Team First

It can be disheartening when I see posts online from Scrum teams that absolutely hated the experience and want to go back to Waterfall.  How can that be?  Agreed, Scrum can be a difficult pill to swallow at first when one is very used to the ebb and flow of the waterfall, but I feel the core principles of the Agile Manifesto and the core principles of Scrum should make sense and eventually impassion the team. So why does this not happen?

I believe one of the main reasons this might be happening is because some of us are not doing a good job as Scrum Masters.  We need to do a better job selling agile to the teams (and implementing, but that is for another post).  Selling does not mean shoving down their throats, blinding following the Scrum rules.  It means put yourself in their shoes and see the change from their eyes.  More meetings?  Possibly.  More transparency?  Of course.  More productivity?  Let’s hope so.  Less risk? Definitely.  These are not bad things for the team, but they may not see it that way initially.  It could take months, or longer, for them to “get it”.  Unfortunately, some of them may never “get it”, or may never want to.

Once we understand the changes that are occurring, not just in process, but also in culture, we can do a better job communicating to the team.  Each team member may see things differently, and we should acknowledge that fact and see if we can find a common avenue to help address issues or concerns across the team.

These types of discussions and transition plans could be well beyond the skill level and experience of a new Scrum Master, so in that case a professional agile or Scrum coach may be necessary.

We must not forget that a large part of Scrum and agile is team empowerment, and valuing people over process.  If the team is resisting, then the process probably isn’t working, and may need adjustment.  Use common sense.  I strongly feel that if the Scrum Master protects and empowers the team to the best of their ability, success in a better process in the end should be inevitable.  It may not be Scrum by-the-book, but I think that is perfectly fine.

Selling Leadership

If you are doing Scrum, or another agile process/methodology, hopefully leadership has bought in and is fully committed to going agile.  Perhaps they have not.  Even if they have stated that they are committed, they may not fully understand what it means to go agile.  Helping to form cross-functional empowered teams, providing dedicated Product Owners, and embracing and leading the agile cultural change, are all things that leadership should be doing.

If leadership does not appear to be fully on board, perhaps the benefits of agile are not clear?  Perhaps we need to do a better job selling agile to leadership.  Agile best practices are in their best interest as well.  Motivated, well-equipped, cross-functional teams working on the highest value items to the business at all times?  Reduce waste in process; increase productivity; increase transparency and thus reduce delivery risk.  Who would not want that?  Continuous Integration is an easy sale, in my opinion.  There are no real negatives that could possibly come close to outweighing the automation and reduced waste and risk that would ensue.  Test Driven Development is possibly more difficult to sell, but I am convinced it can be done with supporting arguments and information.

Everything can, and should, be started in moderation.  That also helps reduce the risk of spending time on processes or procedures that may ultimately fail.  It also reduces the risk that team members will be caught up focusing on areas that are not providing the highest value to the business.

With leadership support, agile teams can do so much more.  It is our job to help leadership see the benefits of going agile, and to try to obtain that support.

Selling the Client

I think much of the same arguments above apply to communicating the benefits of agile to clients or our users.  Shorter feedback loops, more transparency, adaptation to change; I see all of these as positives from a client perspective.  There is definitely a lot of adjustment that will probably have to be made (doing small iterations means delivering smaller chucks of features to the client at a time, which is possibly quite different than what they are used to), but there is a reason for this (see adaption to change and shorter feedback loops).

So if a Scrum or agile transition is not going as well as planned, or preparations are being made for a future transition, keep in mind that we must all be good stewards of agile.  We must be able to sell agile, to a certain extent.

After that, it should continue to sell itself.

How can a team accomplish so much (i.e. requirement docs, design, development, qa testing, etc) in 30 day iterations?   It can take weeks to get a large feature just developed, never mind the extra days for testing, extra days for bugs fixes and more time for regression testing.   I thought it would likely take a few sprints together, 60 or more likely 90 days, to fit all that in.

Those were my initial thoughts, both as a senior dev, architect and development manager.  My experience in those roles were telling me, “wow, this is kinda crazy. Kinda cool, but kinda crazy.”  It must work for somebody, otherwise Scrum wouldn’t be so popular, right?  I think so.

It is becoming more obvious over time where we need to make more cultural changes, process changes and further embrace the idea of team empowerment. What would you do if you had to cram all that effort and tasks into 30 days, and still have a production quality product ready at the end of each Sprint?

We are still moving in that direction, and I think we have a clear path so far.   We are automating a lot;  CI builds, soon Unit Tests, and whatever else we can.   Making the code self-testing, self-validating (read TDD for test driven development) appears to be the only feasible way of getting all the development done (and the testing), while maximizing the teams use of the full Sprint (i.e. no “code freeze” to fit in QA testing at the end of the sprint).

The rough part?  How to quickly move toward TDD on a very large, complex application suite that has been under heavy development for many years.  I will follow up shortly with some ideas on how we might address this issue.

Thankfully, our team has transitioned pretty well.  While we are not following Daily Scrum’s to the letter (i.e. we don’t always stand up; always go in the correct order; or require team members to stick money in the pot if they are a few seconds late), we do keep our discussions short (most cases shorter than 10 or 15 minutes.

