Eric Nehrlich, Unrepentant Generalist

Charlie asked the question last week: “Why aren’t you striving to be a leader in your field?” which has gotten me thinking about leadership, and what it means to be a leader. It also sparked an email exchange with a friend on the topic, which led to some interesting thoughts.

What does it mean to be a leader?

• Is it being an “active participant in professional societies, write popular blogs about your industry, get asked to write articles for magazines and regularly speak on conference panels”, as Charlie suggests?
• Is it being the person on top of the org chart giving orders?
• Is it about knowing more than others?
• Is it being the person who everybody gets along with and goes to with their questions and problems?
• Is it being the person motivating others to achieve more?
• Is it about being impressive?

We can probably think of people in all of these categories who we think of as leaders. So it’s clear that “leadership” is not an easily definable characteristic. Yet it’s like obscenity – we know it when we see it.

Perhaps leadership is about helping people achieve their goals. In other words, if I want to be a leader, I must gain followers, and therefore I must do something that would get people to follow me. I can do that in a number of different ways:

• I can be an industry spokesperson, with the potential to widely publicize followers who add value, possibly turning them into leaders themselves.
• I can be the organized one, who puts together plans, prioritizes goals, makes sure resources are available, etc. so that my followers efficiently use their time and effort.
• I can be the domain expert, with the experience to understand how to turn ideas into reality and the ability to enhance others’ capabilities by providing them with the knowledge they need to succeed.
• I can be the consensus builder, able to bridge different viewpoints and synthesize them into solutions that are better than any individual contribution.
• I can be the inspirational one, able to convince people to reach deep inside themselves to work harder towards a common goal.

One way to measure leadership might be to see who everybody in the room looks to when a decision needs to be made. Just because somebody is the manager on the org chart doesn’t necessarily make them that sort of leader (as Rands’s great story about “The Culture Chart” illustrates). One can be “The Guy” to whom others look using any of the methods described above.

So what can one do to become a leader? Part of being a leader is understanding one’s own strengths and weaknesses and choosing a leadership style that matches one’s tendencies. Gerald Weinberg’s book Becoming a Technical Leader offers advice about finding one’s own path towards leadership. And, on the flip side, it’s hard to take somebody seriously as a leader when they are acting in a way that is contrary to their nature – the engineer trying to schmooze his way to the top, or the MBA spouting half-understood technical jargon.

Mostly I’m fascinated by this idea of leadership that initially seems so prosaically obvious – we all know what leadership is – and yet so difficult to define.

What do you think defines a leader?

P.S. There were a couple crazy months there, both at work and in life. Things have calmed down a bit, and I enjoyed a slothful few weeks to recover from the craziness, but it may be time to pick up the posting habit again.

Anybody that’s been following my Twitter feed knows I’ve been working long hours recently. I’m actually working harder now than I was last year when I was working full time while finishing my master’s degree at Columbia. This would come as a surprise to, well, pretty much anybody that’s ever worked with me, given my tendency to do just enough to get by and no more. So what’s different about working at Google for me? My current answer to that question requires a detour through some other things I’ve been thinking about.

I really liked Po Bronson’s book What Should I Do With My Life? when I first read it (his original Fast Company article is a nice introduction). Trying to find a place for myself in the professional world has been an ongoing struggle, as while I have the capability to do anything, I have often found myself in situations where I was stuck doing the “wrong things”. Bronson recently published an update article with his perspective since publishing the book, and a couple quotes from that resonated with me:

“All jobs have things you hate about them. But real people feel fulfilled by the overall purpose of their organization that the shitty parts are worth putting up with. It’s not what you do, it’s what you’re working towards. …you know the feeling you desire — fulfillment, connection, responsibility, and some excitement.”

I think these last four characteristics he describes are absolutely key for me. I don’t define myself by the specific work tasks that I have to do, as I have the flexibility to do any number of tasks. I aspire to define myself with the meta-work of being a generalist. I want to feel that I am getting to use my unique potential and abilities in the course of doing my job. I need to find some sort of meaning in what I’m doing, or at least be working towards both personal and company goals.

As Bronson observes, there are annoying parts of every job – the question is whether the goal towards which one is working is worth the annoying parts. To put in a larger sense, it is important to know the answer to the question of “Why am I doing this?” In a similar vein, when I have friends considering grad school, I argue strenuously against it and try to dissuade them from going. This is not because I don’t value education – it’s because grad school is really hard, and the only way to get through it is to know exactly why you are going. My arguing against grad school is a way for me to get them to articulate their core reason for going to grad school.

