In waterfall project management, software engineers create a feature and then throw it over to the quality assurance team (QA) for testing. The QA team then creates and executes comprehensive plans. They also register defects while methodically inspecting existing features for regressions that may have been caused by new work. This is how it works in traditional project management, i.e., by combining development and testing into two distinct processes.
Many teams take this approach or other traditional testing approaches. They later discover that with an increase in the complexity of a product, the amount of testing — and QA usually struggles to stay up. Project managers are forced to choose between delaying the release and cutting corners on testing. To complicate things, QA teams are generally rewarded based on the number of problems discovered, which puts developers on the defensive end.
What could be the solution?
What could be a better approach for developers and QA to minimize the number of problems in code?
Also, is there a new way to remove the uncomfortable trade-off that project owners have to take with the increasing complexity of products?
Wouldn’t that result in better overall software?
Agile Testing. The Solution!
Transitioning from traditional to agile testing methodologies
Teams using agile and DevOps strive to produce high-quality new features consistently. However, an agile or DevOps structure cannot accommodate traditional testing approaches. Due to the rapid pace of development, a new strategy is needed to guarantee quality in every build; it is also essential to identify whether these new testing strategies are in-line with the Agile Methodologies.
This blog will discuss ways to identify places where your current testing strategy might not be as Agile as you think, then we’ll discuss how you can adequately adopt an Agile Mindset.
Let’s look at some of the questionable approaches that are clear indicators of poorly implemented Agile practices.
You only use testers for software testing
These testers are merely cheap resources you hire because you don’t want your developers to do testing. This viewpoint opposes Agile testing — If you employ inexpensive employees to “test,” you are not conducting testing with an Agile mindset.
Testing is an activity, not a role. Everyone on the team should take part in this exercise. Even if your Agile team has a position with “quality” in the title (which I think you should really have!), they shouldn’t be the only ones conducting tests.
Some people might object, “But developers can’t test their own code!” They couldn’t possibly be the only ones that test their code, but they should and can.
It was effective in waterfall development to provide a clear distinction between development and test activities by designating all testing to a specific team of testers, but this is incompatible with the agile approach.
You infuse defects into everything
What would you do? When you discover a problem with the user story you are working on during a sprint? Considering a usual scenario, many teams would say, “submit a defect.”
In waterfall development, test teams would instantly have access to a fresh build with fresh features. After that, a testing cycle lasting a day, a week, or a month may begin. Given the number of defects that would be discovered and the time lag between discovery and correction, it becomes necessary to document each one of them.
In Agile development, this documentation is unnecessary.
When you discover an issue, work with the developer to resolve it immediately, preferably within the same day or sprint. If you need to keep information about the defect, include it in the original story.
Adding new, distinct documentation is optional.
You prioritize defects according to their importance
You, therefore, have a defect for a good cause. (Referring to the last point we just discussed!) The waterfall tester would assign severity and priority to the defect right away.
How do you define priority in Agile?
It is merely the sequence in which the defect is added to the backlog. The product owner decides whether a defect is a high or low priority or somewhere in between, and this is communicated by its relative position in the backlog in relation to the other stories and defects.
Giving each defect a different and redundant priority, which is recorded in a particular field, contradicts the concept of a prioritized backlog in Agile.
You discover a substantial number of defects in each story
The waterfall development philosophy was “developers build it; testers test it.” As a result, it was assumed that a considerable number of faults would be discovered when a fresh build was handed to the test team.
Many people’s Agile development practices now reflect this mindset. A story is developed, sent to QA, and numerous problems are discovered. The story is then given back to the developer by the QA for bug fixing, and this process keeps on repeating.
Finding substantial defects in stories indicates that you still consider testing to be a post-development activity rather than something that is done continuously as the story is being executed.
This contradicts Agile practices; a story’s lifecycle across an Agile board should be viewed as a process of steadily gaining confidence. If substantial issues are always discovered in the last phases, that means something is incorrect in the previous stage. Rather than considering your two-week sprint as a two-week waterfall, adjust your testing strategy to uncover these issues earlier.
As a test case manager, you meticulously detail all of the test cases
When a large number of features were thrown on a manual test team, having a plan for completing all of those tests was invaluable. There wasn’t much for testers to do while developers were busy building that first deployment. As a result, extensive test plans are required.
Agile stories should be short, and testing a single story should not have its own test strategy or an inventory of all test cases.
Does this imply that there will be no test documentation? Certainly not. It is still necessary to document what was tested, any required test infrastructure, testing issues encountered, and so on in the story.
Test planning, documenting test concerns and approaches, and so on are critical when testing in Agile. It is optional to detail each and every test case exhaustively.
You prefer to automate test cases
Given that painstakingly enumerating test cases is bad, automating those test cases is even worse.
Many professionals will say, “Automation is necessary for Agile.” Yes! It is, but you should not automate test cases.
Making 99 test cases from your test plan into 99 new automated test cases to be added to your ever-growing test suite is a guaranteed approach to developing an inverted test pyramid.
Inverted pyramids are evil, in case you didn’t know.
Agile teams must actively evaluate their automation on a holistic level. Teams should aggressively reduce test suites by removing unnecessary tests or moving items down the test pyramid.
Prior to prod deployments, you consider extensive regression testing
Testing cannot truly be called Agile if a “regression sprint” is required before you feel confident deploying to production.
Agile becomes less agile as you require more testing.
