T-shirt sizing is an estimation approach that utilizes relative valuations. The estimation approach uses t-shirt sizes such as extra-small, small, medium, large, and extra-large (Alostad et al., 2017; Raslan et al., 2015). Similar to story points, the t-shirt sizing approach can differ from team to team and requires a common understanding of the estimated value selected (Alostad et al., 2017). The strategy works best when a team has estimated previous stories or work items as a group, and the method can provide a measurement for large effort work items (Alostad et al., 2017; Raslan et al., 2015). The t-shirt sizing technique can produce an early estimate to give the business a metric of complexity for determining the level of effort (McConnell, 2006). The early comparison allows stakeholders or requesters to determine if the effort required is worth the business value generated from the effort (McConnell, 2006). The t-shirt size estimation technique offers a simple alternative to executing a more complex estimation process.
The advantages of t-shirt sizing are that as there are fewer values to select and the voting process can be conducted expeditiously. According to Harzl (2017), due to the abstract nature of the approach, t-shirt sizing does not suggest precision. However, with t-shirt sizing, sizes are non-numerical, and the approach is simple and easily understood (Harzl, 2017). The t-shirt sizing method provides a nontechnical, initial estimation projection that is accurate enough to support effective project control (McConnell, 2006). The disadvantages in using the approach are that velocity is hard to measure (performance of the team over time), the scale lacks detail, there is no precise mathematical correlation between the sizes, and a numerical value to track effort actuals is lacking (Harzl, 2017). Thus, the absence of numerical values is problematic in establishing velocity.