To get back to my original question of what makes my job at Google different than other jobs I’ve had in the past is that I can see how I am contributing to making the company work better. I do genuinely believe that Google is making the world a better place, all things considered, as it provides us all with tools that are astonishing in their ability to put information at our fingertips. I can see a future for myself where I get to use all of my skills and talents in that goal. The only obstacle between me and that future is myself – I have the opportunity, the skills, the support to get there.

So while the long hours are irksome, they are not morale-destroying – we were joking this week about who in our group would be the first to snap, and I was surprised to realize that I wasn’t anywhere near snapping. I have hit my limits before and know what that feels like. But in this case, while I am tired and occasionally cranky, I feel like the work I am doing is recognized as meaningful, both to the company and to my future, and that’s far more sustainable for me than working even half the hours on dead-end tasks like technical support.

This has tremendous implications for managers. In a free agent world, getting the top people is no longer about paying them the most (beyond a certain point, I don’t think it makes a difference) or showering them with perks – it’s about giving them challenging, meaningful, interesting work. Managers need to find ways to engage their employees by framing the work that needs to be done in a narrative that propels the employee forward into a desired future. Getting back to Bronson, managers need to work with their employees to answer the question of “Why am I doing this?” And the answer doesn’t have to be existential – it can be as simple as earning the money to support one’s family. But there needs to be an answer, because without that answer in place, the annoying parts of the job will wear anybody down.

Anyway. We’ll see how I feel in another month if things don’t slow down, but for now, it was interesting to think about how this job has given me the belief that I can finally stretch myself in the directions I want to go with my career and life, in terms of building on my interests in interdisciplinary collaboration. It’s not a Great Cause ™, but the belief that I have found a good fit for my skills and talents is exciting enough to me to keep me going through the long hours. That being said, it sure is nice to take a weekend off and actually have time to write down some of my thoughts

I’ve mentioned this idea in several conversations recently, so I figured it was time to blog about it. In particular, I’ve been telling people about my career and how it’s much easier to be right than it is to make the right thing happen. So, like any good wanna-be management consultant, I came up with a two-by-two matrix to illustrate the possibilities.

In case it’s difficult to read, the horizontal axis goes from Wrong to Right, and the vertical axis goes from Ineffective to Effective. I’ve labelled each of the quadrants with an appropriate name.

• Wrong and Ineffective: Fools are wrong about what needs to get done, but fortunately, they are also ineffective so at least they’re not moving the organization in that wrong direction. They’re annoying to have around as they waste other people’s time, but are more obstacles than active hindrances.
• Wrong and Effective: Players are effective at getting what they want even when it’s the wrong things. They’re the ones that play the political game successfully and get resources for their projects despite the project being a waste of time. These kill the motivation of others in the organization, as Players get rewarded for doing the wrong thing.
• Right and Ineffective: Martyrs like to say “I told you so”, as they have a sense of what the right thing to do is, but are completely ineffective at actually convincing others in the organization to do that right thing.
• Right and Effective: Leaders are clearly what you want – people who both can figure out what the right thing to do is, and can also mobilize others to their point of view and create action within the organization. Hard to find these people, of course.

Most engineers end up in the Martyr quadrant, as they have the powers of analysis to figure out the way things should work. But they don’t necessarily have the people skills and observational skills to understand how decisions get made within the organization. So they complain vociferously when decisions get made that they don’t understand, but are singularly ineffective at changing the decisions. I spent several years as a Martyr at a couple different organizations, so I’m very familiar with the feeling.

The engineers are living along one dimension in this space – they see Right and Wrong, and see that they are more Right than the Player, so they can’t understand why the wrong decisions keep on getting made. That’s why I think it’s helpful to introduce this second axis of Effective and Ineffective, to illustrate the axis on which they are failing. I’m still learning to be more effective within an organization (it’s a slow process), but just being aware of my failings is a good first step.

Of course, “Right” can also be optimized along multiple dimensions, and which dimension you choose to optimize on changes the decision as well. Engineers prefer to optimize for the correct technical solution (scalable, clean design, etc.) while marketers might choose to optimize for what the customer wants, sales people commissions, and executives revenue or profit. But that makes my two-by-two matrix too complicated.

