Dr Venu Murthy is a Technocrat with 20+ years of experience in driving digital transformations of fast growing companies and an innovator.
Technical debt and code quality significantly impact customer experience and revenue. As a CEO, overseeing code quality and tech debt management is crucial. A comprehensive tech dashboard facilitates continuous improvement across these three crucial dimensions.
1. Customer experience enhancement.
2. Project/team health.
3. Individual programmer/developer growth.
Identifying The Underlying Causes
In my decades of experience, I’ve found that the cause of compromised code quality stems from the desperation to hit the market faster with a product/feature.
This urgency leads tech teams to rely heavily on subjective feedback during code reviews, further compounded by the challenge of walking-the-tightrope of maintaining the legacy apps as well as developing code using modern technology for adding newer capabilities.
Furthermore, code reviews are quite repetitive and can eat away at the code reviewers’ time, which is generally the TLs (team leads) who would already be juggling pressing issues under delivery pressure.
Negative Current Trend
A prevailing trend in response to these challenges has been to increase the number of code reviewers. However, this reactionary approach merely exacerbates inefficiencies and frustrates developers. They have follow-ups with the reviewers to get their code promoted to the next stage. To mitigate these instabilities, my strategy prioritizes delivering new code with the utmost quality.
Where To Start?
My approach has always been to “shift left” to address issues early in the development process as debt or bugs get expensive as we move toward right (product), so let’s go from left of SIPOC (Supplier, Input, Process, Output, Customer) and work our way toward the end of the value chain.
By embracing a comprehensive competency matrix for Devs and QAs, we ensure all team members possess the requisite skills to produce high-quality code based on the following gradients.
• Problem-solving.
• Coding.
• Design.
• Architecture.
• Delivery.
• Org Building.
Cultivating A Culture Of Excellence
Our commitment to excellence extends beyond technical proficiency, encompassing cybersecurity vigilance and compliance training like GDPR and HIPAA. (My take on this is in an earlier article.)
The competency matrix guides hiring, appraisals and aligning salary revisions with performance. Recognition of innovation fosters a culture of thinking.
I’m passionate about attracting the best, so I participate in the drafting JD for devs and QAs. The team engages in frequent knowledge-sharing sessions, where code reviewers highlight exemplary code and address recurring issues, either during daily stand-ups or through lunch-and-learn sessions for in-depth discussions.
Moreover, even the QAs meticulously examine product backlog items (PBIs) before sprint planning, delving beyond the “happy path” to uncover edge case scenarios. Equipped with a suite of testing tools and IDE extensions, they ensure thorough testing across platforms and versions, addressing code quality and security concerns in real time during development.
Automating Quality Assurance
Moreover, within the automation realm of the continuous integration (CI) pipeline, actions are initiated upon the creation of a pull request (PR). Subsequently, the code undergoes scrutiny at coding gates, which assesses test coverage against predefined thresholds. For instance, a benchmark of 70% coverage is commonly set, subject to reassessment if issues are identified during root cause analysis (RCA). Furthermore, this automated process extends to evaluating code complexity, detecting code smells, identifying potential toxicity and flagging vulnerabilities, providing a comprehensive evaluation of the codebase.
Integrating Metrics Into Performance Evaluation
Code analysis metrics serve as KPIs, seamlessly integrated into performance appraisals.
Only when the TLs either give feedback and the respective changes made is the PR approved and the changes move on to the next development stage, that is to the QA environment, which have also been spawned on need automatically using the IaC scripts provided by the DevOps.
QAs take up the testing for happy path and edge cases; they do functional, usability, regression, smoke testing, and performance testing, and my teams are also accountable for the SAST; compatibility testing is almost automated in the release pipelines. QAs have the KPIs defined in the lines of error points and also the defect discovery rate; they are also responsible for testing the documentation and user guides.
Driving Continuous Improvement
The build and release are architected and maintained by the tech architect (TA) with real-time alerts and notifications set up at various levels of the CI/CD by working with the engineering managers (EM) to have any changes documented as PBIs and pass them through the sprint cycle, and clearly defined escalation mechanisms are put in place to bring to my notice when thresholds are breached. Based on the issues reported and RCAs of incidents, new gates or checks are added or updated to the build and release pipeline using different tooling, like SAST, for vulnerabilities.
Catalyzing Customer-Centric Innovation
On the project level, the product team and EM work collaboratively to document and manage the tech debt, which goes through the sprint cycle. Project dashboards are maintained to display detailed static code quality analysis and insights and the trend analysis, which are depicted through graphs presented to the leadership in all hands and such high-level meetings, which might also include the customers (when we have good news).
On the customer account level, we have the data on our dashboards, which consolidates the number of issues raised and all other support-related stats from the helpdesk for the given time period by the customer success manager and presents to me which is tallied with the trends of aggregated code quality metrics, tech debt and retros to have clearly defined action items for continuous improvement which then make their way as PBIs and go through the sprint cycle.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
