Based on my MBA research on “Mobility as a Service”

Why I care about Customer Centricity

Every product I build, every AI model I use (or fine-tune), and every system I design revolves around one principle: a product only works if it works for real people.

This idea guided my MBA thesis. Although my research focused on Mobility as a Service, what I discovered extends far beyond mobility. The patterns I found are prevalent in fintech apps, marketplaces, AI tools, education platforms, healthcare systems, and nearly any product used daily by real users.

To understand what people actually experience, I analyzed more than 26,500 real customer reviews. I wanted to see what they complain about, what they value, what they ignore, and what they never forgive.

Once I stepped back from data, the message was clear: Human expectations are universal [1]. Industries change. Technologies evolve. But the psychology of trust, frustration, and satisfaction stays the same.

Below are three of the most impactful insights I wish I had understood better before building any product.

1. Negative reviews are signals of collective pain.

One of the strongest patterns I found was the collective response to negative reviews. When a user complained about something, others immediately validated it through likes and comments.

“Negative customer reviews are mostly found in the community responses or upvote aspect, where customers largely share similar negative feelings.”

This does not only apply to mobility. This is human behavior. When something is painful, users gather around it. When something is delightful, they simply move on.

This means that negative reviews should never be treated as angry outbursts. They are early indicators of some of the big issues. When many people resonate with the same frustration, it means the product is failing in an essential place.

Across industries, community-reinforced complaints predict churn, broken expectations, misaligned features, operational weaknesses, bad workflows, hidden bottlenecks. So, ignoring negative sentiment from a set of users (or even one user) can be one of the most expensive mistakes a business can make.

2. The number one customer expectation everywhere is reliability

In the dataset, I noticed people repeated one expectation more than anything else:

“From their point of view, the primary thing they want is to get accurate data.”

Even though this appeared in the context of mobility business, this insight applies to every product category.

Customers want:

• information they can trust
• results that make sense
• predictable performance
• consistent outcomes
• interfaces that behave the same way every time

Reliability is the foundation of trust.
And trust is the foundation of adoption.

This explains why some apps with simple designs outperform sophisticated competitors. Users prefer the product that works every time over the product that looks beautiful but breaks under pressure.

Industries change, but the psychology behind trust does not.

Big ears, small mouth

A business that doesn’t listen to what the customers are actually saying is going to collapse or, if not, be outcompeted by others.

It is one of the earliest tasks to build a channel of communication between between them and their users, and that should be easy and simple:

• A simple email will do (an email that’s actively responding to tickets and collecting feedback)

• A feedback tool embedded into the application will be great, so that users will not have to log out and open their email box.

• A zoom or Google meet link will be perfect, since it helps with building trust, which is everything when it comes to acquiring customers.

Listen. Listen. Listen. Crucial for a successful start.

I have to admit …

Though I just said nice and “smart” things about my work, I have to admit, it is one thing to say you understand these things and it is another thing to actually be persistent in applying them in your business. That’s a struggle I have to deal with every single time I build a product (most of my products, 98% of them, failed by the way, and I’m still learning). [2]

I’ll share more in the upcoming posts.

Notes

[1] "“Ask AI” is not a feature, it’s a UX pattern", from AI, by Kenn Kibadi

[2] "AI is purple 🟪 (LLMs)", from AI, by Kenn Kibadi (see where I mentioned my most recent product).