Anyway, I have found this distinction of Right/Wrong from Effective/Ineffective to be a useful conversational prop recently, so I figured I would share it in case it’s helpful to others.

Google has a program called Self-Powered Commuting, where they let employees track the days on which they get to work via self-powered methods (primarily biking or walking). At the end of the year, they tote up the number of days, and donate a proportionate amount to charity.

What’s amazing to me is how effective this program is at making me feel guilty when I don’t bike to work. I have the tracking page set up to open as one of my initial tabs, and every morning I don’t bike to work, I feel a slight twinge when I have to close the tab without clicking the button. And when I look back at the month, I can see which days I missed and wonder what my excuse was for not biking. The simple act of checking in once a day and tracking how I did is remarkably effective at reinforcing the habit of biking that I want to inculcate in myself.

In a similar vein, I recently found Joe’s Goals, which is a website based on the same concept. It allows you to set goals for yourself, assign them different point values, and track them on a daily basis. For instance, I give myself 1 point for flossing and eating 4 fruits a day, but 3 points for going to the gym or posting here. And now I can see tangible progress towards acquiring these habits – I’ve flossed every day for a week!

It’s silly that such a small thing as checking a box once a day can reinforce habits that I have been saying I should acquire for years. But it does. By reducing my goals down to a daily yes-or-no question, there’s no fudging, there’s no “I’ll get to it tomorrow” – there is only whether the box gets checked or not. I read recently that new behaviors take a month to set into place as habits (Google Answers has a couple references), so if I can keep up these habits for a month, then I can perhaps add a couple more next month. We’ll see.

But I think tracking has applicability beyond personal improvement, specifically in management. One of my favorite management anecdotes is the story of how Andrew Carnegie (I think?) was asked to figure out how to make an underperforming factory more productive. He said all he needed was a piece of chalk. He arrived at the shift change between the night crew and the day crew, asked the outgoing crew how many widgets they had produced, wrote that number on the wall with the chalk, and waited. The incoming crew asked what that number was. He told them. At the end of their shift, the number was erased and replaced with a higher number. The night crew put up a higher number, and soon the factory was outperforming other factories. The story may be apocryphal, but I find it indicative of the power of tracking (and also competition in this case).

Many management texts recommend SMART goals – that’s Specific, Measurable, Achievable, Relevant, and Time-Bound. And those are good, although the method biases action towards easily measurable goals, which may not be aligned with overall organization goals. But the Self-Powered Commuting and Joe’s Goals pages show me that having regular checkpoints is really important as well, so that one can’t be procrastinating on those goals. I suspect that some of the benefit of Agile methods is having a daily stand-up meeting, so everybody knows they have to make progress each day. My best manager checked in with me once a week, to find out what I’d done the previous week, and what I planned to do the next week.

What are the behaviors you want to inculcate in yourself and others? Would tracking those behaviors, either privately or publicly, help? It’s a theory, and we’ll see how the Skinner-ian self-management experiment goes for me. If I’m still sticking with it in a week, I’ll move the Joe’s Goals tracker to my sidebar .

I really liked Tim O’Reilly’s post today about how companies like Google and WalMart are incorporating IT into their organizational DNA. O’Reilly’s post describes how those example companies are mapping out a new way of organizing people built around integrating IT into how the organization functions:

Sensing, processing, and responding (based on pre-built models of what matters, “the database of expectations,” so to speak) is arguably the hallmark of living things. We’re now starting to build computers that work the same way. And we’re building enterprises around this new kind of sense-and-respond computing infrastructure. …It’s essential to recognize that each of these systems is a hybrid human-machine system, in which human actions are part of the computational loop.

I particularly like O’Reilly’s description of the organization as a group mind that incorporates both people and machines, as it fits in with my thoughts on organizational cognition. The organization also incorporates culture, processes and many other feedback loops that structure how the organization accomplishes its tasks.

Let’s start by taking a quick look at two existing organizational models:

• Small teams – the pre-industrial-age organizational model. In a small team, no organizational structure is needed because everybody knows what everybody else does, and decisions can be made organically or by consensus. New team members are indoctrinated into the way things work by social pressures. Whether discussing hunter-gatherer bands or artisan guilds, it’s rare that organizations grew to more than 30 people without splitting into smaller groups. There’s a reason that even modern managers understand the power of small targeted teams. Communications limited the size to which a team could grow, as the number of communication pathways grows exponentially with the size of the team.
• Hierarchy – the industrial-age organizational model. Information and decisions are funneled up to the appropriate decision-maker, and the resulting decision is distributed out to the employees who carry out those decisions. This was ideal in a world of limited communications, as each employee knew that information flowed up to their manager, and decisions flowed down from their manager, so they only had one primary communication link to maintain. Hierarchies also simplified assimilation of new people because the hierarchy defined each employee’s responsibilities, generally in an organizational handbook.

