The real test for trends is time. Out of the dozens of new technologies and approaches that are born every minute, many will vanish by next year while few will grow until they will dominate the field.
This proves to be a challenge for automated testing professionals and newcomers, as it’s difficult for them to know which tools and methods to focus on and what to invest their precious time in. Is the AI trend in test automation tools here to stay? Will Selenium continue to reign supreme? Will Java grow? What new languages should be looked at? You get the idea…
Now that we’re at the midpoint of 2019, we begin to see emerging trends that come to fruition, and familiar ones that stick around and grow bigger and stronger.
I’ve mapped and decided to share with you a few directions worth considering that:
1. Don’t go chasing waterfalls
The transition from waterfall testing to Agile testing is a result of the overall shift to the Agile methodology in software development. This not-so-new approach creates continuous integration and iterations between developers and testers working on the project.Agile is believed to produce products of a greater quality as a result of a less structured, but more attentive, workflow. Immediate feedback is given throughout the process, and detailed documentation is often replaced with checklists.This straightforward approach is perfect for many testing teams and improves communication and results. For these and other reasons, we can expect to see more Agile teams in 2019 and beyond. In other words, the pressure for automation will only grow, as will the demand for faster executions. I predict that professional automation team leads will put extraneous efforts into seeking advanced technologies that can help them run tests faster and increase the quality of both tests and results.
2. DevOps I did it again
We can’t mention Agile without discussing DevOps. This methodology is no new trend and has been going strong for the past couple of years, with approximately 50% of companies implementing it by late 2017, according to Forrester.
2018 has more than lived up to its promise of becoming the ‘year of DevOps’, and we can expect to see this trend continue in 2019 and 2020.The holistic approach to development and testing dictates that the same logic which led us to build a feedback-based process for engineers and testers, will proceed to do so when it comes to feedback from clients and the changes it dictates.With DevOps, every stakeholder is involved – including the testers who now work much closer to the development teams. The process is only truly complete when customer feedback says so. This holds interesting career opportunities for testers, and I predict that testing pros will present higher collaboration with the developers; they will also be required to provide a more comprehensive analysis of identified issues, or root-causes, in order to make troubleshooting and debugging work more efficient.
3. Putting AI to the test
AI-based testing solutions will probably continue to be a leading trend in 2019 and 2020, for two main reasons: The first is that AI is EVERYWHERE, and the second is that it makes perfect sense for optimized automated testing.Test automation is one thing; smart, algorithm-based test automation is another. This is truly next-level testing, which takes into account the behavior of products and users and allows testers to keep their tests relevant without having to constantly fix them by themselves. Besides, challenges such as automated root-cause analysis are simply begging for higher, smarter powers to solve them, especially when coding with Selenium (for instance), which is a brilliant open-source environment. Still, it has quite a few frustrating limitations.AI-based tools will allow testers to not only automate simple tasks but to train the machines on how to identify flaky test patterns, in order to optimize them without forcing the testing engineers to continue adding bits of code to their commands, which are there to reduce false positive results.
4. Don’t stop ‘til you test enough’ – the continuous testing factory
In the era of lean, mean testing machines, Agile and DevOps are joined by continuous testing as a way of improving results and minimizing risk, while accelerating the development process and shortening delivery times.Automated tests embedded into the development timeline allow for better risk assessment and reduction. They don’t replace manual testing, but they cover all bases and scrutinize every stage of the development process, taking into account both consumer behavior and company business goals.
Once again, we see a more comprehensive approach that embraces a lean process in order to view it from a wider angle. Think about it as a shift to a never-ending testing. There’s no ‘testing cycle’, but rather 24/7 testing, befitting of the 24/7 changes of the Agile world.
I, therefore, believe that future testers would have to project plan an efficient operation of constant testing. That means a lot more automation, a lot more support by advanced technologies, advanced infrastructure, network, hosting capacity, cloud resources and more.
Aligning with a product’s testing cycle, even A2Z, is completely different from running a 24/7 ‘testing factory’ and requires different skills. I predict that more and more companies will hunt for test-automation pros that know how to build such “factories”
5. The need for speed
Some of the aforementioned methodologies might make testers worry about the loss of quality testing and results. We often think that the tradeoff for gaining speed and agility is an inferior product. But that would be wrong.Many in the testing arena have implemented solutions such as continuous testing as a way of enabling high-quality results at great speed. Better automation + better procedures = better products.By adding advanced optimization layers to Selenium, for instance, we can speed up the process and make it that much more reliable. That’s what great automation does.But automation alone is never the answer, and teams must also learn to embrace a culture of quality that is manifested in every aspect of their work. We wish there was a way to quickly automate that as well, but we’re afraid that this is one solution even the best script can’t replace.
6. Three’s company
What’s great about automation is that it makes developers and testers feel comfortable with the idea of not doing it all on their own, which opens the door to third-party solutions and codeless tools. Additional and specific layers are added to popular automation solutions like Selenium, thereby solving serious pain for testers.Third party solutions like Shield34 enable coders and testers to keep their development environment of choice (in this case, Selenium) and operate familiar tools in better ways.Shield34 simply adds an optimization layer to all existing and future Selenium tests, without otherwise interfering with them. Selenium coders can continue to code just like before but eliminate flaky tests, generate auto root-cause reports and run tests much faster.
With more tools going the open source route, the coding and testing community can enrich the pool of brilliant ideas and improve everyone’s workflow. One big happy testing family.
Perhaps the word “trend” doesn’t do these approaches justice. They are set to stick around for much longer than fashion styles, and will change the way we think, work, and test. Testers who are busy planning their career should pay close attention to these methodologies and the tools that enable them. They are likely to find themselves required to embrace them at their next, if not current, workplace.