Quantifying Mobile App User Experience Value: Design, Metrics, and Technical Implementation

In the mobile internet era, improving user experience (UX) is critical, yet quantifying its business impact remains a pain point; traditional technical indicators (FPS, FCP, TTI) cannot fully represent overall UX or its ROI.

The authors review existing UX measurement models such as Google Pulse, Google HEART, Alipay PTECH, and Alibaba UES, highlighting common dimensions like satisfaction, task efficiency, performance, usability, revenue, retention, engagement, acceptance, and consistency.

To bridge UX with business metrics, they propose a detailed scheme that defines and measures two key dimensions: booking completion time and effort ("费力度"). The design extracts three core interaction points—page jumps, user actions, and order creation—and abstracts them into duration, frequency, and depth indicators.

The technical solution follows four main steps: metric definition, data collection (埋点) infrastructure, metric calculation, and data visualization. Metric definitions include formulas such as booking duration = booking time – entry time, with careful handling of abnormal cases and noise filtering (P99, thresholds).

Data collection builds on existing front‑end click, show, stayTime, and lifecycle events, extending them with business‑specific parameters (e.g., user identity, product rating, region). A table of key埋点 keys and meanings is provided:

埋点key

埋点含义

app/app_flow/default/show/to

页面跳转埋点

app/app_flow/default/show/click

用户点击（参数扩展）

app/app_flow/default/show/lifecycle

页面停前后台切换

app/app_flow/default/show/stayTime

页面停留时长埋点

app/app_flow/default/show/bizData

业务数据埋点（新增）

Automatic埋点 generation is implemented in the React Native (RN) stack using page, module, and line‑number information, ensuring unique identifiers for each user action. Performance optimizations (simplified data concatenation, hook‑based click handling, blacklist configuration) reduce the impact on first‑paint and TTI from ~100 ms to ~10 ms.

Data processing follows a three‑layer architecture: ODS (raw storage), DWD (detail layer), and RPT (report layer). The pipeline aggregates page‑level timestamps, links them to order data, filters anomalies, and produces business‑ready fact tables for analysis.

Visualization leverages the internal low‑code Rock platform and Sylas service platform to deliver detailed, overview, and trend dashboards, showing per‑order path durations, frequency, and depth metrics.

Analysis of hotel booking data reveals that newly activated users exhibit significantly higher effort and longer booking times; targeted optimizations reduced effort by 5 % without harming conversion rates.

A repeatable SOP outlines the end‑to‑end process: embed埋点 SDK, persist data via scheduled jobs, compute and display metrics on Rock, and perform business‑driven analysis.

In summary, the effort establishes a foundational system for quantifying UX effort in mobile applications, combining front‑end instrumentation, big‑data processing, and visual analytics to turn raw user behavior into actionable business insights.