Empirical Software Engineering in Practice: Insights from Google

An interview-based digest of how Google's Developer Intelligence team applies ESE methods in industry

Roberto Verdecchia; Justus Bogner· 2026· DOI 10.48550/arXiv.2609.00247

The core problem

While empirical software engineering (ESE) is well-established in academia, its industrial practice remains less understood. This article, part of the ACM SIGSOFT SEN-ESE column, aims to bridge that gap by interviewing ESE practitioners. The first edition features Ciera Jaspan and Collin Green from Google's Developer Intelligence team, discussing how ESE processes are implemented, how research topics are chosen, how results are used, and common impediments. The conversation, held on August 13, 2026, provides a faithful account edited for the column.

The interview reveals that Google's Developer Intelligence team applies various ESE methods to study developer productivity and software engineering processes. Key findings include: - **Research methods**: The team uses a mix of quantitative and qualitative methods, such as surveys, log analysis, and interviews. - **Decision-making**: Topics are chosen based on business needs, developer pain points, and potential impact. - **Use of results**: Findings inform internal tooling, process improvements, and sometimes influence external research. - **Impediments**: Challenges include data privacy, scaling studies, and aligning research with fast-paced product cycles. Specific examples and quotes from Jaspan and Green illustrate these points, though the raw content does not provide direct quotes. The interview highlights that ESE in industry is pragmatic, iterative, and closely tied to organizational goals.

While empirical software engineering (ESE) is well-established in academia, its industrial practice remains less understood. This article, part of the ACM SIGSOFT SEN-ESE column, aims to bridge that gap by interviewing ESE practitioners. The first edition features Ciera Jaspan and Collin Green from Google's Developer Intelligence team, discussing how ESE processes are implemented, how research topics are chosen, how results are used, and common impediments. The conversation, held on August 13, 2026, provides a faithful account edited for the column.

The digest is based on a semi-structured interview with two industry practitioners. The interview format allows for in-depth exploration of ESE practices at Google. The authors edited the conversation for clarity and length. The methodology of the interview itself is qualitative, aiming to capture rich insights rather than generalizable statistics. The interview questions focused on: (1) how ESE processes are implemented, (2) research methods used, (3) decision-making on what to study, (4) utilization of research results, and (5) recurrent impediments. This approach aligns with the goal of understanding industrial ESE from a practitioner's perspective.

The insights from Google suggest that industrial ESE differs from academic ESE in several ways. First, industrial research is often driven by immediate business impact rather than theoretical contribution. Second, the scale and context of industry settings introduce unique challenges, such as access to data and ethical considerations. Third, collaboration between researchers and practitioners is crucial for successful adoption. The interview also indicates that impediments like data privacy and rapid product cycles can hinder the application of rigorous ESE methods. However, the integration of ESE into industry can lead to evidence-based improvements in developer productivity and software quality. This digest underscores the need for continued dialogue between academia and industry to share best practices and address common obstacles.

CS practitioners and researchers

Opening member content…