There have been various hybrid organizational models where there are hierarchies of teams and other configurations, but teams and hierarchies have been the basic building blocks for most organizations.

We are in a fascinating time where the number of possible organizational solutions has gotten much larger, as technology has removed the communication limitations that previously eliminated many potential configurations. We are just now figuring out what the new possibilities are, evaluating their strengths and weaknesses, so that we can find the appropriate option for a given venture. To put it in geekier terms, we are starting to map out the vastly expanded search space for organizational structures.

I think O’Reilly’s post identifies one direction, where organizations integrate computers so that certain decisions (like Google ranking web pages) don’t need to be handled by people and instead information deluges are handled by software. One of my interests is in trying to map out other possibilities, what they would look like and how they would fit various organizational purposes. My previous post about the future of organizations discussed how the new limitations may be social rather than technical, which implies that we need to start designing new social structures that can take advantage of the newly available technology.

One possibility that I’m playing with is that of overlapping teams with clearly defined roles. The good teams I’ve been on involved people who trusted and respected each others’ contributions to the team’s overall goals. I’d like to think that a fractal organization could be built off of such teams which each have a team goal, and then each team trusts the other teams to accomplish their goals in order to satisfy the organization goals. There would be a ton of communication necessary to distribute information within the organization to where it needs to go, but I think that is becoming more realistic by the day.

Another possibility is the free agent world, where there are no continuing organizations. Instead, coalitions of individuals are formed for specific projects, accomplish those project by bringing in other people as needed, and then disband to pursue other projects with different people. This would be the endpoint of the world where everybody becomes a consultant in their specialty.

I’m sure there are lots of other possibilities that I haven’t considered. For instance, I’m definitely interested in what we can learn from how World of Warcraft guilds are organized to accomplish their goals when every player is free to leave guilds that don’t work for them. Or how organizations mobilize volunteers to work for them – I’m sure there’s much to learn from Obama’s campaign this year. I’d love to hear of other ideas that people have on how to organize people.

P.S. I finally created the Google group/email list to discuss organizations that I mentioned in that future of organizations post, so go ahead and join up if you’re interested.

Paul Graham’s latest essay claims that small organizations are the future:

“But in the late twentieth century something changed. It turned out that economies of scale were not the only force at work. Particularly in technology, the increase in speed one could get from smaller groups started to trump the advantages of size. …For the future, the trend to bet on seems to be networks of small, autonomous groups whose performance is measured individually. And the societies that win will be the ones with the least impedance.”

This is interesting to me because I’ve been thinking about organizational cognition recently, which is the question of how an organization creates a group mind that knows more than its individual constituents. If the trend is towards smaller organizations, then perhaps the problem isn’t how to get large organizations to operate more effectively, but instead how to facilitate cooperation between organizations. These are similar problems, but existing organizational solutions like hierarchies don’t work for inter-organization collaboration, which creates urgency to find more flexible solutions.

This move towards less formal organizations to accomplish tasks is also covered well in Clay Shirky’s Here Comes Everybody. Shirky cites Ronald Coase’s theory that companies exist because the transaction costs associated with organizing people were more expensive than the associated inefficiencies of not necessarily finding the best person for each individual task. According to Shirky, new technologies lower the Coasean floor and create the possibility of impromptu evanescent gatherings of people accomplishing things together that could simply not have been organized previously.

So how small can organizations get? Are we approaching a full free market world where we recruit different people for each individual project (the analogy I use in that post is movie making)? I don’t think so. And here’s why.

My theory is that the new Coasean floor is going to be set by social trust. While we are in a world where I could hire a programmer to do a task from Elance or oDesk, I have to admit that I would be very nervous about doing so for any critical task. Why? Because I wouldn’t know the person and wouldn’t trust them.

It takes time and experience together to build the trust necessary for a team to function effectively and efficiently. Teams do not begin jelling as high performance units until each member of the team trusts the others to the point where he or she feels comfortable outsourcing parts of their intelligence to them. In other words, even though we have the physical technology now to collaborate informally and spontaneously, we do not have the social technology yet to fully exploit those capabilities (which, now that I re-read that post, reminds me that I need to get back to that topic at some point).

