Tuesday, July 31, 2012

Deconstructing Weekend Testing: A Test Coach Camp Talk


As I was preparing talks for the Test Coach Camp session that preceded CAST this year, I had a number of ideas I wanted to discuss. Most of them were centered around Weekend Testing, since in my "coaching" career per se, that has been where I've done most of it and most regularly. As I was preparing materials for explaining the Weekend Testing model, I had a heretical thought... what if I ran a session completely deconstructing Weekend Testing? If for some reason, the main site were hacked and destroyed, and we had to start it all over again, what would we do?

With that, I decided to make that a core element of what we discussed. It was interesting to hear some of the feedback from participants, those who were interested but never did a session, and those who knew nothing about it at all. I hope this isn't seen as a blind side to the founders or other facilitators, it's not meant to be. It is shared with the spirit and understanding that we have something awesome, but even awesome can strive to do more, be more and get better.

What are some thing we discussed?

The first and most obvious aspects that were discussed was the main site itself.  When we go to the Weekend Testing site, we see prominently displayed the last three sessions of weekend testing that were held. these are experience reports,and they are displayed at the top. On one hand, this is helpful, because it shows what we have done and what we do. This is great for regulars who have missed a session and want to get caught up or see what happened. However, for most curious lurkers, they don't really want to know what has already happened, they want to know what will happen. When is the next session going to occur? Many of the commenters on my talk said that, instead of giving such prominence to past activities, we should be using that slide show to advertise the next testing sessions.

Tied into the previous aspect is how we find out about sessions. As I explained, the tradition has been that we advertise on twitter and via email about a week prior to sessions. It's also been to put the sessions up on the weekend testing site in the forum for announcing sessions. One participant explained to me the process of finding a session. They first:

Had to navigate to the site.
Click on the Forum link.
Click on Next Weekend Testing Session.
Scroll through and find the one that was in their geographical area or time zone that worked for them.

It's been proposed that these sessions and their times and details should be front and center, the first thing people see when they get to the site. I think that's a very reasonable request :).

Other comments related to how we construct the sessions. This is a common question, and one in which each chapter has addressed in different ways. For most of the sessions, we need to expect that they will be one-offs. That means that we need to make sure that what we present, what we discuss, and how we discuss it will be handled in our two hour meeting time. because of this, there are certain topics that we just don't have the time to get in to, such as more advanced automation techniques, or more detailed explorations of a product. We tried this with a approach that we called Project Sherwood a while back, and while it worked well the first few times, the problem was getting people to commit to a regular meeting time. This has led to questions about "asynchronous sessions", or making a weekend testing model that utilized the forums and allowed for a more in-depth discussion that, rather than just being for two hours, might continue for a week or more.

Another thought that was brought up was an interesting question... how could we encourage more facilitation? Right now, we have sessions that have one or maybe two facilitators? What if we were to make it so that there were many potential facilitators in a given session? How would we do that? What roles could we give them? What if one of the goals was a rapid capture of the session, so that the experience report was published within minutes of completing a session, rather than hours or days, which is sometimes the norm? What if there were ways that we could leverage the skills of the collective of testers to help streamline the process of making experience reports, so that the announcements become the experience reports, and the next announcements go up as quickly as possible? It may mess with the egalitarian model of everyone coming into the sessions on the same footing, but maybe that's OK.

There was a lot of other areas we talked about as well, and I am looking to engage the other facilitators and maintainers of the Weekend Testing site and approach as to how we can scale our creation and help it to grow and be useful to our participants. To that end, I'll ask all of you the same question... if for some reason, the Weekend Testing site were destroyed and we had to rebuild it, what would you like to see us do. More to the point, what would YOU be willing to do to see that vision become a reality :)?

No comments: