Up Front & Iterative Planning

You know the question…”what kind of projects are best suited for agile?”.  To me, this is the same as asking “what kind of projects are best suited for empowered teams, technical excellence, servant leadership, reducing waste?”.  Would there ever be a time when you would not want these things?  I know, that is a very smart-allec response, but I just can’t help myself.

I then ask “what’s the other option?”.  This surprisingly stumps people when I ask this.  We all know there is no such thing as “true” waterfall in software.  So, to help them along, I clarify what I “think” they are asking which is “what kinds of projects are best suited for big, upfront planning and design vs. iterative & incremental delivery?”.  People almost always agree with the restatement of the question.

I believe that everything can and must be done iteratively and incrementally in software.  However, the level of “up front” may change with the type of project.  Gutting out an old accounting system and replacing it with an Oracle or Great Plains solution will require more up-front planning and analysis than a green-field Web 2.0 social networking application.

So, fellow agilists, “up-front” is not a bad word, it has to be done.  We know this already with release planning, and even sprint planning, we just don’t call it “up-front”.  Just be careful…VERY careful.  You run the risk digressing into waterfall.  At times, there will need to be more “up-front” than other times.  Always be asking yourself, “what is the soonest we can deliver value to customer?”.

Sweet Exercise to Teach Empowerment and Self-Organization

I taught a one day agile course to a group of IS leaders. I wanted to come up with an exercise that illustrated that to be efficient, you have to reduce hand-offs, allow teams to self-organize, and leaders must be servants.

So, with my wife, we came up with “Project Pinwheel”. Below is how it works.
Supplies
  • Straws
  • Paper fasteners
  • Paper copies of the pinwheel pattern to cut out
  • Scissors
  • Markers
  • Paper punch
  • Shoe-box size plastic containers (to hold the supplies at each table)
Process
This is the process that I used. Feel free to change it to suite your needs.
At each table, there were 5-6 students. Each plastic shoe-box contained enough supplies for 20 pinwheels (would need less most likely). Also, make sure that you make samples of what the pinwheels are supposed to look like for each group.
There are two rounds. Below are the slides I showed to the class for each round.

Round 1

Your job is to create as many pinwheels as you can in 5 minutes. Take a minute, and assign the following roles at each table:

  • Cutter – owns the scissors and completes the cutting
  • Decorator/Designer – owns the markers and creates the design for the pinwheel
  • Hole Puncher & Paper Fastener – owns the hole puncher and the paper fasteners
  • Folder – does any necessary manipulation or folding of the paper during the creation of the pinwheel
  • Tester – tests the pinwheel when it has been finished. Verify that it has been decorated and that it at least moves a little when someone blows.
  • Manager – responsible for telling each team member what to do. The manager will communicate the tasks to the team members. The team members are not allowed to see the instructions.
  • At your table there is a box. Each box contains the instructions and the supplies.
  • No one is allowed to step outside their role.
  • The manager is the only one that can speak, by instructing the team members. Each team member is only allowed to speak to the manager.
  • If the pinwheel fails testing, the tester must hand the pinwheel back to who they think caused the “bug”.

At this point, the teams DO NOT see the sample. They don’t even know it exists.

Now, start the timer. After 5 minutes, they will likely create 0 pinwheels.
Round 2
[Slide for round 2]
  • Manager, you are now a servant leader. Please do whatever it takes to help the team.
  • Team members are allowed to help others out.
  • You can cross role boundaries.
  • Everyone can read the instructions. You can use the instructions as a guideline, but you can now be creative in how you create the pinwheels.
  • First, take 2 minutes to discuss how you will work together to be more efficient. Then, you will have 5 minutes to create as many pinwheels as possible.
Here, the team goes through a time-boxed planning session of 2 minutes to figure out how to best make the pinwheels before the 5 minute pin-wheel making session. Oh, I also pull out the sample pinwheel, so the team can have a collective understanding of what the “vision” is.
During my first session, each team made between 5-10 pinwheels. The ones who made less had issues with the “servant leader” thing, which made for a great discussion afterwords.
This is a bit of a pain in the butt to set up, but in the end, it is WELL worth it, and now I have the supplies for many classes to come!
Below is the instruction sheet that each team used.
Project Pinwheel Instructions
  1. Cut-out the pinwheel on the solid lines only.
  2. Decorate both sides of the paper pinwheel.

  3. Cut the dotted lines from the four corners to the center circle. Try not to cut into the center circle.
  4. Use the hole punch to poke a hole through the four tiny dark circles. Use the hole punch to poke a hole through the straw about 1/2 inch from the top.

