Tuesday, April 01, 2008

one down, who knows how many to go

Went through my first real CEO interview yesterday. While a few very early stage companies have approached me, this was my first full, meet-the-board CEO interview of an accomplished, 100+ person company. Very interesting day, since I can't imagine a harder challenge than trying to pick the right outsider to come in and lead an experienced, successful group to the next level. I'm sure there is a "101 Questions to Ask a Potential CEO" instruction manual somewhere, but I haven't seen it. In my opinion, it comes down to connecting with the current executive team, founders, and board of directors. Do you inspire trust and confidence in your ability to raise the performance of those around you, to face difficult challenges, and to add an effective voice to the decision making process.

The company I visited yesterday had , I thought, some particularly good questions around conflict resolution, style, and vision that were fun to answer and generated more questions on both sides. Like most high-performance technology companies, they were a little less organized than they wanted to be, but the result -- a rotating subset of the interview team -- seemed pretty effective and kept me on my toes. I may incorporate that as a strategy in the future when I'm back on the interviewer side of things!

And, no, I don't know whether it will go forward. I think I did a good job presenting an accurate picture of who I am, what my strengths and weaknesses are, where my passion and experience aligns with their needs, and where we aren't a good fit. Even if they feel I'm who they want, there is no guarantee I would take the job. But I am still interested and spending the day being tested by a really smart group was a blast.

Of course, I feel like I should take this opportunity to talk about Paul Graham's "You Weren't Meant to Have a Boss" post that has been generating a fair amount of blogging noise, both pro and con. I thought his post was particularly apropos, since I'm currently in the process of talking to large technology companies (>10,000 employees), medium companies (>100), small companies (>5), and startups (<5). They all have positives and minuses, although my biases are pretty clear since my last three jobs were at early stage startups. In fact, my last time working for a large company was Lockheed in 1994.

But that doesn't mean I think working for a large or medium company would be foolish. Larger companies have expertise in many areas -- people management, for example -- and resources rarely available to startups. With a young daughter, I might decide to trade off for a different risk-reward profile than I have in the past. I might view it as an opportunity to learn by working in a different environment.

However, no matter the scale of the organization, the key question is how hierarchy is used (or not used.) Hierarchy is often spoken of as a singular structure -- usually the org chart -- but is actually made up of at least two independent components: control and communication.

  • Control: Who can tell you what to do
  • Communication: How information flows between people in the organization
Now, when we think of traditional hierarchical organizations, these two functions are superimposed. You have a boss. He tells you what to do, gives you salary reviews, and fires you. Sometimes, as in matrix management structures, there are multiple hierarchies, so that the person who tells you what to do may be different from the person who reviews your performance.

But, these are not the only choices. Think about control for a moment. Forms of government are a reasonable approximation of management structures:
  • Monarchy: typical management hierarchy, with decider at the top
  • Anarchy: either nobody tells anyone what to do or everyone can tell anyone what to do
  • Republic: everyone selects a group to tell them what to do
  • Direct Democracy: everyone participates in every decision
Or, consider communication parallels:
  • Broadcast: one to many
  • Telephone: one to one, sometimes few to few
  • Blogging: many (well at least a few) to many
Organizations have the freedom to choose whichever structures make the most sense based on their size and requirements. More importantly, different groups and levels within an organization can choose different combinations. When many-to-many communication is failing because of team size -- how many emails can you process a day? -- either adopt a different communication strategy or create interfaces between teams. When you need all hands on deck to fight a fire, democracy might not be your best option, but once the fire is out and you need to innovate, bring on the anarchy!

Which brings us back to organizational size. Clearly, strategies that work for 3 or 5 -- direct democracy, many-to-many communication -- won't work for a company of 10,000. Many-to-many communication scales as O(n^2), so if everyone is trying to engage with everyone else they are going to spend all their time just reading email. Worse, if all 10,000 have to vote on every decision, the stress of knowing everything so that you could alway vote wisely would be crushing.

One response to this is a traditional hierarchy. However, it this hierarchy need not be fractal. If you have teams of 5 or 20 or 50 that operate smoothly with different structures than the company as a whole, that is fine. In fact, it's better than fine -- it's almost certainly what you want! A good example of this were how game teams were run at PCP&L. The team helped build the design, schedule, and budget and then was generally cut lose to exist as an insulated pocket within the company. For all intents and purposes, the game team was a little, entrepreneurial startup, except that you had the surrounding company for health care, payroll, etc. Wholly owned subsidiaries can operate the same way, with the employees within the sub virtually unaware of the structure and requirements of the parent company.

The question comes down to how flexible will the group I work with be allowed to be, because all projects have changing requirements throughout their life cycle. A small company or startup means flexibility is limited by funding and resources. In a large company, that flexibility is more limited by culture and habit. Some people might argue that a small company's limitations are better because "you control them", that it is better to be denied flexibility because you ran out of money rather than some pointy-haired boss' random decision. Maybe, but in my experience, not being able to execute due to a limitation sucks either way.

So, for me, choosing based on company size is the wrong metric. Instead, the question should be about how open your design space is, what opportunities you need, what impact you hope to have, and which constraints most impact these vision and goals. For some, that will mean a startup, for others a big company.

And that will be OK.


Avi said...

I think 'constraints' is a great metric to choose, but I think you know it cuts both ways.

_Some_ will argue that constraints are important, as they tend to keep you focused. A company with too much money and perceived time may wander -- take on projects or hire chaff that ultimately hurt their long-term goals. They may build in a culture of waste that's hard to eliminate.

However, I'm more of the view that living too close to the edge means too many opportunities for random setbacks to knock you out.

The most common constraint I've seen: founders don't raise enough money and don't hire enough do-ers vs. managers, counting on the perpetual death march to meet their goals.

The mark of good management is planning things such that you can achieve your vision with mere-human levels of effort.

It's the difference between maximizing returns, and maximizing success, IMO. And success is multi-dimensional, no matter what the VCs say.

Startups seem a lot like raising kids in that way. Raise a spoiled kid = bad. Raise a starving kid = cruel. Raise a kid who learns how to achieve his/her goals = just right.

cory ondrejka said...

Constraints are important -- as you well know :) -- hence the importance of using them as a metric rather than a second order metric like company size. Good VCs, in my experience, recognize the complex relationship between success and returns.

情趣用品 said...



A片,色情,成人,做愛,情色文學,A片下載,色情遊戲,色情影片,色情聊天室,情色電影,免費視訊,免費視訊聊天,免費視訊聊天室,一葉情貼圖片區,情色,情色視訊,免費成人影片,視訊交友,視訊聊天,視訊聊天室,言情小說,愛情小說,AIO,AV片,A漫,av dvd,聊天室,自拍,情色論壇,視訊美女,AV成人網,色情A片,SEX