一句话:提出一套面向百人规模中型企业、无需专职DX团队的开发者体验分析与改进方法,并给出情境化措施。
本文关注中型企业(约100名开发者)中开发者体验(DX)的可持续改进问题,这类组织通常无法设立专职DX团队。作者提出一套多阶段、系统化方法:首先进行以定量问卷调查为主并辅以项目分析的社会技术分析,共47名员工完整作答且未见系统性无应答偏差;随后开展由7名参与者组成的定性焦点小组工作坊;同时对企业代码仓库进行静态分析以识别可能影响DX的痛点。研究综合技术、组织与社会因素的交互,得出一组情境敏感的候选改进措施。经验表明,即使不设专职DX团队,也能以较低成本识别改进方向;所提措施在当前情境下可立即行动,但其长期效果尚待实证检验。
In recent years, development teams have broadened their focus beyond performance measures such as delivery speed and feature scope, now considering also factors like working conditions and developer satisfaction. These aspects have been systematically collected and summarized under the term Developer Experience (DX). The present paper sets out a structured concept for sustainably improving DX in a medium-sized corporation with approximately one hundred developers, where establishing dedicated DX teams is not a viable option. A detailed sociotechnical analysis is presented, examining how technical, organizational and social factors interact and how these interactions can be intentionally shaped to foster a productive, motivating and enduring development environment. A multi-stage, methodical approach was employed to analyse Developer Experience and to enable reasonably founded conclusions. The initial stage comprised a thorough, predominantly quantitative survey supplemented by project analyses. The second stage of the research involved the execution of a qualitative focus group workshop. A total of 47 employees completed the survey in full, with no indication of systematic non-response bias. The focus group, comprising seven participants, was utilized to further explore the prevailing circumstances. Another pillar of the study was a static analysis of the corporation's repositories to gain pain points that could influence the Developer Experience. This approach resulted in a set of context-sensitive candidate measures intended to enhance DX in medium-sized organizations. The experiences demonstrate that it is feasible to discern areas for enhancement with low effort and without the necessity of a dedicated DX team. While the identified measures are immediately actionable within the studied context, their long-term effectiveness has yet to be empirically validated.