  5. Hole punch the middle of the center circle on the paper pinwheel where the dotted lines converge.

  6. Make the holes on the four points meet at the center circle.

  7. Push the ends of the paper fastener through the holes on the pinwheel. Then push the fastener through the center circle.

  8. Place the straw on the back side of your pinwheel and push the ends of the fastener through the hole in the straw. Open-up the fastener by flattening the ends in opposite directions.

  9. Test the pinwheel to ensure it is functional. If the pinwheel is not functional, determine which step you need to send it back to in order to gain functionality. Repeat testing again until pinwheel is “complete”.

Only decorated, functional pinwheels that can at least minimally move when blown can be counted as “complete”.

Below is the image to cut-out to make the pin-wheel.  You can also find the image just by searching for “pinwheel pattern” on google.

The quest for the perfect agile tool

I’ve been searching off and on for the perfect agile tool for years. My “dream” tool would be a one stop shop for releases, stories, tasks, acceptance criteria, test cases, points, etc., etc. Oh, and by the way, it would be INTUITIVE.

When I enter a new team, I start the search over, and almost always end up back to stickies and/or spreadsheets. Currently I am using Acunote (http://www.acunote.com), and it seems to be doing the trick. It’s missing a lot of things, but it’s good enough for our purposes.

I’ve learned to not worry about tools until there is a solid foundation. The right team (technical lead, Product Owner, Scrum Master) and a product backlog must be in place first. Then, the core Scrum team can be formed because it is clear what needs to happen. Once the Scrum team is in place, ensure they are self-organizing. NOW, it’s okay to start looking at tools … as a team.

In the meantime, my quest will continue, and maybe someday that perfect agile tool will surface. But, if it doesn’t, I’ll just stick to stickies and spreadsheets and be happy!

What makes a successful waterfall project?

I recently posted a question on linked in wanting to hear from people who have either led or been a part of a successful waterfall project.

Here’s the link. There a some awesome answers. What’s interesting is that what is required for a successful waterfall project is clearly defined scope, engaged sponsors, and a great, empowered team. Sadly, I rarely see all of these occur in the wild. And, these are the primary elements of what’s necessary for an agile project, explaining why there is a higher success rate with agile projects.

The Curse of Traditional Software Development Methodologies

I’ve realized that very few teams follow the waterfall methodology by the book. However, I think that most shops try…really, really hard. Of course we WANT to define every requirement and scenario that could possibly happen. Once that’s complete, THEN we can start building. And, if there is a new scenario, or “missed” requirement (heaven forbid), then the developers can point at the requirement gatherers and say “AH – HA!!! You screwed up, and now you must suffer the consequences!”.

Well, this is just silly, we all know that change happens, but yet we try to fight it every step of the way. That’s what requirements change management is, right? Its to prevent change, i.e. just say NO! To remain competitive in today’s marketplace, we just can’t think like this anymore.

Below is a list of beliefs that we (including myself) really wish was true. If they were true, our jobs would be so easy! Of course, it wouldn’t be as fun 🙂

  • The customer knows exactly what they want.
  • We understand what the customer wants.
  • We understand who the customer is.
  • We can define everything up front.
  • The technical team knows how to build what the customer wants.
  • We can accurately predict how long it will take to develop what the customer wants.
  • There won’t be any problems (or very few).
  • Documentation is a true measure of progress.
  • Nothing should change. If it does, it’s [fill in the blank] fault.
  • Hours worked is infinitely directly proportional to value delivered.

Okay, so none of these things are really true. Since we all like to pretend, we have failed over, and over, and over again. I believe that this failure in traditional projects has caused stereotypes to evolve, including:

  • The technical team is arrogant and lazy.
  • The technical team doesn’t care about the customer.
  • The technical team is lying. It can’t be as hard as they say.
  • The customer doesn’t respect the amount of work necessary to deliver what they want.
  • The customer is stupid.

This failure has also caused some pretty weird behaviors to spring up.

  • A tendency to over-specialize and divide ourselves into silos.
  • The desire to control and direct every move made by every team member.
  • Forgetting that development is creative and innovative and can’t be treated as a commodity (this causes some real bad implementation).
  • A culture of command and control.
  • The desire to create a process for EVERYTHING in an attempt to control.
  • A fear of sharing “too much” information with other silos…I mean other teams. If we share “too much” information, then the other teams may want “change”.
  • A fear of sharing too much information with the customer. Again, for a fear of change.

