© 2019 SourceMedia. All rights reserved.

4 steps to reaping the software benefits of continuous testing

The business demand for high quality software continues to accelerate, and it’s no longer acceptable to deliver new software releases quarterly or even monthly. Changes to software development processes through Agile methods, continuous integration (CI) and continuous development (CD) processes have increased delivery velocity, but software testing can cause a bottleneck.

But there is an antidote — continuous testing (CT). By integrating tests throughout the entire software development process, you arm yourself with rapid feedback on the business risks and quality of the software while you’re still building it.

CT shifts testing activities left to be concurrent with development activities and not postponed to the end of an Agile sprint or end of your project. CT can cover any type of test, such as functional, API, UI, regression, performance, security, etc. With continuous testing, these tests run continuously and concurrently, are reliable, have fast execution times and are automated.

Benefits of continuous testing

Evolving your software development process to include continuous testing is a necessary investment so your team’s software is high quality and delivered quickly. Testing mitigates the risk of poor software quality by identifying defects before there is impact to your customers, your business operations and your revenue.

  • Avoid Technical Debt Testing is good, but continuous testing is better, because you can find and fix more defects sooner to avoid the accumulation of technical debt. Software technical debt includes defects in your software not yet discovered plus the backlog of lower priority defects waiting to be fixed. Technical debt adds to the complexity and cost of maintaining your software over time. With CT, you can take action immediately to fix defects to avoid adding to your technical debt.
  • Increase Developers’ Productivity CT decreases the time and cost to fix a defect. Research by Perfecto disclosed that a software defect fixed on the same day it was detected took only one hour, while it took eight hours to fix if detected at the end of a two-week sprint. It’s quicker for your developer to fix code when the code is still top of mind. By reducing the time spent on fixing defects, your developers can instead build new software functionality.
software developer 40.jpg
The silhouette of an attendee is seen while he works on an Apple Inc. laptop computer while participating in the Yahoo! Inc. Mobile Developer Conference Hackathon in New York, U.S., on Tuesday, Aug. 25, 2015. The Hackathon is an opportunity for mobile developers to come together and hack around the Yahoo! Inc. Mobile Developer Suite. Photographer: Victor J. Blue/Bloomberg

  • Increase Test Engineers’ Productivity Continuous testing builds automated tests into the CI/CD process to check code quality before deployment. CT shifts your testers away from labor intensive manual UI testing to automated API tests, which you can develop early in a sprint and which requires less maintenance. Your test engineers can then focus on creating more automated tests for high priority business critical functions.
  • Quicker Automation ROI Automation requires an investment in building or acquiring automation frameworks/tools and resources. Continuous testing provides a quicker return on investment (ROI) for automation, because automated tests are continuously executed. The more frequently you run these automated, the quicker their development cost is paid off (i.e. breakeven point) and a savings is realized over manual testing.

Getting started with CT

Leaning on CT removes the testing bottleneck while reducing your costs and increasing your team’s productivity. CT is a true game changer for any software development team. So where to begin? Follow these steps to adopt CT into your software development process.

  • Implement Agile – If your team isn’t yet using Agile methods to build your software, now is the time to implement it. Using Agile ensures that quality is a team responsibility and testing is integrated throughout the development process. Plan ahead for adequate testing hours in your sprint; otherwise, CT will fall by the wayside. Include test engineers in your sprint planning to review user stories for testability and to better plan their manual and automated tests. Your user stories’ Definition of Done (DoD) must include both coding and testing completed. Scrum masters — the facilitators of the Agile process — are instrumental in integrating testing within each sprint.
  • Implement CI/CD Your developers must commit their code and integrate it more frequently within a sprint so it’s continuously available to test; otherwise, it is impossible to shift testing left and perform continuous testing. Because of this, you’ll need a solid CI/CD process to easily integrate and deploy code to various environments, such as your development, QA or staging environments. Once you have this, you can then perform testing as soon as each software feature is completed instead of waiting until the end of a two-week sprint.
  • Design Proper Tests – Proper test design leads to high confidence that when a test fails, it’s due to an actual defect and not a brittle test by design. So instead of designing tests for long, end-to-end business workflows, your tests should be small and modular with a single purpose. Once they’ve proven their stability, you can then combine these small tests into larger workflow tests. This design approach allows you to repurpose the tests and avoid extensive maintenance on the test code as the software changes.
  • Implement test automation – You cannot achieve continuous testing without automation because automated testing helps you keep up with the pace of development activities. You can start the test automation as your developers concurrently build first the back-end services. Your automated tests should focus more on APIs and less on the user interface (UI). By using automated API tests, you can efficiently test more data combinations for a quicker feedback loop. For consistency when doing this, you’ll need to put an automation framework/tool in place — either open source or commercial — using the same technologies as your development team.

Implementing continuous testing into an already fast-paced software development process may seem daunting. But if you’ve already implemented Agile and CI/CD, then you have the prerequisites in place to fully integrate testing and reap the time, cost and risk mitigation benefits of CT.

By beginning your software development modernization journey with CT in mind and following these four essential steps, you’ll be able to deliver high quality software faster to meet your business needs.

For reprint and licensing requests for this article, click here.