Early on in the Scrum transition, one of the team members brought in a soft stress ball to the Daily Scrum.  They ended up volunteering to go first and then tossed the ball to somebody else (who inadvertently all eyes were on so they went next).  Since one of those first meetings, that is what the team has done.  They don’t go in order, or sit in the same place, but when they get the ball, they quickly go over the important information:

• What did I do yesterday
• What do I plan on doing today
• What is getting in the way of me getting my work done (Impediments)

Sometimes the team will go into more discussion right then and there when they think it is warranted, but they will keep it short (if it requires more time, they will parking lot the discussion and continue with it after the meeting concludes).  This is the format the team prefers, and it seems to be working really well – so no reason to change it.

One other transitional issue we had early on, was that most team members would look directly at and provide an update only to me (Scrum Master), rather than addressing the entire team.  First Tip:  I listened and gazed around the room, but avoided long-term eye contact with the person speaking.  They tended to start looking at others for eye contact, which is what I wanted.  Second Tip:  Due to other meetings and conflicts, on some days I was unable to make the Daily Scrum (but I was lucky enough to have another trained Scrum Master on the team to help run things).  I think my “not going” helped the team figure out how to update themselves and not worry about just updating me.  I would then always follow up with a few team members to see if anything critical or Impediments were discussed that I needed to be aware of.  Obviously, I still try to make every Daily Scrum that I can – but if I cannot, I feel it is critical that they still meet and talk to each other (as I am just there to facilitate and provide guidance when needed).

I am curious what other Scrum Masters do to encourage their teams to transition to a Daily Scrum rather than a status meeting?

We have not been using Scrum and Agile for a long time.  We are a little more than a year into our transition from waterfall to agile.

The transitional road has not been without its bumps, twists & bends, and occasional detours, but I think we are on the right track and working our way towards a more agile and productive team.

I plan to write a series of short posts to help myself and hopefully others reflect on our mis-steps and successes.  What worked for us may not necessarily work for you;  Nor to I want to give you the false impression that we are completely agile and following Scrum to the letter, because we have some gaps here and there.  Some gaps we are trying to close, and some gaps I am not sure we will even bother to address.  Our ultimate goal is to become more agile with each iteration, and continually find ways to improve our process, improve our team and the product at the same time.

This is tiny snapshot of where the project was more than a year ago:

• Requirements (what requirements?) were typically conversations or just an email with some description of what somebody wanted.  This was not necessarily a horrible thing, as it allowed the developers, business analyst and project manager/business owner to be creative and discuss the requirements as we went.  However the lack of documentation did cause problems along the way.
• Our release schedule was our iteration schedule (a release roughly every 3-4 months or so).  A rough plan was devised at the beginning of a release cycle, most of the time.
• Status meetings.  The team would meet roughly once a week or so, sometimes less frequent, with the project manager to go over each task status.  A lot of unknowns and large tasks resulted in a lot of “I am still working on it” from week to week.

Currently:

• We are still on a 3 month release cycle (as required by the business), but we work in 30 day iterations (or Sprints).
• Status meetings went away and replaced with a more natural quick “team discussion” called a Daily Scrum.  Here is where my team deviates a bit from the Scrum rules:  we answer the required 3 statements/questions 1) What did I do yesterday? (or before the meeting) 2) What do I plan on working on today? (or after the meeting) and 3) What impediments are preventing me from getting my work done?  Instead, each team member still goes through all three items, but more freestyle – just normal talking to the team.  Sometimes they take more time and get deeper in the details (rather than parking lot the discussion), but that is a conscious choice the team made, and it does not seem to be disrupting the flow or speed of the Daily Scrum too much.  As the team does not sit in a single room together, this is one of the way they mediate that and have a quick open discussion.
• Detailed requirements are being drafted, but as Use Cases.  We will likely start moving towards User Stories and using Story Points for estimating the product backlog.  Our BA attempts to focus on Use Cases a sprint ahead of the team, but this is not always the case, and the team will still work together to understand the requirements as they go.
• Review meetings:  The team demonstrates functionality and what they worked on during the sprint to our Product Owner.  This is a good place to get additional feedback on the overall direction, vision and change requests of the product from the Product Owner – which would make their way into the Product Backlog.
• Retrospective meetings: Arguably the most important meeting for the team from a cohesive, self-managing, happy and productive team standpoint.  This meeting has gotten noticeably more important, more detailed and more productive as we go.  The team is starting to notice more items that get in the way of them being productive – and are doing a better job in bringing these items up and discussing them with the team.  We try to tackle as many items, at least one, each sprint so we can always feel like we are changing something for the better – and so that the team feels that they are in control.  These discussions also occur at times during the Daily Scrum or anytime during the sprint.  We bring them back up during the retrospective so we can discuss and document it.  Depending on the item, we may try to address it during our sprint (if it is small enough, and makes sense to address it immediately rather than wait, and does not have a negative impact on what the team already committed to).

Both lists a actually much longer, but it should serve as an idea of the progress we are making.  The short of it is I think the team, and the Product Owner, is starting to see a lot of positives from the transition to be more agile.  The team is much more motivated and happy, and the productivity and product quality is continuously improving over time.  There is a lot of ground left to cover, and it may take us a while to get there, but this was definitely the correct direction to take from the fork in the road.

I will follow up with more stories, and links to useful articles and sites that I found to be helpful.