Please, stop the madness. Let’s stop pretending that traditional development works and start working together.

Hard Questions for Agile

I am seeing a lot of the same questions come up over and over again recently. Now that larger organizations are starting to see the benefits of agile, more and more questions are coming up.

Everyone (including myself) has a desire to have a prescribed method, process, checklist, etc. that they can use to make sure they are doing things the “right way”. In Scrum, there is a pretty clear list of activities, artifacts, and roles. People struggle because these lists provide the “whats” but not the “hows”. For example, when I tell teams that they need to have one product owner for that project, they are sometimes at a loss; “How do we find one product owner? Our organization won’t support that!”.

Below is a list of questions that I have seen come up over and over again. I’m going to address them over time (as I have more time).

  1. How can agile work on large projects?
    • I know this is being addressed slowly, but there are just not enough case studies out there. The standard “Scrum of Scrums” just doesn’t cut it anymore, meaning that there are serious complexities that aren’t addressed in the standard definition.
  2. How do I work with a vendor that does not want to take and agile approach?
  3. How does interaction design work in an agile environment?
    • I have not met any interaction designers (HCI type people) who really buy into the concept of iterative, incremental development.
  4. Our company is committed to PMBOK, how can agile fit in this environment?
  5. Our company needs to make financial projections that are 5 years out. How can we do this using an agile approach without trying to plan everything up front?
    • We in agile know that it is a myth that you can accurately plan everything up front.
  6. Our company needs to be CMMI compliant to be able to work with government entities. How can we be CMMI level 1-5 compliant and still follow the agile values?
  7. We have a project that’s in crisis. How can agile rescue our project?

I’ll add more as I hear them, but these are the biggies. Feel free to provide input or links that addresses any of these questions.

Agile won’t work on [fill in the blank]

I have four kids ranging from 15 to 3. For some reason, people with no kids really like to give my wife and I advice on how to raise kids. Usually they will provide analogies using their favorite pet. Or, they will tell me about their second cousin twice removed who “just had a baby”, and here’s what they PLAN on doing. I just politely nod and say “thank you”.

So, what’s the point? Well, I tend to get a lot of advice, or “words of wisdom” regarding agile development and how it won’t work on certain projects. I’ve heard it won’t work on complex projects, big projects, geographically disperse projects, mission critical projects, non-software projects…you name it. Everyone who tells me this has either a) never worked on an agile project or b) worked with a poor implementation of agile.

Instead of politely nodding and saying “thank you for your advice” as when I’m given child-rearing advice, I’ve recently decided to challenge these ideas. Of course not in a confrontational way, but just in an inquisitive way. “What experiences have led you to that belief?”. “Why can’t an offshore project implement agile methodologies?”.

Below is a post that I saw on LinkedIn:

As with any project management tool, it is situational. If, like in many software development environments, there are one or two key people involved in each and every task, Agile tends to break down. In this case, there are better methodologies to employ. Like any concept, I think the rub is that no technique fits all situations.

This is clearly someone providing advice on agile development that clearly has never worked in an agile environment, or has suffered a poor implementation. I couldn’t resist responding. Below is my post.

There is a very broad spectrum of agile methodologies out there that will cover any situation. Being agile also means adapting the process to fit your needs as it’s an empirical, “inspect & adapt” approach to managing projects. For instance, typically Scrum & XP are implemented together. Scrum only addresses project management issues, while XP addresses development issues.

The failed agile implementations I have seen were caused by one or more of the following:

  1. The team is not empowered and self-organizing.
  2. Team members are working on too many projects, resulting in excessive context switching.
  3. Management is committed to command and control (related to #1)
  4. The customer or customer representative is not involved.
  5. Processes are not allowed to evolve.
  6. There is no dedicated product owner, i.e. someone responsible for the ROI that the team can go to (getting rid of the multiple boss syndrome).
  7. The ones leading the agile implementation have not been properly trained in agile.


So, agile breaks down when the organization is not ready to accept its principles. If the organization is not willing to change from command & control to servant leadership, empower the team, assign ONE product owner, and get the customer involved, then it will fail.

Hopefully I didn’t come off as a jerk on this. I just couldn’t resist.

If you feel that agile just won’t work on _____, I encourage you to find out for yourself. There are plenty of cases where it has worked on huge, complex, mission critical projects.

Struggles of the New “Agile” Consultant

In the Beginning, There Was Chaos

I’ve always wanted to be someone who mentored and taught others how to live up to their potential. Naturally, I wanted to become a consultant. The path that lead me to where I am will have to be the subject of another post. For now, I’ll just summarize.