Agile testing should always aim to make all completed work, as part of the story, production-ready. The further away your testing is from being considered agile, the more there will be between finished stories and production-ready code.
Another way to look at this is to evaluate the definition of “Done” in your user stories. When time pressures arise, it is quite simple to begin cutting things out. The more you cripple your “done,” the less Agile you become.
You split development sprints from testing sprints
In collaboration with QA, developers create a lot of stories. At the end of each sprint, however, tasks related to testing or automation remain unfinished.
Instead of addressing the underlying issue (story sizing, estimation, dev-QA teamwork, etc.), the team opts to use a “follow-up” test sprint method, i.e., stories are developed in one sprint. Automation of those stories and testing happens in another sprint.
These repeated test sprints are an admission of failure. They lead to a more isolated, serial division of labor between development and test activities, which is the exact opposite of where your process wants to go.
So, how do you properly adopt an Agile Mindset?
The only approach to guarantee the delivery of high-quality software is through rigorous quality procedures and practices, which a knowledgeable and diligent QA can only carry out.
But what exactly does a diligent QA mean?
There are numerous factors for a diligent QA, but ultimately, we at AnArSolutions boil it down to adopting what we refer to as an Agile QA mindset.
What is an Agile Testing Mindset or Agile QA Mindset?
Let’s understand some guidelines and instructions to help you understand this mindset and how to foster it.
A QA should not be proud of identifying a bug; instead, they should be dedicated to preventing them. Anyone with an Agile QA perspective will only take joy in discovering a defect during the story testing phase. Instead, they should devise creative methods for avoiding bugs throughout the design or development phases. A defect uncovered early in a story, during analysis or design, costs substantially less than one discovered later in the story, during testing.
As a result, QA should be included from “Sprint 0” when the product is first formed and framed by research.
A QA should never presume that a feature has been correctly implemented
An Agile QA must have a driven mindset to break the system in order to provide a great product. After all, the QA’s responsibility is not only to ensure that everything is in working order.
When testing anything, QAs must be both destructive and inventive. One method is to create mind maps for test scenarios; this allows QAs to go beyond the current test site, examine the broader ramifications of a given test, and find potential branching.
A QA should not declare a test complete if it has only been run using a standard use case and test data
Any testing relies on use cases and test data. Trying multiple sets of test data and combinations in different use scenarios can give us more confidence in the tested feature.
In our experience, problems only occur when unusual data inputs — called “corner scenarios” — are employed.
Examples include the following:
- A bug occurs only when an unrelated background process is executing.
- At that exact moment, multiple people are attempting to upload big files.
A QA shouldn’t have to redo things repeatedly
It is generally good to have pre-made automated tests, data fetch helpers, or data setup helpers available when running tests or locating test data. You will ultimately save time, money, and effort by doing this.
Consider a project with no automation tests to act as safety nets because the manual regression cycle will take longer. As a result, releases will be severely delayed.
A QA must simultaneously ensure that nothing has been over-engineered. When planning to automate something, return on investment is an essential factor to consider.
A QA must know that quality is the entire team’s responsibility
It’s critical to understand that the entire team, not just the QA, is accountable for quality.
This mentality is essential because it guarantees that everyone keeps an eye on quality from the beginning of the design process until it is made available to the customer.
Following the release, we should employ instrumentation and user surveys to track how the feature or application is being used.
How to adopt an Agile Testing Mindset?
Let’s look at some ingenious ways you should adopt to help develop an Agile Testing Mindset.
Workgroup collaboration and communication
The Agile methodology relies on people to adapt to business demands and help drive the development process. Individuals must therefore be heard in order to build an agile testing attitude.
The testers must be taught to speak up and communicate, as the Agile process is built around delivering software continuously and satisfying customers. The ultimate objective for all parties involved is to produce the finest product and experience possible.
One of the following options must be present for optimal team collaboration:
- Creating Cross Dependencies — Clear objectives and work separation between several teams offer freedom for the teams while ensuring that different members do not do the same job.
- Removing Departmental Silos — The absence of departmental silos enables a successful agile transformation, which may necessitate a change in culture and structure to support the team organization.
Discover Rapid Evolution
Agile testing training may be the most effective way to shift to the agile approach. Developing a practical agile testing attitude, supported by adequate training, will allow for a smooth transition to the agile process.
Value Customer Feedback and Collaboration
Customer satisfaction is accomplished in agile testing through continuous software delivery. Although some customers may need more interest in how your product is made, these customers rely on the result to be satisfactory.
Rather than the customer simply receiving the end product, which may have difficulties in how they use it, if not impossible, to fix at the moment, an agile testing approach identifies glitches and issues as they arise, leaving the consumer with an overall better end product.
Second, Agile allows for greater customer control. Customer collaboration has a more significant role in agile testing than contract negotiation.
The customer plays a bigger part in the development process, and as a result, they are more able to provide feedback at every stage. This way, the product can be aligned with customer needs and expectations, yielding a considerably superior result.
Through this article, we aimed to help organizations successfully transition to agile testing; you can find out flaws in your current QA strategy and implement the suggested methodologies to quickly cultivate an agile mentality and appropriately adapt agile testing practices.
It’s up to you.
Agile is iterative and comes with numerous advantages; the overall quality of the outcomes improves. Furthermore, emphasizing users and corporate values results in happier consumers. Short testing schedules provide early and predictable delivery and predictable costs and timetables.