So where does the social trust Coasean floor lie? Katzenbach and Smith suggest that the highest performing teams have between 6 and 15 people – the lower bound is set by not having enough variety of skills within the group to really create a group mind, the upper bound by communication inefficiencies. That range sounds right to me as well, based on my own experience with various teams at various companies. To really get an answer, we’d have to map out the performance curve of groups as they grow; in other words, 2 people working on a project together might get less done than those 2 people working independently because of the communication overhead, but they might be more effective because they can bounce ideas off of each other. How that scales up to 3, 4, 5, or 10 people depends on the people, and the organization, and the communication technologies in place. But I would guess that the sweet spot is in the 6-10 person range.

If that is the team size which is most efficient from a social trust perspective, we return to the original question I posed above: how do we facilitate communication and collaboration between such small teams? What are the social and physical technologies we can use to transfer knowledge and expertise so that teams can build off of each others’ work? I don’t know what the answers to these questions are yet. Some people would suggest semantic knowledge management technologies to parse knowledge and distribute it automatically to the right people. Others would suggest quantitative approaches where measuring for the desired results will spur appropriate action. I tend towards humanistic approaches where trained generalists build the bridges between such teams, but I’m slightly biased as that’s one of the roles towards which I strive.

I think these social technology design questions have the potential to created fantastic productivity benefits over the coming decades. We’re hitting the limits of what physical technology can do for us. We have more and more powerful computers that sit idle most of the time, as users stare at them trying to figure out the interface. No amount of technology seems to remove the need for meetings to synchronize the organization. And we’re at a fascinating time when the physical technology Coasean floor has been removed, opening up new experimental possibilities for social technologies to help solve these organizational questions. I plan to continue exploring this topic, and hope that you will join me.

P.S. To be specific, a few of us from Convergence08 are starting a regular get-together where we exchange ideas on the topic of how organizations think and work, and share articles and resources via email; in fact, this post was inspired by discussion from that list. Let me know if you’d be interested in joining us.

In my post about Philip Zimbardo’s work, I mentioned the concepts of situational vs. dispositional tendencies. One might see these as being obscure cognitive constructs. However, a recent situation made me realize that beliefs about these tendencies have direct consequences on management styles. So let’s dig into this some more by starting with a description of the two tendencies before getting into consequences for management.

Dispositionists believe that our tendencies and behaviors are fixed because of the kind of person that we fundamentally are. They search for explanations in a person’s character to explain their behavior. So when a criminal steals, they try to find the “flaw” in that person’s character that would make them perform such an act. This explanatory search can concentrate on nature (looking for genetic correlations) or nurture (looking at the childhood surroundings), but the search is for a character trait within the person that explains their actions.

Situationists like Zimbardo think that people’s behaviors, while influenced by their parents and upbringing, are dominated by the situation in which they act. The Stanford Prison Experiment is the most glaring example of this, where intelligent, well-adjusted college students turned into abusive guards and mentally unbalanced prisoners within three days of being placed in a prison environment. Because of these overwhelmingly powerful situational effects, if somebody performs “evil” actions, it is not necessarily an indication of a fundamental character flaw on their part; instead the situation must be examined to see how it contributed to the actions.

When comparing the two, the dispositional viewpoint is easier to understand, with a simple narrative to explain somebody’s actions (“Lucifer is a bad person, which is why he did bad things”). The tricky thing here is that saying something is something raises warnings flags for me (see my review of Wilson’s Quantum Psychology for a longer take on the difficulties of “is”-ness). Attributing a characteristic as a fundamental component of something, as is implies, simplifies the narrative, but at the cost of making us more vulnerable to the true complexities of life (I’m reading Taleb’s The Black Swan right now which expands upon this idea). The dispositional viewpoint also has dangerous consequences in how we raise kids: treating intelligence and talent as fundamental characteristics of children actually retards their development, as they don’t even try to improve themselves. I think that while we have dispositional tendencies, we need to recognize the situation defines how we behave. But I’m going to stop with the discussion of the tendencies themselves (since others have done it better), and focus on the managerial consequences of these two ways of thinking about people.