After college, I went in to retail. What else could you do with a Psychology degree? After about four years in retail management, I went back to college to get a Computer Science degree. I had various jobs where I performed the functions of systems engineer, software engineer, DBA, SEO, etc. Then, I graduated to development management where I managed a couple of teams at a large software company. Surprisingly, managing a software development team isn’t that much different that managing a retail team.

While managing these teams, it was hard to ignore the fact that they were in total chaos. My company was pushing a home-grown agile software development methodology. It sounded good because there was more focus on the customer, but it just didn’t work. So, I went on the journey of finding what agile development meant. Certainly it didn’t mean “no documentation” and “no design”, right? Well, it didn’t, but that’s how we were acting. The teams were working super long hours. The engineers had no idea how to estimate. The team really didn’t know what was supposed to be delivered.

Let There Be Scrum!

Then, slowly through several “iterations”, we implemented scrum. I played the Scrum Master role. The path we took could be mapped by books that were used. The books that paved the road for us were “Agile Estimating and Planning”, by Mike Cohn, then “User Stories Applied”, by Mike Cohn, and then finally “Agile Software Development with Scrum”, by Ken Schwaber. As we went through each book, we implemented these new things we learned incrementally.

So, we did it. We successfully implemented scrum. It was awesome. Why was it awesome? Here are some of the key reasons why:

  • Freedom – The team was free to do whatever it took to make software. The management team didn’t care if we used Scrum, XP, or Waterfall, they just wanted software. We were continually getting beat up for the crap we turned out. We finally got tired of it, and management let us do whatever it took.
  • Engaged Teams – I was blessed with a wonderful team. Was every single person on the team engaged and wonderful? NO. But, these things have a way of working themselves out. The team, with a little guidance, quickly became a high performing, self-organizing team.
  • Awesome Product Owner – Our product owner on our team was amazing. She was strong, had technical sense, and strived to play by the Scrum rules. She mastered the product backlog. Sure, maybe at times she was a bit “too” engaged throughout the sprint, which caused minor distractions at times, but that was to be expected and easily corrected.
  • Individuals and Interactions Over Processes and Tools – You probably recognize that this is part of the agile manifesto. I listed this explicitly because this was a big deal. Originally, we were trying to use all kinds of tools; QuickBase, TeamTrack, TestDirector, Excel, etc. It was becoming a mess. We were spending so much time making sure everything was up to date. In addition, we were spending time generating reports that no one was reading. In the end, we used the tool with the lowest possible impact to the team; the white board. The teams sprint backlog was documented on the white board, where we tracked the daily status.

    Okay, to be honest, we didn’t “just” use the white board. We also used a tool called “ExtremePlanner” to hold the entire product backlog, and help with sprint and release planning. This was basically the playground for the product owner, so there was no overhead experienced by the team.

Don’t get me wrong, the road wasn’t always smooth. There was a lot of history that we had to work through. There was a pile of crap that we called software that was released to the public, and we had to fix it. But, we did.

Enter the Agile Consultant

I really wanted to use my experiences over the last couple of years, both good and bad, to help others make the transition to common sense, i.e. Scrum. I know that most software development shops really didn’t know how to make software that the customer wants without smashing the deadline, scope, or quality. That’s exactly where we were.

Our company closed our office in Omaha, and I didn’t want to move to the home base in San Diego, CA. This was the perfect opportunity to step into an “agile” consultant role. By certain divine intervention, a local company was looking for a consultant who was an expert in agile software methodologies. That was my calling.

My first job as a consultant is working with a company that was (and still is) in trouble. Some promises were made by executives that are now long gone, and the current team has to make good on those promises. Below is a sample of some of the core problems at this division:

  1. Legacy software – Most companies who have been around for 5+ years have to deal with legacy software that is fragile and extremely time consuming to make any changes. This is just about the worst case that I’ve seen or even read about. The code is so poorly architected it is a significant challenge to make even the smallest of changes.
  2. Turnover – Approximately 95% of the original team left. Most of the people who left held knowledge that was not documented nor was it shared. Just about everyone working on this project is new (within 6 months).
  3. Poor development infrastructure – The team is using VSS, which performs file-locks when code is checked out. This is a serious bottle-neck. Due to the poor design of the code, there are only a few files that most programmers have to touch. In addition, there is no consistent build or test environments or process.
  4. Lack of core experience – There are some key personnel, such as BA, QA, Management, that have no experience in those disciplines. Not only do they have to learn the product, they have to learn how to do their job.
  5. Micro-Management – There is a desire to manage every last task the team has to do. This has had a very damaging effect on the team. Most of the team is paralyzed when they have to make any decision for themselves. In addition, the management style is militant and somewhat cold.

