Eight‑Dimension Low‑Code Capability Measurement Model

Under the low‑code banner, many platforms compete, offering various application scenarios (PC admin, mobile H5, mini‑programs, cross‑platform like React Native), core functions (UI orchestration, logic flow, service orchestration), and interaction methods (form configuration, drag‑and‑drop, rich‑text extensions).

Common doubts arise: can mature commercial products meet real business needs? What are the real problems when standing on the shoulders of giants? Which stage are we in now, where is the next stage, and how to get there?

To answer these questions, a capability measurement model is proposed to make the evolution of low‑code platforms traceable.

1. Business Scenarios

The first dimension is the breadth of business scenarios covered. More scenarios indicate stronger low‑code capability. Example classifications include:

Product: B2C, internal admin;

Business: marketing activities, feedback forms, simple content display, complex rich interactions;

Endpoint: mobile, PC web, mini‑programs.

These categories can be further refined by importance (core, important, peripheral) and variance (technology stack, user groups) to quantify how well a platform supports target scenarios.

2. User Groups

The second dimension measures the size of the user base. Users are typically grouped by technical proficiency:

Specialist technical staff: front‑end, back‑end, DBA, etc.;

General technical staff: developers with some coding ability who can use expressions and events;

Non‑technical staff: product, operations, business, admin personnel without coding experience.

Platforms targeting non‑technical users must abstract functionality heavily, lowering the usage threshold and expanding the user base, which is a key indicator of platform maturity.

3. Capability Completeness

The third dimension reflects the completeness of technical expression. A fully complete low‑code platform can express everything that hand‑written code can, covering UI, front‑end logic, API calls, back‑end logic, and data models, enabling end‑to‑end development without platform limitations.

4. Material Inclusiveness

The fourth dimension assesses the platform’s ability to ingest various inputs, such as existing business components, third‑party modules, or non‑standard modules. Supporting custom components is essential for code reuse and maintaining development efficiency in long‑term projects.

5. Product Richness

The fifth dimension measures the diversity of output artifacts:

Final products: functional modules, pages, applications;

Intermediate products: business components, blocks, templates;

Basic products: UI components.

Multiple artifact types enable flexible integration, hybrid low‑code/source‑code workflows, and downstream customization.

6. Link Coverage

The sixth dimension evaluates how much of the full production chain (requirement → design → development → test → release → operation) the platform covers. Two improvement approaches are:

Inclusion: bring out‑of‑scope stages (e.g., testing, deployment) into the platform;

Connection: integrate with external tools (e.g., code repositories, IDEs) to bridge gaps.

Example integration diagram:

原料协议
物料资产 ----------> 低代码平台
产物协议
低代码平台 ----------> 发布平台/代码仓库
中间产物协议
UED设计工具 --------------> 低代码平台
接口描述协议
API管理平台 ------------> 低代码平台
数据描述协议
低代码平台 ------------- > 数据Mock平台

7. Collaboration Efficiency

The seventh dimension measures how efficiently different roles collaborate using low‑code. Examples:

Product managers can generate high‑fidelity prototypes and adjust content without developer assistance;

UED designers can connect design tools directly to the platform, eliminating manual hand‑off.

Design‑to‑Code automation further reduces the need for specialized coding skills.

8. Intelligence Level

The eighth dimension evaluates how intelligent the platform is. Intelligence is defined as the ability to assist or replace human decisions, such as automatic recommendation of layouts, component styles, images, copywriting, or even generating UI combinations for A/B testing. Configurable, data‑driven low‑code provides the structured input needed for recommendation algorithms, enabling data‑driven automation.

By combining all eight dimensions, the low‑code capability measurement model is established: