September 15, 2017
Managing technical debt with intelligent software quality analysis
By: Rajendra Prasad

Here’s an interesting conundrum that several of our clients face: up to 90 percent of their IT budget goes into maintaining legacy systems, squeezing the investments into next-generation systems and innovation.

Legacy environments carry both an operating cost and a cost of change such as remediating the code, applications, architecture and infrastructure—often referred to as technical debt—that becomes prohibitive as the costs continues to pile up. In fact, 85 percent of executives believe that legacy hinders their ability to move to a more digital model. [1] The push and pull of innovation versus legacy is a typical zero-sum game— the only way to boost budgets for new IT systems is to reduce the amount spent on servicing the technical debt.

Legacy systems eat up IT budgets for many companies—that’s called #TechnicalDebt. It doesn’t have to be that way.


Software quality analysis to measure technical debt

So, how do we at Accenture go about helping our clients reduce technical debt? An important first step is understanding and measuring the debt to determine which areas must be remediated first. This is where software quality analysis can be a powerful aid. A code and system level quality analysis to scan for violations of good architectural and coding practices, security vulnerabilities, duplicate processes etc. as well as effort required to fix those violations is a good indicator of technical debt. We often use CAST, a leader in software analytics and measurement, to estimate technical debt.

CAST Application Intelligence Platform (AIP) uses advanced diagnostics to identify the most serious structural flaws adding to the technical debt of the application. The debt is expressed as a dollar amount, and can be analyzed at the portfolio, demographic group or application level. CAST AIP not only quantifies technical debt, but also gives the development team all the necessary tools to proactively manage and reduce it.

At Accenture, we are taking this analysis to the next level by adding intelligence and cognitive capabilities to CAST’s analytical chops. We have integrated CAST’s suite of tools with Accenture myWizard@, our automation platform powered by artificial intelligence, and Accenture secure cloud to completely automate application analysis and reporting. At any point of time, project managers can view current state of all the applications in their portfolio, measured across several parameters such as efficiency, security, robustness, transferability and changeability along with a Total Quality Index (TQI). Project managers can determine which applications are most vulnerable and need immediate attention, evaluate technical debt based on software flaws found in the application and violations of industry-accepted best practices, measure the contextual complexity of the objects where the violations incur, and estimate the cost of labor to rectify the violations.

Improving TQI and reducing technical debt for a leading retailer

For a leading retail group, we used CAST to identify bugs and benchmark the code quality across individual quality factors like performance, security, robustness and maintainability of the retailer’s applications. CAST not only identified quality violations but also highlighted relevant code segments and recommended an action plan for improving the application’s TQI. By following the action plan, the number of violations came down from more than 1,800 to just 27 over a period of two quarters. Overall, the team was able to improve the TQI by 47 percent while reducing technical debt by a whopping 82 percent with the help of CAST.

Control future debt

Technical debt can become a serious concern if left unaddressed, in some cases costing millions to remediate. Some technical debt has a big impact, while most of it is pure hygiene. The key is to understand exactly where the technical debt lies and what its impact is. Software quality analysis tools like CAST use insights from data and cutting-edge analytics to measure the debt correctly and determine which areas must be remediated first, allowing IT teams to reduce technical debt by as much as 20 percent with a laser sharp focus for maximum impact.

[1] Research by Accenture Strategy: Is Technology Debt Bankrupting Your Competitiveness? 

Popular Tags

    More blogs on this topic