Doing. Well.

This is my final post as the CG-LIMS project manager.  It has been an absolute delight to serve you.  I am so proud of the work we've done together.  I know it has been a tough road at times.  If this business was easy, we all would have left it for more challenging assignments long ago.  Hopefully you choose to stay because this is important work.  Aircraft don't fly and boats don't get underway without the support of IT systems that are over 20 years old.  We're replacing the oldest systems with more modern, integrated tools.  Once we replace ACMS at the unit level, we will have fielded a mission critical system that maintainers and operators trust with their lives.  That's an awesome responsibility, and an awesome privilege.  The system you deliver will be an integral part of the great and noble missions accomplished using the assets managed in CG-LIMS.

Before I close, I'll share something Jim posted on his cubicle.  He can't look at the computer without seeing it out of the corner of his eye.  It's a single slide from the presentation I shared with the whole development team when Mythics came aboard in March 2012.  Given the many changes in direction we've all experienced, it was nice to see something that is as relevant today as it was on day the dev team was formed at ALC.  I shared three main topics that day:  Project basics, What success looks like, and How Mythics can add value.

Here are the five things on the "Success Looks Like" slide that Jim posted by his screen:

Reduce the burden on our field units 
Poster child for Agile COTS implementation 
Transparency within team and users 
SDLC docs match real work of development team 
Replace ACMS and AMMIS with COTS tool that meets enterprise requirements and scales to replace other systems

That's as true  today as it was in March 2012.  That's what success looks like.

Right at the top is the most important thing we must accomplish:  Reduce the burden on the field.

Finally, I want to share just five of the things I think we're doing well.  I could have listed 500, but here are just five things I think we did well together.

1. We brought real users into the development process from the start.  From the day the development team stood up in E City, we've included end users in the development team in a real way.  It was easy when the user was based in Elizabeth City.  We also found a way to make it work with the real end users were in Mobile.  One of the keys to making an agile process work is having user involvement at the beginning of a sprint to define the work and agree on the acceptance criteria, then having a user as one of the folks who affirms the done-ness at the end.  We've done that.  Your respect for the users of the system and your persistence in keeping them involved will be one of the keys to the team's further success.  Keep asking the users great questions

2. We had the guts to pilot a prototype when the whole release wasn't done.  It's easy to not share something with the customer because it's not perfect yet.  We had planned to start the pilot while we finished Release 3.  But when it came to actually sharing our work and letting Mobile use it, we met our natural internal resistance to shipping.  There was no perfect time to start the pilot. We wanted to tool to be done enough that we weren't wasting the unit's time.  You should take pride in the respect you all have for the end user's time.  But we'll never get it right without sharing our work and letting users use it in an operational environment while it's still fixable.  I think we timed the start of the Release 3 pilot well.  Even though we didn't have the MDL done, we had enough done to start getting feedback.  It turns out we needed that feedback to find out that some of the things we'd planned to push to release 4 were "must haves" for release 3 from the user's perspective.  I encourage you to be equally thoughtful in future releases about getting feedback on early prototypes, while it's still early enough for people to tell us that we need some help.  

3.  We have tapped into the experience, lessons, and knowledge of the Air Force and the Marines using cMRO and EBS.  Anyone can learn lessons by making their own mistakes or looking back on their own successes.  But that's expensive and takes a long time.  It's much better to do it the way you've done it:  by reaching out directly to the GCSS-MC PMO to learn their lessons on what worked and didn't work with implementation, and by reaching out directly to the USAF developers with firsthand knowledge of the requirements that drove many of the functional changes in cMRO RUP 5 so it could support military aircraft.  Those relationships will bear fruit over the long term as you continue to cultivate them.

4.  We delivered solid, useful, context sensitive help with the UPK.  Based on feedback from the folks in Mobile and lessons from the Marines, we're making it even better.  This level of support will help show our future users that we care about them and want CG-LIMS to be as easy-to-use a tool as we can possibly make it.  I know there's some inherent complexity in EBS and cMRO -- "Complex" is the product's first name after all -- but I think users will appreciate the investment we're making in them by delivering good online context-sensitive help.  If we can deliver on a promise of being one click away from an lifeline from anyplace in the application, it will help turn them into fans of our work.

5.  We serve as a living example of a team doing its best to configure and deliver a COTS system using an Agile methodology.  In addition to all the CG-LIMS stakeholders who participate in our Agile reviews, folks are joining in from all over the Coast Guard to experience what we are actually doing so they can steal what works and apply it to their projects.  You should take pride in your willingness to share so openly and transparently.  You're making a bigger difference that you'll ever know.  

6.  You shipped Release 1 less than six months from the we completed source selection for the software.  (Remember the software award we made on December 22nd, busting the myth that nothing gets done between Thanksgiving and Christmas?  That's another thing we did well.)  Shipping Release 1 in less than six months took incredible prioritization and discipline.  We showed how much working software we could deliver to production in a fixed amount of time.  That was an important first step.  You transformed CG-LIMS from a vision to working software.

Yes, I lied about stopping at five.

Now I invite you to take a moment to think about -- and share as a comment --  something you think we've done well as a team.

I thank you for the privilege of serving you as the CG-LIMS PM.  I wish you all continued success as you continue to deliver, one small step at a time.

(crossposted at