# Towards Using Data to Inform Decisions in Agile Software Development: Views of Available Data

## TL;DR

This paper explores how data can be used to improve decision making in Agile software development, emphasizing the types of data available and the potential benefits of data-driven approaches.

## Contribution

It characterizes the data types and decision levels in Agile teams and advocates for increased adoption of data-driven decision making in Agile practices.

## Key findings

- Data types and decision levels in Agile teams are identified.
- Data-driven decision making is highly applicable but underused in Agile.
- Using data can improve decision quality and project outcomes.

## Abstract

Software development comprises complex tasks which are performed by humans. It involves problem solving, domain understanding and communication skills as well as knowledge of a broad variety of technologies, architectures, and solution approaches. As such, software development projects include many situations where crucial decisions must be made. Making the appropriate organizational or technical choices for a given software team building a product can make the difference between project success or failure. Software development methods have introduced frameworks and sets of best practices for certain contexts, providing practitioners with established guidelines for these important choices. Current Agile methods employed in modern software development have highlighted the importance of the human factors in software development. These methods rely on short feedback loops and the self-organization of teams to enable collaborative decision making. While Agile methods stress the importance of empirical process control, i.e. relying on data to make decisions, they do not prescribe in detail how this goal should be achieved. In this paper, we describe the types and abstraction levels of data and decisions within modern software development teams and identify the benefits that usage of this data enables. We argue that the principles of data-driven decision making are highly applicable, yet underused, in modern Agile software development.