Test automation is a big topic. There are so many different areas to talk about: tool choice, jumpstart, cross platform, services, cloud… Each of these areas have changed so much in the recent past that they could each be worth their own magazine issue.
As software development has changed so dramatically in the past decade — from same sprint Test Automation in agile, to Selenium as a universal browser test tool, to service testing using containers in DevOps — so has Test Automation.
But at the same time I am noticing a new trend that has not been as obvious, or as striking in the past. It’s the phenomenon of the ‘have’ and ‘have nots’! That is, those who have significant Test Automation programs and those who do not.
Some organizations have become quite sophisticated in their automation, some struggle with the very basics, and most organizations are somewhere in between.
From the first survey we did in spring earlier this year, we found that there was a very large number of companies who do not have Test Automation at all. At the same time, we know that there are companies where the Test Automation software development project is nearly as sophisticated as the production code project. Tens of thousands of automated tests with sophisticated development and maintenance methods are running on hundreds of virtual machines or real devices in a state of continuous testing.
Between these extremes, there are many levels of automation. For example:
- Small, but hopefully, effective automation, such as one automated test or smoke test suite that run against each build.
- A few happy path workflow, or transaction tests, or full transaction tests that touch perhaps — all the outside services or major functionality and are run against each platform.
- A small regression suite with test design and maintenance and cross platform abilities, but also with gaps, perhaps, of various services, subsystems, or devices and platforms.
- Significant automation program, high coverage diverse platforms, significant maintenance, and manageable.
- Sophisticated automation that scales at a high volume, with an automated process that runs on a significant amount of systems and has its own development team.
As always there are constant demands to go faster, do more with less, and automate more.
Three aspects of this situation put pressure on test automation programs:
- There are also universal goals to test automation, such as higher coverage and lower maintenance.
- That each level of automation has its own unique issues.
- That with Continuous Integration and Continuous Delivery in DevOps, each of these levels need to tie into the ALM tool chain and run automatically.
A critical task when optimizing your test automation suite, or beginning to build a test automation program, is recognizing the issues. This is where LogiGear Magazine can help.
I suggest that you look at the infographic “Top 10 Must-Haves for Test Automation” to help you recognize areas you can improve on in your practice.
In this issue, we take a dive into the world of API testing, a growing skill for testers in automation. But more exclusively we have a fresh cover story breaking down voice apps and the effect it has on Test Automation.
Prashant Hedge tells us everything we need to know about API testing in the Blogger of the Month spread. Lawrence Nuanez acts as our guide in “Climbing Mount Automation.” The 3rd survey results are here to help us reflect on test automation, based on the feedback from our readers.
Beware of the dangers of Test Automation in yet another exciting infographic, and don’t miss out on your chance to win a $100 Amazon gift card by taking our last survey in the State of Software Testing Survey Series — this one is on “Modern Distributed Test Teams.”