PREP: A Post-Requirements Effort Estimation Method in Scrum's Sprint Grooming

Effort estimation in the software development process using Scrum is performed to determine the time the development team needs to complete user stories and the sprint's complexity level. Inaccurate effort estimation can result in user stories from the previous sprint having to be carried over to the next sprint, budget overruns, project delivery delays, and incorrect schedule estimates. The inaccurate estimation occurs in one product development in company XYZ, with average estimation performance for only 57.12%. This performance is less than expected, 75%. This study designs a method that the development team can use to improve the accuracy of the effort estimation, focusing on improving the quality of user stories and the estimation process in an e-commerce company in Indonesia, XYZ. The effort estimation method design developed, PREP (Post-Requirements Estimation Procedure), is then evaluated by looking at the reduction in the number of bugs and calculating the estimation accuracy with Balanced Relative Error bias (BREbias). The PREP method experiment in product development in XYZ shows bug reduction from 46.7^% to 15.9% and increases the accuracy of the estimates by 23.63%. The effort estimation method validation shows a positive trend for delight, effort, and functionality aspects but requires a high level of understanding and more time to implement it.

I. Introduction

The agile software development method is an excellent way for companies to face competition, with 58% adopting Scrum as the chosen agile method [1]. Scrum allows companies to move faster, accept change, and adapt to dynamic business needs [2]. One stage in Scrum is sprint grooming which serves as a forum to unite the minds of the product owner and the development team about the tasks they must complete in one sprint [3]. An estimation method is used to determine the level of complexity and the time required for the development team to complete a task. Story points often represent the estimates [4]–[6]. This estimation phase is crucial because it ensures that development is completed within the planned time [7]. However, inaccurate estimates can impact the user stories in the previous sprint needed to be carried over to the next sprint [8]. In addition, this can result in budget overruns, project delivery delays, and incorrect schedule estimates [4], [9]–[11].