In a dispositional workplace, life is relatively simple – you interview candidates, find the ones that have the right fundamental attributes (e.g. “Smart and Gets Things Done”), and then focus on removing obstacles to progress so that these people get things done, in accordance with their nature. It’s a nice, tidy view of the world. Unfortunately, I think it’s too simple, as it ignores the influence of the system on the attributes that people display – people that are tremendously successful and effective under one system might be completely ineffectual and unmotivated in another system where their strengths go unused, as sports teams find out each year in free agency.

Situational management is about designing the system to match people with the appropriate environment to get the desired organizational results. This system design can take a couple forms:

• Designing a financial incentive system that rewards appropriate behavior, although Robert Austin cautions us as to the difficulties with this.
• Desiging a culture and vision that reinforce the desired employee characteristics towards a common goal, such that employees “believe in the mission they are trying to accomplish and know that they are contributing to its success”, as a former CEO of Southwest Airlines puts it.
• Understanding employees’ strengths and weaknesses and giving them jobs that leverage their strengths and minimize their weaknesses. The canonical example of how not to do it is promoting a great software engineer into management, since the manager mindset is completely different than the engineer mindset. Good management in this scenario is about placing people in situations that maximize their chances for success while contributing to organizational goals.

What’s I like about this conception of management is that it means that management is a design position. Management in a dispositional world is about hiring the right people and then getting out of their way – it’s passive and uninvolved. Management in a situational world is an iterative systems design problem with constraints – managers have to pick a vision, align employees with that vision, work towards the vision, re-evaluate progress, possibly pick a different vision that aligns better with the strengths of the employees, etc. It’s an ongoing active process that involves being involved with all aspects of the business, understanding how employees work best, what the organization’s capabilities are, monitoring the environment outside of the organization to understand how to align potential outputs with environmental demand, etc, and using that understanding to better design how the company works. This is the type of manager I aspire to be someday.

Over the past seven weeks (good golly, where does the time go?) at Google, I’ve noticed a funny habit of mine. Whenever I overhear a conversation involving something that is related to my team’s work, I drop whatever I’m working on and wander over to listen in. Now, one might guess this is due to my slacker ways and desire to gossip as much as possible, and one would not be entirely incorrect. However, I was listening to one such conversation earlier this week as two of my groupmates were trying to suss out exactly how an analysis should work, and realized that there’s more going on here.

Such conversations are where an organization’s thought processes are made visible (audible?). In other words, if an organization could be perceived as a group mind, then those conversations are the equivalent of watching the synapses between neurons firing. It’s the only way to get insight into how the group mind operates. This may sound a little wacky, but I’m inspired by Edwin Hutchins’s book Cognition in the Wild, where he extracts observations about cognition by watching how a group mind in the form of a navigation team operated. In the best case, meetings can be a reflection of this organizational cognition, an aspect which Peter mentioned in the comments of my meetings post.

So by listening in on such conversations, one can map out how the organization operates. Which assumptions are taken for granted? Who are the stakeholders mentioned regularly as needing to be consulted? How are decisions made in such conversations – is it by consensus, persuasion, or hierarchy? These sorts of observations may not be directly relevant to one’s job, but it’s invaluable to an amateur anthropologist like myself in understanding the different forces that are at work within the organization.

Even better, conversations often arise when new organizational territory is being mapped out, either in the form of current assumptions being questioned, or in the development of a response to a new stimulus. When everything is running smoothly and there’s a defined process for how to do things, there is no need for conversation as everybody knows how to do their job. But when one reaches the limits of one’s understanding, then one has to consult another person, and such consultations are where somebody like me can see how the organization, in the form of its constituents, learns. To use the framework of Latour, conversations are where we can see the organizational Collective perform the Consultation process, where it grapples with an outside influence (what Latour calls “Perplexity”). Seeing the Collective go through a round of growing and learning is exactly what I mean by saying that conversations are a window into the cognitive process of the organization.

This is also exactly the sort of fuzzy stuff that I once would have scorned as a hard scientist and logic-driven engineer. These sorts of ephemeral observations about an organization are difficult to quantify and would not have even been on my radar ten years ago. And now they are the sorts of things that capture the value I bring to an organization, as my ability to attend meetings and listen to passing conversations and extract this sort of organizational knowledge is a testament to my ever-improving observational skills. I liked Rands’s description of it as the culture chart, as opposed to the formal organizational chart. The culture chart isn’t written down anywhere and is supremely fuzzy – it can only be intuited by reading between the lines of conversations.

