11/21/2005

You say you want a revolution ...

Revolution and evolution. Destination and directions. Do not confuse your need to have a company or team vision with how they get there.

I always say that my goals are revolutionary and I make every attempt to have my methods be evolutionary. Sometimes even little common sense things can take a long time. When it comes to change, trust is key. When I took over the Solaris (TM) team at Sun, no one tracked progress of the team's efforts. I resisted instituting such a process until I had fought some battles on their behalf. Even then I posed it as a question to my staff "How can we know if our teams are working consistently and successfully?"

You need to think big and bold in order to ignite your team's imagination but you must translate those thoughts into practical tasks. Both thoughts and tasks are important. Did you ever try to give a driver step-by-step directions without them knowing the destination. It is unsettling and disconcerting for the driver. The same is true for your developers. If you cannot articulate your team's destination and your company's destination, they will be unsettled, less motivated, and more prone to leave.

Sometimes change is necessary to reach your destination. The ability to change is important to any organization. Brokering change is difficult. It takes time and patience. Most of the time you need to continue to produce product while you make change. It is akin, as one of my previous managers used to say, to figuring out how to re-tool the engine of an airplane mid-flight and still land on time. Do not underestimate what developers look at as change. From my understanding, before GE took on Six Sigma (TM) they developed a process to help adopt change because they knew they could not make the Six-Sigma-discovered changes without such a mechanism.

Think big and plan thoughtfully!

More later ...

11/10/2005

Schedules, Schedules, Schedules

Why create schedules? How do you make schedules? How do you get your team buy-in to schedules?

Schedules are about predictability. This may be obvious to you but get lost on the way to the troops. It is always important to talk about rationale. You need to be able to predict product availability so your sales team can sell and your executive team can predict revenues for budgeting and investor purposes. You need to be able to predict integrations points between constituents on your team like QA or docs and development. Your developers need to predict when they need to work harder or longer and when they are done and can go home for the weekend.

Schedules are not rocket science but they do require effort. First you need agreement on what you are building and therefore you need both requirements documents and preliminary engineering specifications and you need them to be consistent. Next you to do project decomposition and there are a lot of books on this subject (just google software project decomposition). Next you must define integration points. Some hints:

  • Only schedule 4 days a week
  • Update the schedule weekly and do not let slips go unattended
  • Don't let people go home if they do not finish their tasks
  • Make people sequentialize their tasks even if it takes longer. Mounds of partially completed tasks will result in slips
  • Allocate enough time for docs, qa, integration and field test feedback

Many engineers will tell you that software is not schedulable. They are wrong but it something that can only be believed when experienced. Get some people who have done it before as part of your team. Make your first release smaller and eminently accomplishable to help get your team experience. Goals have to be top down but schedules have to be bottoms up --- do not dictate dates.

Persevere.

More later ...

11/08/2005

What do you do when you make a mistake?

Many of us grew our careers through the engineering ranks. Engineering is more black and white than management. Either your code works or it does not. When an engineer becomes a manager that kind of thinking has to change. Everyday managers are faced with things that do not work like people leaving their teams, irate or lost customers, bugs in their products, etc.

Managers cannot afford to be paralyzed by setbacks. They need to obtain the mentality of a baseball player that loses a game one day and goes out with energy to win the next day. The most successful teams have winning percentages of 50% and the most successful ballplayers have hitting averages of around 30%. Managers have to learn how to sleep well after a failure and get up the next with energy and win.

Some of the aforementioned setbacks will be due to manager mistakes. These may be the hardest to overcome. I have personally done things like shorten development cycles or said the wrong thing at the wrong time resulting in missed schedules or unhappy peers and employees. The question you need to ask yourself is what do you when you make a mistake?

Here are some things that work for me:

  • Get up the next day with energy
  • Do not be embarrassed
  • Take responsibility for your actions
  • Acknowledge the mistake before you are told about it by others
  • Look at the problems ensuing from the mistake as just things to be solved
  • Put a plan together
  • Work harder and longer to make up for your actions
  • Work with your team to deal with the problem
  • Be open minded
  • Learn from your mistake

These are not easy things to do but they can make all of the difference in your success as a manager.

More later ...

10/27/2005

How do you make progress on your roadmap in the face of firefighting?

This is an ongoing question from the moment that you have a product and even a prospective customer. Real-time requirements for assistance, demos, features and bugfixing are common diversions for many engineering groups. So how do you handle this natural competition for resources?

The first thing you need is a roadmap. Without clear plans and goals you won't even know what you are not doing. Time can slip by easily in fire-fighting mode and along with it market opportunities.

The next thing you need is a budget. You need to legitimize the time spent on diversions from the roadmap. Do not expect your developers to handle a full development load along with a second development load for bugfixing. Another key to successfully managing resource competition is to firewall resources for new development (see Christensen's 1997 book The Innovator's Dilemma).

