5 Ways To Improve Collaboration Between Software Testers And Developers - 99tests
Download the whitepaper to learn how to make your QA and Dev teams work together as a cohesive unit toward a common goal. In a development group, there are times when the QA and developer relationship can grow up noticeably exhausted. The main question arises. A healthy relationship between QA & developers can help in finding and fixing these bugs quickly. With good collaboration and smart.
They are getting all the pain of the waterfall process and missing many of the benefits of Agile. To truly operate in an Agile environment, entire organizations need to be Agile. Sure, that is easy to say, but what does that actually mean? It means that organizations need to be able to respond to change quickly and to iterate their software in response to that change. Here is an example that will make things a bit more clear Staggered iterations mini-waterfall Suppose a team is developing a feature for call center software that allows a customer representative to provide detailed notes about a call to their supervisor for review each week.
The software developer on that team's project writes the code for that feature during the current iteration, while the QA team is happily testing the features the development team had completed last week.
The development team finishes the feature and in the next iteration, they move on to another feature and hand the call notes feature over to QA to test. Well, it just so happens that when the QA team starts testing the feature, the customer who is going to use the software tries it out as well and discovers that they actually want the feature implemented in a completely different way.
Obviously, at this point it doesn't make sense for QA to continue testing the feature, so they stop. But now, the developers are already working on something else, so they either have to stop what they are doing and start working on the call notes feature again, or they have to put off the changes to the feature for the next iteration. Even if they manage to complete the requested changes to the feature in their current iteration, it will be a whole other iteration before the feature is tested.
How working together makes you more agile So, following an Agile process alone does not make one agile. In order to really be agile, in the true sense of the word, meaning an organizations can respond to change rapidly, they need to have their QA and development teams working together each iteration.
QA and Developers: Playing on the same team
Instead of developers completing code and throwing it over the wall to be tested in the next iteration, a truly Agile team will have developer and QA team members work together during the current sprint to both develop and test a feature.
By doing this, a team is able to respond to any changes immediately and is truly able to iterate the development of their software.
Going back to the example with the call center software. It can be seen that if the QA team was testing the call note feature in the same iteration that the developers were developing it, making changes midstream would be a much easier task. The developers could simply make modifications to the feature and give those modifications directly to QA in the same iteration. No time would be lost waiting for the feature to make it completely through the pipeline. Development and testing in the same iteration How to create a unified team Of course this is all easier said than done.
How can a team develop and test a feature in the same iteration? Doesn't the feature need to be developed before it can be tested? If a team looks at a feature as an atomic element that can't be broken up into smaller parts, then yes, it does.
But, most features are not unbreakable stones. Most features can be broken up into smaller pebbles which can be developed and tested independently of the whole. To look at features and develop them this way requires coordination and communication between the developers and QA analysts.
Developers vs Testers or How to Avoid Conflicts
When a new feature is going to be worked on in an iteration, the development team needs to meet with the QA team together to talk about exactly how the feature is going to be broken up and what exactly is going to be tested. The term test driven development, or TDD is used to describe the practice of writing failing unit tests before writing code in a software project. This same idea can be applied at a higher level to the development and testing of an actual feature to allow the tests to drive the development of that feature.
In this scenario the development and QA teams meet together and the first thing that is decided upon is the high level tests that will be used to verify the correctness of the feature.
Developers and QA analysts work together to define, at a high level, what test will be run and created to test the feature. Developers then start writing the code that will be necessary to make those tests pass, one test at a time. Each time the development team has enough code created to make a test pass, that code is handed over to QA to execute that test against the code for the feature. One by one, the code required to make each test pass is written and tested and little by little over the course of the iteration the feature is both developed and tested at the same time.
This process of developing features one test at a time requires the entire team to work together as a single unit to complete the work for a sprint. It requires developers to understand more about the testing process, since they will need to know how the features they are developing will be tested.
It requires the QA analysts to know more about the development process, since they will need to be aware of when certain parts of a feature are ready for testing.
Inevitably, when bugs are found using this process, they are handled immediately. When QA member finds a bug, it indicates that a particular test that a developer thought should pass, does not pass.
Work doesn't move forward until the failing test passes, so bugs are always fixed as the software is developed, not after. At least not bugs that are found by normal testing. Metrics and Dashboards Most Agile teams use either a burndown chart or a wallboard with different lanes to show the progress of items being worked on and to let the team know what the priority of work is. These tools can be an important part of getting development and QA teams to work together Having a central location where team members are able to see priorities and the progress of work being done during the iteration, helps the team to have a unified goal and to visualize how work will progress during the iteration.
When adapting a truly Agile process, as suggested in this paper, many teams struggle with a huge backlog of QA work being pushed to the end of the iteration. Breaking down features into smaller components and testing them as they are finished can reduce this problem, but it can also be very helpful to have a burndown chart or wallboard that clearly shows the work backing up before it becomes an issue.
Communication is key This whole process hinges on one very important thing: For many organizations this is the single hardest barrier to overcome. In many development shops QA teams are not used to communicating with development teams with the frequency that is required to be successful in an Agile environment.
There is no time for indirect methods of communication like complex bug tracking systems and long emails. Always think of your users and don't test just to say that you performed a test. Users don't care how many tests you ran on your application—they care about the product quality and that it answers their needs.
Share responsibility It's very simple: Everyone should be responsible for the quality of the product. In an agile team, there's no more "us" and "them.
QA should test the whole system. Yes, QA are the gatekeepers, but everyone in the agile team should have the same responsibility for the quality of the product. Three best practices ] 3. Choose your battles As a gatekeeper, you can't fight over every single defect. Understand which fights are worth fighting and where you can let go. Otherwise, everyone will spend a lot of time fixing things that are not that important. Define your own "red line" of things you simply won't compromise on, and focus only on those things.
For example, I am very particular about initial interactions with a product e.
Conversely, I've avoided fights about installation tweaks for on-premises solutions that only get installed once. Many teams set up a "defects committee," which predefines showstoppers vs.
Getting QA and Developers to Work Together
This helps focus everyone's efforts. Be constructive about defects No amount of testing will ensure that you have zero defects. Some will always escape even the most rigorous testing procedures and be discovered by external users. The key is to keep calm, learn from these "escaped defects," and improve your next release.
Developers love to ask QA engineers, "How did this get past you? We conduct risk-based testing and test the user flows we see as most important and common according to the time we have. In some cases we consult with product management, commercial stakeholders sales, pre-sales, etc. If something gets through our web, we do a debrief to discover what happened and why we missed it, and, we create an automatic test for escaped defects. Create visibility into your activities Visibility improves collaboration and trust in any team, and agile teams are no exception.
You shouldn't assume that developers or any other stakeholders know what you're up to, how you're testing, or what your standards are.
Review what you're planning to test with the developer. When you share your work with developers, they can start paying attention to the things that matter, upstream.
Having bug-hunt days, with additional stakeholders from product management, support, and architects, not only widens the scope of testing efficiently, but also gets more eyes to scrutinize the product. Publishing important lessons learned from customers has the added benefit of establishing you as a subject matter expert who represents the end user.
Don't punish developers or users I often hear QA threaten that they won't approve a feature because of low quality. In my opinion, this is really the worst thing you can do as a tester. Think of the outcome here: By not approving a feature, you alienate the developer, and, worse, your users will not have the chance to use it.