By: Vengalarao Pelapolu, Specialist, Program Management Office, Canvas by LTIMindtree

A group of developers first introduced the Agile methodology as a more effective and practical approach to software development, and it is now revolutionizing the way businesses operate. Today, banks, manufacturers, research and development facilities, hospitals, and even airports use Agile for project execution. For software development, the adoption of Agile stands at a whopping 86%[i].

Agile frameworks such as Scaled Agile Framework (SAFe), Large-Scale Scrum (LeSS), Nexus, and others gained popularity as the scale and reach of Agile frameworks expanded. Enterprise agility and business agility are fundamentally altering the way organizations are structured.

Agile’s fundamental premise is that organizations are complex adaptive systems, not just a collection of departments that can function sequentially. The COVID-19 pandemic hastened the widespread adoption of Agile because organizations realized that it is a matter of survival and sustenance and innovated newer methods to be flexible and adaptable to customers’ ever-changing needs. Agile does just that!

Finding the right Agile fit!

When Agile is implemented at scale in institutions, the depth and thoroughness of implementation may vary. Organizations use various assessment methods to determine the maturity of being Agile as an indicator of true agility. Such evaluations typically include factors related to systems, procedures, tools, people, and culture. Organizations use key performance metrics (KPIs) to evaluate Agile attributes at the team, program, and business levels. Many of these measures are kept as simple as possible in their construction and design to avoid complicating the inherent simplicity of Agile. Burn-down charts, for example, show the progress of work completion.

Velocity is a team metric that shows how much work the team has completed in an iteration. While these are useful for visualizing product increase and resource solutions, there may be underlying patterns, trends, and insights in an Agile implementation that can be understood to sustain and continuously improve Agile at the business level. Because the volume and variety of this data will be diverse, organizations must employ more sophisticated data analytics techniques and tools to uncover these insights.

Understanding the quality of user stories is one example. Almost all Agile practitioners believe a user story must be simple and relevant. To ensure the quality of user stories, some use attributes such as INVEST.

We can use a natural language processing text analytics tool to determine whether the user stories written by the team are of high quality and follow the INVEST pattern. Similarly, to determine whether the product team will be able to release a product incrementally on time successfully, we could use certain mechanisms to determine the success rate of releases.

Agile environment empowered with insights!

The data generated in Agile is abundant. However, the generated data often goes underutilized. Here are four examples of how data gathering can be used in Agile methodologies to visualize and uncover insights:

1. Utilizing tree structure to decide release readiness
   Any Agile team must determine whether they are prepared to release. Although frameworks like Scaled Agile support release on demand, it is not always certain if the customer’s releases on demand will live up to their expectations. According to studies, Agile projects are three times more likely to be successful than waterfall projects[i].A combination of factors can be used to predict release readiness. Factors such as code quality, story burn rate, test coverage, defect fix efficiency, and managing product owner goals will determine the release’s success. These factors can be modeled using machine learning by setting thresholds at each level to determine whether or not the team must release. The success probability of a release can also be calculated using the same method, after which business leaders must decide whether to proceed with the release for a given probability.

2. Using predictive analytics to estimate flow predictability
   Many organizations seek predictability and agility when delivering products and services. While Agile methods encourage empowered teams to self-organize around value, predicting what the team will deliver in future releases can take time. This is due to the inherent nature of Agile, which emphasizes embracing change and complexity over sticking to an irrelevant plan unnecessarily (or artificially).
   On the other hand, business leaders may need to know the product output and a timeline to plan for cash inflows and market launches and answer shareholders. This predictability of business value can be calculated using data science predictive analytic techniques. The features delivered over iterations can be used to model the rate and predict the next release using time-series techniques. Another option is to model the factors that influence delivery outcomes. Both approaches will give business owners a reasonable perspective when making critical decisions to estimate delivery outcomes.

3. Using text analytics to assess the quality of user stories
   One of the most important artifacts in Agile implementation is the user story. If the quality of user stories is poor, it may impact multiple aspects, such as design, code, and the usefulness of test cases. Furthermore, if the user stories are unclear, it will confuse developers, designers, testers, and even customers, preventing the development team from producing high-quality products and services.
   As user stories are not quantitative, measuring their quality is difficult. This is where advanced analytics techniques, such as text analytics, can help. Many mechanisms can scan the user stories in the Agile tool and determine their clarity, coverage, and consistency. This will also aid in better estimation because poorly written user stories can be rewritten to be more consistent with an ideal reference.

4. Using database and code analysis to manage technical debt
   If not managed properly, technical debt will haunt the product team for a long time. Simply put, it is the amount of money you are willing to pay to meet the customer’s immediate needs. According to organizations, approximately 20% of the budget for new product development is used to resolve technical debt issues[ii]. Using advanced analytics techniques, organizations can save millions of dollars.The accumulation of this debt will undoubtedly affect the product’s stability and maintainability over time. Analyzing databases and code to determine when to refactor can assist in keeping technical debt under control.

Uncovering hidden insights through data

Data science has advanced in recent years to handle highly complex events and now powers many artificial intelligence applications around us. However, the use of data science or data analytics techniques in product development, particularly in an Agile environment, is still limited. The usage of AI in translating unstructured data into structured data can go a long way in deriving insights that can assist in accelerating software delivery.

Therefore, if organizations can use analytics to uncover hidden insights and patterns in an Agile implementation, it can provide enormous value to teams, organizations, and, ultimately, customers. To know more on how AI can assist in insights-fueled software delivery acceleration, visit Canvas by LTIMindtree

References

[i]  Agile Is Trending: 3 Ways Agile Makes Work Better, Tracy Brower, Forbes, April 24, 2022: https://www.forbes.com/sites/tracybrower/2022/04/24/agile-is-trending-3-ways-agile-makes-work-better/?sh=19c467e84398

[ii] Agile vs. Waterfall: Comparing Project Management Approaches, Abhay Talreja, Medium, May 1, 2023: https://medium.com/teaching-agile/agile-vs-waterfall-comparing-project-management-approaches-b1d1a7c9c08c

[iii] Breaking technical debt’s vicious cycle to modernize your business, McKinsey, April 25, 2023: https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/breaking-technical-debts-vicious-cycle-to-modernize-your-business