt.

"Can AI understand the unique context of GCOO?"

The Failure of the First Attempt

• User: "The transfer discount isn't applied."

• AI classification result:

  • Category: General discount inquiry

  • Urgency: Low

  • Responsible department: Unassigned

Seeing the results, we sighed. 'Transfer discount' was a core service of GCOO, yet AI only understood it as a simple discount inquiry, failing to grasp the user's mental model.

Learning the Language of GCOO

First, we organized the unique terminologies of GCOO:

• Transfer discount, pickup zone, return area, parking discount, season ticket, QR scan...

The idea of 'teaching GCOO language to AI' brought a chuckle, yet it was an accurate expression. This process played a key role in helping AI understand the context of GCOO through domain knowledge building.

The Growth Diary of AI

Initially clumsy, AI improved day by day.

• First month: Basic classification

  • User: "The brake feels off."

  • AI classification: Device malfunction / General

• Second month: Beginning to understand context

  • User: "The brake feels off."

  • AI classification: Safety-related / Urgent response needed

  • Additional action: Temporary lockdown of the device

• Third month: Pattern detection

  • User: "It happened again..."

  • AI analysis: Reoccurrence detected, checked previous feedback history, suggested fundamental resolution

This demonstrates the learning cycle through which AI learns and improves from repetitive data.

Unanticipated Discoveries

Special patterns started appearing by time of day:

• 8 AM: Related to unlocking

• Lunchtime: Inquiries about return locations

• Evening commute: App performance issues

Mistakes Were Opportunities for Learning

Sometimes AI made erroneous conclusions. Yet, those mistakes were helpful. Analyzing the feedback AI wrongly classified uncovered new aspects of the service previously unnoticed.

4. Real-time Monitoring: How AI Changed the Daily Routine

"Now, We Can Finally Work"

Two months into AI adoption, work methods completely changed. No longer did we have to check hundreds of rows and columns every morning.

Moments of Change

"Ding-" At 9 AM, an unfamiliar notification sound rang.

[Emergency Alert]

• Location: Exit 3, Gangnam Station

• Issue: Brake System

• Status: Danger Detected

• Recommended Action: On-site inspection needed

Now, AI reads, classifies, and prioritizes first. Urgent safety issues are relayed immediately to the on-ground team, and suggestions for service improvements are sent to the product team.

The Change in Work Dynamics

Previously, incessant concerns prevailed:

"Which department should see this feedback?"
"How urgent is this?"
"Have there been similar issues before?"

But now, it’s different. AI identifies context, checks history, and tags relevant teams automatically.

Insights Generated by Data

Every Friday afternoon, we review data. It's now less about 'problem-solving' and more about 'envisioning the future of our service.' This process, part of heuristic evaluation, is a crucial activity for identifying problems and deducing improvement directions.

5. "From Feedback to Service": Beyond Data to See the People

Changes Beyond the Numbers

Six months after AI adoption, the changes in numbers were clear:

• Feedback processing time: 12 hours → 2 hours

• Urgent issue response: 24 hours → Within 1 hour

• Feedback omission rate: 15% → Less than 1%

But the real change lay beyond the numbers.

A Small Piece of Feedback Created Big Change

"The map shows e-scooter and bike locations too cluttered, making it hard to find them. In a hurry, I want to see what’s available at a glance..."

Among hundreds of pieces of feedback that came in daily, certain patterns stood out. Here, a new challenge was discovered.

"How can we display information more intuitively on the map?"

This challenge required reducing cognitive load while maintaining reliability.

The Start of Data-driven Decision Making

As we discovered patterns in individual feedback, the need for systematic analysis arose. This led to the beginning of the monthly report.

The Evolution of the Monthly Report

Now, a special moment arrives at the end of each month. It’s time to review the comprehensive report AI generates by analyzing all feedback from the month.

It offers not just statistics, but actionable insights:

• Users’ real voices: Recurrent patterns and emerging needs

• Current state of service: Improved areas and persisting issues

• Areas that need improvement: Priority and impact analysis

• Tasks for next month: Specific action items

Collaboration Enabled by Data

Monthly reports offer new perspectives to each team:

• Product Team: "This pattern indicates a need for new features."

• Development Team: "This issue seems to require fundamental technical improvement."

• Operations Team: "It would be beneficial to modify the on-site response manual this way."

A Journey of Continuous Improvement

The change that started from small feedback has now evolved into a systematic service improvement process. Though AI has become an essential tool in this process, ultimately, it’s all about 'people'.

Data indicates the direction, but the final decisions and executions are up to people.

6. "Making Every Moment Meaningful": GCOO’s New Challenge

Standing at a New Starting Point

New goals were written on the whiteboard in the conference room.

"Answer before the user asks"
"Make every moment special"
"Design a better mobility experience"

This goal considered overall user experience beyond mere usability.

Towards the Next Steps

We are currently preparing certain things. Not just 'responding' to feedback, but providing information that’s truly helpful instantly.

For example:

• User: "When will the 30-minute free coupon be available?"

• AI: "Let me guide you on the current discount events..."

Innovation Driven by Data

The monthly AI report is now an important resource awaited by all employees. Each team, based on this data:

• Product Team: Set directions for new services

• Development Team: Decide priorities for technical enhancements

• Operations Team: Update on-site response manuals

The Future We Dream Of

AI continues to evolve. More precise context understanding, quicker pattern detection, and more meaningful insight extraction. But this is just the beginning.

New Challenges

Current concerns of GCOO's UX team are:

• Real-time information: Automatic notification of ongoing improvements, proactive response to FAQs, customized provision of relevant information

• Advanced service improvement: Feedback-based automatic improvement suggestions, usage pattern analysis and optimization, personalized experiences

• Evolution in team collaboration: Real-time data sharing, strengthened interdepartmental collaboration, improved decision-making processes

An Endless Journey

Every morning, countless pieces of feedback still accumulate. But now, instead of being a burden, it marks the beginning of new possibilities.

"Commuting is enjoyable because of GCOO."
"Thanks to you, I’ve been around every corner of Seoul."
"I’ll look forward to working together in the future."

Reading such messages, I recall meetings continuing into early mornings. It reminds me of many trials and errors, yet one thing is certain: we will continue to create better services.

In Conclusion: Making Every Moment Special

GCOO, accompanying all your moments. We will continue to listen. Until every feedback leads to meaningful change. And we realize, true innovation starts with listening to the voices of users, not just technology.

A new morning has dawned. Surely, feedback will accumulate again today? But it's alright. We now have reliable automation and AI on our side.