Management Thoughts

I get asked about my management and leadership style often. It’s an important question, one that often gets short shrift in silicon valley where engineering execs are hired based on their apparent technical competence (and, when fired, it’s always for other reasons – exercises poor leadership, misses deadlines, or simply doesn’t get the job done).

I’ve had a long-standing interest in questions of motivation, morale, and teamwork: How can a manger inspire a team to a higher level of performance? What motivates people to "be extraordinary" when that is what's required to meet a significant challenge? Many books have been written on that subject and I discuss some influences on my approach to management and leadership below.

The Classics

My influences start with Peter F. Drucker, whose timeless principles still apply in today’s environment. Drucker especially emphasized the value of a company’s human capital and the need to maximize those assets through effective management. He was writing about such core concepts as metrics and management by objectives over fifty years ago. His foundation principles include fairness, and deep respect for everyone in an organization.

Drucker never experienced the tsunami of information today’s managers must cope with. Stephen R. Covey had a lot to say about instilling a discipline – one’s habits – into the practice of management, and how important it is to understand the context within which we work (especially to begin with the end in mind!). But what I like most about Covey is that he brings ethics and principle-centered leadership into the equation. After all, we spend most of our waking hours working; and if we can’t feel good about it other than bringing home a paycheck, then we lose balance in our lives. (There’s a lot of that going around these days.)

There’s a category of books about leadership and achieving high performance from the world of sports, music and other domains which recognizes superior achievement that I think is relevant to business. It’s somewhat ironic that ideas for how to instill principles of unselfishness, making sacrifices to achieve a higher goal, and reaching a mental state (often referred to as “getting into the zone”) in which the highest performance is possible seem to get more attention outside of the business world. But there may be parallels in how, for example, Bill Walsh instilled his “Standard of Performance” on a team widely regarded as one of the worst in sports, so that they became perennial champions. Maybe business leaders would benefit from learning how Michael Tilson Thomas inspires the players in the San Francisco Symphony to perform much better than they ever had under his predecessors, although it’s mostly the same musicians as before. And when Bill Russell describes how he got in a zone and outperformed his competitors to the tune of 11 NBA Championships, aren’t there potentially lessons to be learned from that?

Jim Collins and Level 5 Leadership

Jim Collins is another significant influence. I got a lot out of BHAG (big hairy audacious goals) and "try a lot of stuff and keep what works" in Built to Last; but what inspires me the most is his research on characteristics of an effective leader, summarized in this article in Harvard Business Review, Level 5 Leadership: The Triumph of Humility and Fierce Resolve. Collins identifies a paradoxical combination of personal attributes of the most successful executives: a deep personal humility (giving credit to others for successes, having a calm demeanor); and an intense personal will (unwavering resolve, utterly intolerant of mediocrity). You can think of this balance as the yin and yang of level 5 leadership:

My takeaway is: Get my ego out of the equation, and always set the bar high. The worst thing a leader can do to a team is expect too little. In my opinion, the right formula is this: Start by building a world-class team, communicate the vision, remove roadblocks, maintain high professional standards – and expect great things from the team.

Servant Leadership

Over 40 years ago Robert Greenleaf, founder of the Greenleaf Center for Servant Leadership, published some essays on what he coined servant leadership, a practical philosophy that replaces traditional autocratic leadership with a holistic, ethical approach. Here’s how Greenleaf defined servant leadership:

"The servant-leader is servant first… It begins with the natural feeling that one wants to serve, to serve first. Then conscious choice brings one to aspire to lead. That person is sharply different from one who is leader first, perhaps because of the need to assuage an unusual power drive or to acquire material possessions…The leader-first and the servant-first are two extreme types. Between them there are shadings and blends that are part of the infinite variety of human nature."

Servant leadership applies at the team or departmental level, too. Treating other departments as customers creates the right mind-set for macro-level servant leadership. When this is reciprocated by other departments, great things can happen.

Robert Greenleaf died in 1990, but servant leadership has experienced something of a revival of late, in part due to its association with agile software development methodologies as I’ll explain in greater detail below.

Theory U

C. Otto Scharmer, author of Theory U: Leading From the Future as It Emerges, is a senior lecturer at MIT and a consultant who has helped develop leadership programs for companies such a Google, HP, Daimler, PricewaterhouseCoopers and Fujitsu (as well as non-profits including World Wildlife Fund and African Public Health Leadership Initiative). In his book Scharmer synthesizes and distills ideas from management sciences thought leaders, psychologists, sociologists, captains of industry, and other thoughtful sources on the nature of thinking, social dynamics, motivation, and communication. Scharmer’s work was influenced by, among others, Rudolf Steiner (whom I have read extensively). Because of the depth and breadth of his work, Scharmer is not a particularly easy read but one is well rewarded for the effort invested.