So that’s my rationalization for why listening in on conversations when starting a new job is important. It’s the best way to understand how an organization operates: which assumptions are stated (and more importantly, not stated), which stakeholders matter and which can be ignored, etc. And, in case you’re wondering, all this listening to conversations is why I have to stay at work until 9pm a couple nights a week to catch up on my individual contributions. Once I am more efficient at my assigned tasks and up to speed on how Google works as an organization, I hope to get my hours down to a more reasonable number.

Ferran Adria (yes, that Ferran Adria) stopped by Google today while in town for his book tour promoting A Day at elBulli. He spent most of the hour talking about innovation and his approach for trying to invent a new culinary language within the universal language of cuisine, with pictures of some amazing nature-inspired desserts, apparently from the other new book, Natura, by Albert Adrià.

But then he asked a question which elicited a laugh from the Googler audience: “Are there any website experts out there?” His question, which he claims he’s been asking for years and not getting an answer, is why website designers don’t put a map as the first thing on their home page. After all, they should cater to the user and make it easy for the user to find what they are looking for. He forced his webpage designer to make a navigation page be the first page at the elBulli site, but wonders why everybody doesn’t do that.

This elicited several answers from the audience. My answer was that the goal of many websites is to get you to stick around on their site – if you can find what you need easily, then you leave their site. I later thought of the analogy to a Las Vegas casino, where the navigation is intentionally designed to make it hard for you to leave the gaming area. Another answer mentioned “That’s why Google exists”, saying that search replaces standard navigation, which makes the good point that a designed information architecture doesn’t necessarily match a user’s worldview and may make it harder rather than easier to find what he or she is looking for (reminding me of Clay Shirky’s point about user-driven classification systems). Another person suggested that most Googlers would probably agree with him, but unfortunately bad designers outnumber good designers.

But I still like my answer, because it gets to the heart of design – what is the purpose which you are trying to achieve with a design? Many websites are designed to have you click around as much as possible, to pile up ad impressions and the chance that you might click on an ad before leaving. Google, on the other hand, has the stated goal “to have users leave its website as quickly as possible”. Is one of these designs better or the “right” way? Adria might think the Google way is clearly the right way, but I believe that design can not be evaluated in an absolute sense – designs must be evaluated within the context of the desired goals.

I’ve often found this attitude helpful when navigating the world. Rather than cursing the stupidity of a design, I wonder why a design might have turned out in that way, and what alternative goals the designer might have been working towards. Making things cheap is a common one. Making companies more money is another. It’s rare that a design can be created in isolation and free of constraints (and some would argue that constraints make for better design), so it helps to understand the constraints, whether philosophical or economical, that guided a design.

I’ve also been finding it helpful to remember this goal-first attitude in my first month at Google. It’s very easy to start going down an infinitely deep hole of documentation or data analysis, and occasionally I have to shake myself and say “Okay, wait, what exactly am I trying to accomplish here, and is what I’m doing moving me towards that goal?” It’s particularly tempting at times to do what is easy rather than what is right (aka “If you have a hammer, everything looks like a nail.”).

Getting back to a recurring theme of mine, setting big goals makes it easier to make all other decisions; for instance, Adria is intent on continuing to push the boundaries of innovation in cuisine, so he closes elBulli for six months a year while he and his team experiment with new ideas. Could he make more money by staying open? Of course, but that’s not his goal.

No real deep thoughts on the subject, just that Adria’s confusion as to why people would design something that was clearly inferior in his eyes got me thinking, so I figured I’d use it as an excuse for a blog post.

P.S. Yes, this post was mostly to get to say that it was cool to see Ferran Adria in person. And to get something up – my brain’s been pretty fried from trying to absorb way too much information at work. Hopefully, once I get settled in, I’ll start posting more – as usual, I have several proto-posts in the “couple ideas simmering” stage, but haven’t had the brainpower to expand them into full posts. Maybe I should just start doing scatterbrain posts like my book review roundup posts, where I spent a paragraph on each idea. Hrm.

As I am learning the lay of the land at Google, I’ve been initiating one-on-one meetings with people around me so that I can learn what they do and gain some different perspectives on what my group produces. I wrote up my notes after my meeting with one coworker, and after he reviewed what I wrote, he noted that we’d covered most of the material I wrote up in the first fifteen minutes we’d talked. The other thirty-some minutes had been much less information-rich.

