How to Build Deep User Knowledge in Crypto Payments

The following is by Mesh SVP of Product, Dacheng Zhao

‍

When running a startup, it’s tempting to focus on quick answers–user surveys, dashboards, support tickets, interviews. And while those tools are essential, they often produce shallow knowledge: isolated facts about what someone said or did.

‍

But great products are built on deep knowledge: strong mental models of how users behave and why. That kind of understanding isn’t just about knowing the answers. It’s about being able to predict them before you look.

‍

In this article, I’ll explore what “deep knowledge” really means, why it matters, and how to build it through a simple but powerful habit: guessing before you check.

Shallow vs. Deep user knowledge

Shallow user knowledge is knowing what a user said in an interview or how they clicked through a flow.

‍

Deep user knowledge is knowing how they’re likely to act in unfamiliar situations–the tradeoffs they’ll make, what confuses them, what motivates them. It’s the difference between knowing what happened and being able to model what usually happens and why.

‍

This is especially important when building something new. When you're navigating ambiguity, you can’t just rely on past behavior. You need to develop a working theory of the user: how they think, what they value, where they get stuck.

‍

Without that theory, product decisions become reactive. You wait for data instead of predicting outcomes. You gather facts without understanding patterns.

Shallow vs. Deep ecosystem understanding

The same is true for understanding competitive landscapes or partner ecosystems.

‍

Shallow ecosystem understanding means knowing a company launched x.

‍

Deep ecosystem understanding means knowing why they’ll never launch y (and that they’d go to war to protect z).

‍

If you’re building infrastructure or partnerships, understanding incentives, power dynamics, and strategy is critical. The most successful operators aren’t just reacting to news, they’re anticipating moves based on how each player thinks. That kind of intuition isn’t built from headlines or press releases. It’s trained by pattern recognition.

Guess before you check

The most powerful habit I’ve found for building deep knowledge is deceptively simple:

Before you go look something up, guess what you think you’ll find:

• Before you ask a user for feedback, write down what you think they'll say.
• Before you check your product analytics, write down your hypothesis.
• Before you ask ChatGPT or check documentation, predict the answer.

This “guess before you check” approach does two things:

1. It makes every data point a learning opportunity–not just a fact, but a way to test and refine your model.
2. It builds intuition. Over time, your guesses get more accurate, and your mental models get sharper.

In machine learning terms, this is online supervised learning. You form a hypothesis, check it against the truth, and adjust your weights. If you skip the guess, you lose the feedback loop. The learning is slower and shallower.

‍

“Guess before you check” works because it makes your brain pay attention. It forces you to commit to a position, even if it’s a weak one. Then, when the real answer appears, it either confirms your model or forces you to adjust.

‍

Without that commitment, your brain absorbs the answer passively. There’s no friction, no comparison, no memory. You just consume information and forget it quickly. But when you’re wrong, it sticks. You remember the gap between what you thought and what was true. That’s how real learning happens.

Closing thoughts

Building deep knowledge isn’t about memorizing facts–it’s about developing working theories of users and systems. It’s neither fast nor glamorous, but it works and it compounds over time.

‍

So next time you’re about to run a test, open a dashboard or ask a teammate for data, then take 15 seconds and write down what you think you’ll learn. Your future product instincts will thank you.

‍

Want more like this? Subscribe to Mesh Weekly.

‍