Scharmer’s main thesis is that we have a “blind spot” in our understanding of leadership and transformational change. Scharmer calls it the “invisible dimension” of leadership, even though it is our source dimension, as shown in the figure below:

Scharmer describes a conversation he had with the CEO of a major financial services company. The CEO said that, after years of organizational learning projects and facilitating corporate change, the probability of success of a major project depends on the inner condition, the inner place from which they operate or the source from which all of their actions originate. How can that be improved?

We know a great deal about what leaders do and how they do it. But we know very little about the inner place, the source from which they operate. In professional sports this is an area that gets a great deal of attention. But not so much in the business world.

Think of this example: If we try to understand how Leonardo da Vinci creates a masterpiece such as the Mona Lisa, we can study it in comparison to all other paintings to see if we can apprehend its most salient qualities (the results, or the “what” of his work). Alternatively, we might hope to observe him at work while he’s painting (the process, or the “how”). But what about when he is staring at a blank canvas, just as he’s beginning to paint. What is the source of his art? There’s no question it’s an inner quality—a source dimension. What can we learn about that?

At its core, leadership is about shaping how individuals working as a group attend to a challenge or problem. This is key: Albert Einstein advises us “No problem can be solved from the same level of consciousness that created it.” Creating an environment wherein teams collectively achieve a higher level of consciousness that can be directed towards the challenge at hand is imperative for groups to reach their highest potential. We need help attaining the higher level of consciousness, as Einstein advises, to solve difficult problems. Theory U is about helping teams achieve a higher level of consciousness, which enables them to co-inspire each other to higher levels of performance.

What is Theory U? It’s a process that ultimately enables a team to co-evolve – grow as a group – to achieve the highest collective performance possible. It begins with how we communicate. Typical business conversations are conducted in “downloading” mode: habits of thought are re-confirmed; people talk nice, but generally stick to a predefined script (I knew you were going to say that). The level of communication needs to progress from downloading to factual, object-focused talking and then to the next level – empathetic listening that’s based on inquiry and deep listening. Ultimately that enables the progression through the U-Process, as depicted below:

When one considers the subtitle of the book – leading from the future as it emerges – it’s easy to imagine it as a new-age inspired manifesto rather than the thoughtful, practical source of leadership wisdom that it is. However, the concepts of co-inspiring, co-creating and co-evolving – when applied in an environment of a smart, well-motivated team – can represent the source of a true business advantage.

My Management Style

In describing my management style I am specifically discussing how I apply many of the principles above to the task of being an engineering executive.

Outside of the technical domain, my primary areas of focus are people; and process.

People

The first question of leadership is not what, or how, but who. Jim Collins says about his research into great leaders:

We expected that great leaders would start with vision and strategy. Instead, they attended to people first. They got the right people on the bus, moved the wrong people off, ushered the right people to the right seats – and then figured out where to drive it.

My mandate is to put the best team on the field that I can. Building a world-class team is necessary (but not sufficient) for building a world-class company. In fairness to investors, executives and the rest of the company, we set a high standard and ensure that everyone is performing up to that standard. If they can’t – or won’t – then it’s not a good fit for them and we will replace them with someone who can and will meet our Standard of Performance.

People are my first priority, always. Human capital is the company’s most valuable asset and should be dealt with accordingly – by demanding a high level of contribution, of course, but also by treating them respectfully, fairly and honestly. People won’t be motivated unless they are honored as professionals and human beings worthy of our highest respect. That helps foster the type of company culture that I want to work in, and one that other high performers gravitate towards as well. (One way to foster this is company-wide is to encourage your team to treat other departments as customers.)

But it’s called “work” for a reason, and my attitude towards the people on my team is to ensure we’re performing at as high a level as is reasonably possible. To this end, I focus on the “four C’s”: Communication; Commitment; Competition; and Customers.

Communication: It’s easy to say “communicate more” or “there’s never too much communication” but the truth is more complicated. If one person “communicates” an essential fact to another, but it’s buried on page 27 of a 40-page document and it hasn’t been highlighted or singled out, then effective communication probably hasn’t taken place. On the other hand, if someone is communicating a point of view and the listener is in “downloading” mode, then effective communication probably hasn’t taken place in this case as well either. The only measure of communication is its effectiveness. To communicate well is not just a style issue, it requires thought and effort. We always try to establish processes that support required communications, but in dynamic environments more is required than can be anticipated and defined by standard procedures.