The next thing you need is some process. Decent schedules which represent all of the activities engineering is responsible for is a must. If new opportunities arise, do not assume the team can just lump more work on. I like schedules to be aggressive but accomplishable -- where the key word is accomplishable. Put a process in place that allows for change. If a new feature is needed, will you fund it with new resources, re-target resources working on some of projects, etc. ? Outside influence can begin to look like a fire-fight. Changing directions too often will be inefficient and depress developers.

Finally you need leadership. Understand your team's capacity. Understand your company's financial and competitive situation. Aggressive but accomplishable.

More later ...

10/24/2005

Live to work or work to live?

This is a question that many people ask implicitly or explicitly. It is a value that changes during a lifetime. Teenagers might feel work minimally to live only to become twenty-something work-aholics to become burned-out forty-somethings trying to reconnect with their families. The lucky among us find balance. But let's go back to the question, what should you do? How should you think about work?

I have the pleasure of being the dad to a wonderful teenage girl who is starting to be interested in the inevitability of work. She wants to earn enough to do stuff. She doesn't feel like she needs to live my lifestyle (yet:-). She is getting pressure from her school to think about careers. So I will lean on what I tell her as I write this entry.

I tell her she has to do two things: find something she enjoys doing a lot and be passionate about and find something that will pay the bills for the lifestyle she chooses.

In reading the first item, I think the keyword is passion. When I started working a colleague of mine said "If you are not willing to put in more than eight hours in a job, why are you willing to put in eight?" The upshot of his statement was that when we work we invest our most alert and most valuable hours of our day and we should use those hours doing something we love and are passionate about. I agree with him.

If you only work to meet your financial needs, you will be miserable. Sometimes we have to do things we would otherwise choose not to do. You need to understand this aspect of your life. However, I suspect that most of us can figure out how to do something we are passionate about and still pay the bills.

When I hire people, I look for people who have that passion. I do not want someone just paying the bills or putting in time.

So how do I answer this for myself? I live to be passionate and excited about all aspects of my life and today that includes work.

More later ...

10/19/2005

Why write a book?

Today I formally launched my book 100 Questions to Ask Your Software Organization. The official press release went out. The book is available for sale. But why write a book? What was I trying to accomplish?

I had been writing some form of the book in my head for 9 years. When I sat down to write the book, it popped out of me. I wanted to write a practical, easy-to-read, scalable guide for software managers and executives who have software organizations reporting to them.

As I walked into organizations as a full time employee or as a consultant, I was applying some of the same methods time and time again. I wanted a guide to remind myself about those methods. I wanted a concise mechanism to help communicate with the team about my methods and my intent. I wanted a basis for discussion about better software management.

I talked to a lot of people along the way. One of those people was a venture capitalist who said he needed a better way to understand and evaluate software organizations in his portfolio. With that comment, I adopted the question format. I have been told by readers that it helps the momentum of the book. The questions lead the reader in the right direction and the words that follow strive to help the reader understand what is important to consider as they develop answers for themselves.

There is a crisis in this industry. Late and poor quality software are the norm and not the exception. Disgruntled and burnt-out employees are almost expected. Customers assume they cannot trust vendors based on real historical evidence.

We can do better. It is not rocket science.

Realistic budgets and roadmaps. Clear roles and responsibilities. Effective and efficient processes.

Leadership.

With this book I challenge you all to start a new dialogue. To begin anew. To be better.

More later ...

10/07/2005

What does "Done" mean?

According to Webster, done is the past participle of do. According to Webster, do's definition includes produce, perform, execute, commit and carry-out. The problem with these definitions is that they do not speak to the agreement required by multiple parties as to the contextual semantics of the state of being done. In other words, when do all parties agree a product is produced or a task carried-out? This is a topic worth exploring because it causes angst, misunderstandings, delays in products, missing product quality, etc.

I recall a start-up I worked in a number of years back that had a fairly large release to accomplish. The release included new hardware and new software from the firmware through GUI. I was brought in to run the software group and more explicitly to execute. We had all of the typical problems: top-down dates, too much work for the number of resources, poorly defined features, unmade critical decisions. I worked on all of these with the team. I helped force clear and complete decisions. I helped set expectations about what really could be accomplished and divided the work into required and second tier. I helped the team make real schedules. When we got to the code complete milestone, my engineers said they were complete but we were not complete. Why did that happen?

I made the mistake that everyone shared the same concept of the word "done". To me, done, was, well, done! My engineers and even some of my managers thought that close was good enough. Of course this caused all hell to break loose. The QA folks could not test, the product management folks screamed publicly, and the engineers felt put-upon.

So now I take the time at team meeting to express my definition of done. When a project starts, I look each developer in the eye and ask them if they stand by their schedules. I require specs that can be compared with requirements documents so there can be no misunderstandings. I require developers write reproducible unit tests that can verify tasks do what they say they do before check-in.

Were my engineers bad or unique? No to both of those questions. Many engineers are not taught the meaning of the word "done" or worse yet, they are castigated for it. This is the new millennium and there are some basics when it comes to development. Take the time to agree with your team as to when they are done ...

More later ...