Soft vs. Hard: Uncovering the Hidden Thresholds of User Demand Analysis

Internet's boom created a new profession—product manager. Everyone claims the title, but the core skill—user demand analysis—cannot be faked.

A story about a king, a demanding princess, and a clever servant illustrates how misunderstood demands lead to futile efforts.

The team’s boss reflected, "Demand—maybe even God doesn’t know what humans need," inspiring deeper analysis.

With the rise of user‑first slogans, many self‑proclaimed product managers flood the market, believing they can fill perceived gaps without solid research.

Amid this hype, the article asks whether we can truly dive into rigorous user demand analysis, introducing the theme: hidden thresholds behind demand analysis.

User demand analysis skill schools

Two distinct schools exist:

• Soft school : focuses on ideology, psychology, user experience, group effects, deep interviews, and behavioral simulation.

• Hard school : emphasizes data modeling, profiling, big‑data analysis, and quantitative testing.

Why these two schools?

Before the internet, software served enterprise customers, so low‑cost solutions and tech‑background product managers dominated. The internet opened consumer markets, lowered entry barriers, and attracted non‑technical talent, giving rise to the soft school, while tech‑savvy designers formed the hard school.

In practice, most analysts blend both, but industry hype often traps them at various thresholds.

Soft school thresholds

1. Level 1 – Treat yourself as the user: Over‑generalizing personal assumptions as universal needs.

2. Level 2 – Colored‑glass limitation: Personal background and biases prevent true empathy with diverse users.

3. Level 3 – Lack of systematic psychology: No structured knowledge of psychology or social behavior, leading to fragmented insights.

Crossing these levels yields top‑tier soft‑skill analysts.

Hard school thresholds

1. Level 1 – Data are facts: Over‑reliance on raw data without context can mislead. Example: monitoring 2,789 WeChat accounts during a holiday showed overall sharing increase, but focusing on four major cities revealed a different pattern.

2. Level 2 – Data without viewpoint: Comprehensive metrics lack actionable insight without a guiding hypothesis, illustrated by classic “beer‑and‑diaper” misinterpretations.

3. Level 3 – Data describes everything: Excessive detail obscures core user needs; analysis becomes an end rather than a means.

Balancing both schools creates a more accurate grasp of user demand.

Our insight

The team debated whether the soft or hard approach should dominate in China’s internet landscape, concluding that neither alone suffices. True insight emerges from combining psychological empathy with rigorous data analysis, remembering that technology abstracts the world rather than merely describing it.