Good communication is hard. Since we mostly work on teams and each team member has access to different information sources, one needs to consider context and other perspectives to determine what information should be communicated. The mode of communication is important as well: much if not most technical information should be communicated in writing, and email is the typical mode (although wikis, blogs and forum discussions may be more effective at times). Sometimes a phone call or face-to-face interactive discussion is the better option, especially if there’s unexpected information, a sensitive topic to discuss, or immediate feedback is required.

Many processes are established to further effective communications, including: status meetings, daily Scrum meetings, distribution lists for key events and updates, product backlog updates and transparent access to development status artifacts, use of and notification rules for tracking databases, performance reviews, and so forth.

A quick word about performance reviews: I’m not a big fan, but they’re mostly necessary – especially when they are tied to annual salary increases.  The main problem I have with reviews is the notion of providing feedback on an annual basis; that’s not enough of that kind of communication, and it comes too late. Instead of feedback I prefer to focus on feedforward: It’s more valuable to an employee to be told in advance what is required -- and in near real time how well the objectives are being met -- than to wait until year-end to find out what should have been done.

Commitment: In general, we make two types of commitments to each other: informal, routine commitments (such as “I’ll send you that report tomorrow”); and more critical, “Capital C” Commitments (such as “this will be released on July 21”). The distinction is important: As we work together, we rely on each other for various tasks, reports, deliverables, pieces of information or other workplace artifacts. Lowercase c commitments are subject to the usual unpredictable events and stresses of the work environment: Of course we intend to meet these commitments, and in most cases we do. But it’s not the end of the world if an unanticipated complication or higher priority interrupt inhibits our ability to meet that commitment (although we should always communicate a change in status if there’s the expectation of a deliverable).

“Capital C” Commitments are different. When such Commitments are made, it must also be made clear (to both parties) that this is critically important. There is a legitimate and urgent need to the business for such Commitments. When unanticipated complications arise or we get interrupted, we are expected to overcome the impact of such inevitable instances of Murphy’s Law. Failure is not an option. We make personal sacrifices, we call upon additional resources if possible (including our own reserves), we work nights and weekends, we do whatever it takes to meet such Commitments. We are measured by our ability to meet and honor such Commitments.

There’s another category of “Capital C” Commitment, and that is in honoring our understandings. In this case, the Commitment is not related to an event or tangible deliverable, but to how we work together, or how we treat each other. For example, I may Commit to directors or managers who report to me that I will not “go around” them in communicating with their staff except under urgent circumstances when they are not available, and if I do I will always give them a heads up. Many such examples could be made; those Commitments need to be taken seriously.

Competition: It’s a tough world, and we have competition: competition for customers; competition for limited venture capital; competition for the best and brightest people; and competition internally for budget dollars, support resources, and so forth. We don’t often deal with competition directly, but it’s a motivating force for much of what we do. Competition is behind why we have a limited budget, why we have a tough deadline, why we add a customer-requested feature late in the release cycle, and why we work so hard.

Competition presents a unique challenge to the development organization: achieving defensible product differentiation. What gets much of our attention in building products and services is the set of prioritized requirements -- referred to as the product backlog in many agile methodologies -- that represents the combined wisdom of customers, sales & marketing, industry analysts and thought leaders, and relevant standards. But great products typically boast of innovative concepts, breakthrough technology, or some form of engineering-conceived special sauce. Engineering needs to fully understand the competitive landscape from a product technology point of view, and should consistently challenge itself to innovate in a way that achieves a measurable and relevant benefit to the customer. That’s a tough standard, but it’s what we like about competition – it brings out the best in us.

Customers: Peter Drucker stated that the purpose of business is to create customers. I had to think about that for a while, but it’s a perfect way of expressing a company’s fundamental objective. And although we expect to do so with great products, positioning, sales, support, operations and financing, there’s one other thing that’s almost always required: good customer references. But why aim so low? The objective, in my opinion, should always be to have great customer references – outstanding references, rave references -- from customers whose trust and respect you’ve earned to the degree that they will go out of their way to tell others about it, emphatically and with conviction. The kind of reference that if a prospect hears, they will almost certainly become a customer. But such a reference can't be asked for, it must be earned.

Ironically, the customers from whom I've received the highest praise in the past were those where something went wrong, sometimes really, embarrassingly wrong. But that’s not what they seemed to remember: What they wanted to talk about – for years, in some cases – was the amount and level of support the customer got when things did go wrong. What we heard from these customers was that they expect things to go wrong from time to time, sometimes seriously so. What they don’t expect is the vendor to pull out all the stops in order to fix whatever’s wrong – in some cases, whether it was our product or not.

Communication; Commitment; Competition; and Customers. If it sounds like a formula, it’s not. People, especially technical professionals, are complex, multi-dimensional and sometimes capable of extraordinary feats. Software development is a group endeavor and therefore a social process. Smart people thrive in an environment where they are surrounded by other smart people with a common goal, where their contributions are honored, where they’re treated with respect, and where they can have fun while they’re accomplishing great things.