He extended this observation to claim that most meetings were like that – they started strong, but then lost steam as people became fatigued. Speakers become less clear in making their points, while listeners become less able to absorb information, so meetings tend to get more inefficient the longer they go. This makes sense to me; as anecdotal evidence, I certainly noticed that by the end of the two and a half hour classes we had at Columbia, I wasn’t processing as well as I did at the beginning.

What’s interesting about this claim to me is that it assumes that the purpose of meetings is to transfer information. That leads to the question of whether other types of meetings would have similar dynamics, or whether it’s possible that some meetings might get more productive as they got longer, rather than less.

In particular, I think any sort of brainstorming meeting takes time to get rolling. The participants need time to ensure everybody is making the same assumptions and to establish common patterns of interactions. It takes time for a group to achieve the state of “flow”, where everybody is moving towards a common purpose. Despite the initial dip in “productivity” while alignment is happening, the payback once everybody is synchronized is tremendous. I’ve written about the dynamics of group flow before, as well as the importance of removing discordant elements that are impediments to achieving flow. Having come up with one counterexample, I started wondering about the dynamics of other types of meetings.

Progress report meetings are a bad idea in general, as the information is generally better sent out via other methods so people can skim what isn’t important rather than making everybody sit through everything. Admittedly, a poorly written email is more confusing than an in-person meeting since presenters have the advantage of instant feedback, but that only highlights the importance of writing clearly. To run some pseudo-numbers, an hour-long meeting attended by ten people uses ten person-hours. Let’s say the presenter could write a clear email or slide deck to present their material, but it takes them four hours to write it instead of the one hour they take to present it unclearly. Let’s also say that the clearly written material now takes people ten minutes to absorb instead of an hour of clarifications. Now the total time spent by the group is 5.5 person-hours (4 hours by the presenter, 1.5 hours by the other nine group members), for a net gain of 4.5 person-hours. Obviously, those numbers are made up, but they illustrate the point that it is often worth sacrificing the time of one person to save the time of the group.

Another type of meeting might be a consensus meeting, where people are trying to agree on a course of action. I think this is similar to a brainstorming meeting in that it takes time for everybody to clarify assumptions and make sure everybody is talking about the same thing and understands the tradeoffs of various decision options. At the same time, it also has the perils of the information transfer meeting in that the longer the meeting goes, the lower everybody’s decision making ability gets due to fatigue. So there’s a need for such meetings to be well-managed, preferably with a clear agenda making the assumptions explicit. The moderator also needs to move the meeting along when it gets bogged down in details that are irrelevant to the decision at hand, so that the decision gets made while people are still relatively fresh.

The point about moderators reminds me that meetings need to be better managed in general. Potential meeting attendees should have a clear idea of what the point of a meeting is before getting there. In an ideal world, every meeting would have a clearly defined set of success criteria, so that potential participants would know what they should expect out of that meeting, and be able to decide whether that is worth their time before attending.

Another danger of meetings is meeting creep. As another coworker observed, any regular meeting that is successful eventually loses value precisely because of its success. If a meeting is delivering useful results, influential people start attending, which induces other people to put their own items on the agenda so as to reach those people, which causes those influencers to go attend other meetings, increasing the value of those meetings, which starts the cycle again. The dynamics of meetings do not just apply to individual meetings, but also to the rise and fall of certain meetings.

I have sometimes wondered if there should ever be regular meetings, or whether meetings should only be called on a one-off basis to address specific problems. Meetings often become entrenched and keep on happening because they have always happened, rather than having a clear agenda each time. While one former coworker extolled the benefits of having regular status meetings so that everybody could check in with each other, I tend to believe that’s best handled by informal mechanisms. Admittedly, if I have to impart information to everybody in my group, a meeting is more efficient than talking to everybody one on one, but still more efficient is an email with the caveat that I have to take the time to write clearly.

I don’t really have a point here, other than to present an incomplete taxonomy of meetings. I’ve been fortunate so far in being able to duck most meetings at Google, because nobody knows I am yet. Unfortunately, as a large company with lots of smart people who want to have their opinion heard, Google has a lot of meetings. In some ways, it reminds me of being at CERN or BaBar, which were particle physics collaborations involving lots of smart people who wanted to be heard, and involved continuous meetings; my advisor at CERN had literally 40 hours of meetings a week to attend, and could only do actual physics work in the evenings. So thinking about how to make meetings more efficient might be a useful topic when I start going to more meetings myself.