Erik Brickarp gets the nod for my last session today. After facilitating three talks, it feels nice to just sit and listen :).
Erik's talk focuses on "A leap towards Context-Driven Testing". When chaos and discord start raining down on our efforts, sometimes the best break through comes with a break with. In 2012, he joined a new team at a big multinational telecom company. That team had a big, clunky, old school system for both documentation and loads of test cases (probably based on ISO-9001, and oh do I remember those days :p ). What's worse, the teeam was expected to keep using these approaches. To Erik's credit, he decided to see if he could find a way out of that agreement.
The team decided they needed to look at the product differently. Rather than just focus on features and functions, he also decided to look at ways that the project could be tested. In the process of trying to consider what the test approach had to be, they moved from multiple spreadsheets to web pages that could allow collaboration. By using colors in tables (as they used previously in cells) they were able to quickly communicate information by color and by comment (reminds me of Dhanesekhar's tutorial yesterday ;)).
By stepping away from iron-clad rules and instead focusing on guidelines, they were able to make their testing process work more efficiently. Of course, with changes and modifications, this welcomes criticism. The criticism was not based on the actual work, but they were upset that the junior team member went behind the back of the organization to "change the rules". Fortunately, due to the fact that the work was solid and the information being provided was effective, informative and actionable, they let them continue. In the following weeks, they managed to make the test teams deliverables slimmer and more meaningful, faster to create and easier to maintain. By using a wiki, they were able to make the information searchable, reports listable, and easy to find.
Erik admits that the approach he used was unprofessional, but he was fortunate in the fact that the effort was effective. As a lesson learned, he said that he could have approached this with better communication and could have made these changes without going behind their backs. Nevertheless, they did, and so they have a much more fun story to tell. The takeaway here is that there is a lot of things we can do to improve our test process that don't specifically require corporate sanction. It also shows that we can indeed make changes that could be dramatic and not introduce a ton of risk. Support is important, and making sure the team supports your efforts can help testers (or any team) make transitions, whether they be dramatic or somewhat less so.
Additionally, if you have a hope to change from one paradigm to another, it helps a great deal to understand what you are changing to and how you communicate those changes. Additionally, make sure you keep track of what you are doing. Keeping track doesn't mean having to adopt a heavy system, but you do have to keep track. Exploratory testing doesn't mean "random and do anything". It means finding new things, purposefully looking for new areas, and making a map of what you find. When in doubt, take notes. After all that, make sure to take some time to reflect. Think about what is most important, what is less important, and what I should be doing next. Changing the world is important, and if you feel the need to do so, you might want to take a page from Erik's book. I'll leave it to you to decide if it makes sense to do it in full stealth mode or with your company's approval. the latter is more professional, but the former might be a lot more fun ;).