Process

Processes should be defined that expedite team efforts rather than getting in the way. There are many aspects to this, but here I focus on agile software development methodologies.

Agile is the term used to refer to a class of development methodologies that are incremental and iterative. Scrum is by far the most popular agile method; others include XP (eXtreme Programming); Unified Process (UP, Agile UP, or Rational UP); Evo; Feature-Driven Development (FDD); Test-Driven Development (TDD); Dynamic Systems Development Method (DSDM); Crystal; and others. There is no “best” approach other than what works for a given team; in many cases, a hybrid is used (for example, Scrum + XP is widely used).

I have used Scrum in multiple settings, and have been impressed by it’s positive impact – especially where requirements are uncertain. For larger-scale developments I would be tempted to blend in aspects of Agile UP since it incorporates more SDLC artifacts into the process. In either case I think it’s important to incorporate test-driven development (TDD) concepts into the agile process.

Why do organizations adopt agile development methods? Generally, it’s for improved visibility to the project status from stakeholder, better adaptability in a dynamic environment, greater business value sooner in the development process, and overall risk reduction according to VersionOne as shown in the chart below:

While these are great reasons, I would say the biggest benefit is that engineers love it compared to traditional approaches, primarily because they are in much greater control over what they do and how they do it. And that’s a Good Thing: There’s no question in my mind that the best outcomes result from a team of well-motivated engineers who feel in control over their environment. But another key reason engineers love agile is it enables them to be heroes in customers eyes, since customer prize responsiveness more than anything. And agile allows development teams to be much more responsive to customer needs.

With Scrum, developers are fundamentally empowered to do their jobs in a team that collaborates together and makes its own technical decisions. According to Ken Schwaber, one of the Scrum founders and author of Agile Software Development With Scrum in 2002, Scrum teams are “self-managing, self-organizing, and cross-functional” and therefore control their own destiny.

But isn’t that like anarchy? Not at all. The Team (Capital T for a Scrum Team) is self-directed, but works from a prioritized Product Backlog that represents the interests of all stakeholders, including executive management. Scrum calls for three roles that work collaboratively to ensure alignment with business objectives and transparency to all stakeholders: Product Owner; ScrumMaster; and Team (typically 6-9 people, but can vary). The Product Owner represents customers’ requirements as well as requirements of other business stakeholders, and determines what are the highest priority features for each sprint (called the Sprint Backlog). On any given sprint it’s the Product Owner’s responsibility to ensure that the team is working on the highest priority items at that time. This is sometimes referred to as “just in time” requirements. And because requirements are dealt with on a just-in-time basis, feature creep and requirements uncertainty are minimized.

The agile project manager for Scrum is referred to as the ScrumMaster. He or she is responsible for ensuring that the Scrum process is followed, and for removing roadblocks as identified by Scrum Team members. Because the nature of agile project management is one of facilitation rather than top-down control, the ideal characteristics of an agile project management role such as ScrumMaster are those of a servant leader. Servant leadership, described as an executive model for large companies in the latter decades of the previous century, has now emerged as the preferred management style for agile software development.

The Scrum process involves one or more time-boxed sprints to a release (typical sprints are 4 weeks or 30 days). Ideally, each sprint should result in potentially releasable code – functionally complete (for the backlog items included in the sprint), refactored if necessary, tested, and documented. Within each sprint, there are daily scrum meetings (sometimes referred to as “stand-ups” – meetings so short and focused that they can be done standing up). At the end of a sprint, a review meeting takes place which usually includes a demo to the Product Owner and any interested stakeholders: One of the tenets of the Agile Manifesto is “working software is the primary measure of progress”.

The following diagram from TargetProcess shows the overall flow:

A variety of Scrum-specific project management tools exist, including from companies such as VersionOne, Atlassian, and Agilebuddy. These tools generally include dashboards for “at a glance” project status as well as burndown charts, defect tracking, velocity trends, analytics and a variety of reports.

Scrum – or for that matter, any agile methodology – can’t be plugged in to an organizational environment as a cookbook formula. The amount of what Craig Larman defines as "ceremony" in Agile and Iterative Development -- what we typically think of as structure (development phases), deliverables (artifacts of the development process), and process (workflow and authorizations) -- are unique to each development, based on dozens of factors. This requires wisdom and judgment, and, like the software being developed, should evolve incrementally and iteratively in every organization.

Summary

There is no single “best” management style; what works for any executive is a function of his or her values, character, and inner condition. In this post I have outlined some of the practices and leadership themes that have worked for me in over 20 years of executive experience.