2017 Testing Trends: Prioritizing Improvement Initiatives

2017 testing trendsEvery testing team has a list of initiatives for 2017 that they want to work on when time presents itself. Some of these are related to creating better software, others are focused on improving process so the team can consistently release on time.

Inevitably, these initiatives become less important than the latest release or a needed fix to software already in production. The initiatives are often ignored for months, only considered again when tasks approach a due date.

Creating a list is a good place to start. Prioritizing that list and planning to invest in the most important can help the team accomplish their goals.

3,000+ people recently joined us for a special webinar Keith Klain, Head of the Software Quality Management practice and Tekmark Global Solutions,  to learn Keith’s  top testing trends to watch for in 2017.  You can catch the free webinar here.

Here are some of the biggest 2017 testing trends will see in the next year, along with advice on how to set priorities.

Faster Releases

Programmers are slicing tasks into small pieces, and delivering new software in two weeks instead of several months — thanks to Agile. Unfortunately, testing has not adapted as well. What most people are calling Agile is really a waterfall compressed into two weeks [Is Your Testing Team  Really Agile?]. Programmers might deliver the first new software of a release in the second week of a two week sprint. On the scheduled release day at the end of week two, the test group might get the last new code. They are testing and reporting new problems up to the last minute, whether or not they manage to get through every new feature or code change.

Testers have to adapt their approaches to a two week cadence if they want to help release software faster or, in this case, on time. There are two main ways to adapt testing work to faster development:

  1. Collaboration– Traditional testers find many easy problems, such as pages that fail on submission and errors from bad data, in test environments. These problems are considered low hanging fruit, not difficult for a tester to discover. Instead of waiting to harvest low hanging fruit, a tester and programmer could work together in the development environment to discover these problems earlier. Collaboration can save time in what is called the find, fix, retest loop.
  2. Technical Practices– Technical skill becomes important once collaborative work has reduced the number of bugs for a tester to find. A technical tester can do a variety of work:from discovering javascript problems in a browser to actually writing code. Instead of waiting for a new build, this kind of tester can help a programmer improve the code, he can build new automation frameworks and tests, or he can work on build and continuous integration systems.

Collaborative and technical work are tools to help programmers reduce the amount of time it takes to ship new code.

Collaborative and technical work are tools to help programmers reduce the amount of time it takes to ship new code.

Making Better Release Decisions

The most pressing question in every sprint is, ‘”are we ready to push to production?” Most teams answer this question based on feelings rather than data. Programmers routinely send new code to the test group with few or no unit tests. Testers will usually stop working on a new feature or code change when the number of bugs found per test-hour starts to drop. Instinct and gut feelings can be helpful, but they don’t make it clear whether a product is ready for release.

Testing coverage can be split into two categories:

  1. Technical– Technical coverage tools are installed on the build system. These tools analyze code and produce reports on how many lines, methods or functions, or variables are used by a test.
  2. Non-Technical — Non-technical coverage help the tester  think about parts of the product that the customer will be more familiar with, like features, menu options, and configuration choices. The combination of these test categories helps managers understand what has been tested so far, and what parts of the product are still untested.

Teams with an understanding of test coverage can make informed release decisions.

Fixing a Skills Gap

Some companies select employees right out of college as a strategy to keep costs down. The result is not surprising: stressful workdays and unhappy customers. What was intended to help a company save money ends up costing more in required repairs and customers who choose not to renew their contracts. Investing in technical employees can actually save money because skilled programmers make good software the first time which helps retain customers.

The best way to develop skill is by creating an event that happens regularly during normal working hours. Companies that do this make it clear that they value their employees’ development enough to give them time to learn during the work day. There are many ways to run skill development sessions. A team could focus on a problem it is currently experiencing, like building an API testing framework, and meet for an hour to discuss and solve problems. A session might be based on a rotating list of skills, like domain testing, using databases, or note taking. A group might use something like the dice game to create a scenario for people to think about testing in a different format.

The most important thing is to create a regular event and choose subjects that keep people energized and growing.

Investing in technical employees can actually save money because skilled programmers make good software the first time which helps retain customers.

Pick One

It’s easy enough to take a list like this, hand out one item to every senior leader, come back in a year, and see that not much traction.

Limiting Work In Progress, or WIP, could itself be a goal for many software groups. If that is the case, then adding a lot of improvement ideas will weaken each one. Instead of trying to do everything, pick the one that most defines where the department needs to be in a year, limit WIP to one or two, and release new initiatives only have the first one is actually complete.

Of course, do to that, the team needs to know what done is, so start with acceptance criteria, how the organization will know the improvement idea is done. With limited WIP, a clear direction, and defined acceptance criteria, the project is well on it’s way to success … even before it starts.

Check out our On-Demand Webinar: 2017 Software Testing Trends and view more great, free content from QASymphony in their Resource Center.

Leave a Reply

Your email address will not be published. Required fields are marked *

More Great Content

Get Started with QASymphony