I knew that this would be a challenge going in. We changed our approach several times. Below are some things we tried.

  1. Implement Scrum on a small project and use the success to convince others to adopt. We would play roles on the Scrum team, product owner and Scrum Master.
  2. Educate everyone on Scrum, then let them try it themselves on any product they decided, with us helping and coaching only. We wouldn’t play any roles on the scrum teams.
  3. Implement Scrum on the flag-ship product, and then let the other smaller projects fall in line. We would stay very close and coach the new Scrum Master, product owner, and team.

We actually did all of these things in order, and are now on #3. Scrum is well underway.

We went through quite a few struggles that we weren’t expecting.

  • We assumed that everyone would welcome us with open arms and that there would
    be a high level of trust. Boy, was I wrong. Most people were quite resistant. Many have come around after seeing some of the small successes, but there are some that I see that will always be resistant.
  • We underestimated the learning curve of Scrum. Yes, Scrum is simple. But, as Ken Schwaber (one of the original creators of Scrum) frequently states, “Its incredibly difficult to implement”. For example, creating a list of prioritized work has proven to be incredibly difficult. The product owner believed (for a while) that everything had the same priority … HIGH! We are planning on conducting additional training sessions to accommodate.
  • We didn’t know how much trouble the project was really in. As we all know, not all projects can be saved. It doesn’t matter what process you use. Again, as I’ve heard Ken Schwaber say, with Scrum, you’ll know how screwed you are after 30 days. This couldn’t be more true.
  • There is no trust amongst the team. QA doesn’t trust development, development doesn’t trust QA, BA doesn’t trust anyone, etc. What has been incredibly helpful here is our daily scrums. Just some time for the entire team to get together every day has helped the team members come closer to a trust relationship.
  • People don’t understand what self-organizing and cross-functional mean. The concept that you would have people from different disciplines on the same team is crazy! What’s more crazy is that you would allow the team to work on what they want to. When we introduced this concept, it was like we just told them they should believe in elves. I was shocked. To me, these concepts are common sense for any good manager. As Scrum is working, and we are coaching everyone along the way, these concepts aren’t approached with dismay and confusion anymore. It will be a slow process, but I think that the team will come around.

I’ve learned a lot while implementing Scrum at my first gig as an agile consultant. It is very different being in the role of consultant vs. full-time employee. At this client, we are not empowered enough to affect change as much as we would like. I think we will successfully implement Scrum, and eventually the client will be happy. We’ll come out other end with some battle wounds, but it will be well worth it. Will the project succeed? I don’t know. After the first sprint, the velocity isn’t what executive management was expecting. Actually, not even close. But, its honest and realistic. Something they have never had.

Improving Scrum Roll-Out

I know that there can’t be a fine-grained plan to implement agile development on every team. Every team will have unique needs that will have to be addressed. However, our overall approach will be different.

  • We will spend a significant amount of time with the management team, getting to know their problems, their personalities, and where they want to be. Each one of them will have different perspectives that will eventually have to be addressed. We need to set expectations better than we have in the past.
  • We will work at building trust and teaching Scrum and agile development from the top-down AND bottom-up. With our current client, we used a bottom-up approach, and it just didn’t work. However, if we would have only pushed it from the top, the team members would have totally rebelled. We will have to carefully map out our approach at our next engagement.
  • We will spend significant time educating the team on agile development and Scrum. This will have to be continual. We did a “light” introduction, and then dove right in. The team really needed more than this.
  • We will make sure to not get sucked into putting out fires, which is what happened at our current client. Now, this is a tricky one. Most likely there will be fires that they will need help with. If so, we will have to introduce a separate “fire-fighting” team. While the process improvement team continues with improving the process and coaching the members of the Scrum team.
  • We will need to know exactly what level of empowerment we have and/or need. While management may say that we can do whatever it takes, we will have to make sure that everything is communicated clearly and frequently. And, we will get push-back, as we’ve learned at this current client. We will now be more prepared.

We have learned a ton. Implementing Scrum as a consultant is much different than implementing Scrum at a company that you work for full time. It was very naive of me to think there would be no difference.

I will add posts as this project progresses, and as I learn new things about Scrum. I apologize for the length of this post. I promise future posts won’t be like this. I do hope that this has helped someone out there who is starting their career as an agile consultant.