In Defense of Big Companies
I wrote this to the q-combinator email list on 6/14/2012. It is pretty rough and unedited.
The big company vs small company control vs innovation comment is interesting. I'm currently working for Disney Parks and Resorts Online as a Product Manager Intern. (A bit about me since I just joined the group after Jacob Cole referred me. I'm a senior at MIT in Management and Comp Sci. In high school I had a Facebook app with 1.3 million installs. I’ve worked at Deutsche Bahn, the quarter million employee German national railway; State Street bank; Altman Vilandrie, a management consultancy for large telecom firms; and NextJump, a 150 employee “start up”) More details: http://wiki.theplaz.com/Resume)
I’ve worked at 4 big companies, 1 medium sized one, and 1 small one. I am not 100% sure why I seem to be drawn to big companies.
Since I did a start-up in high school, I didn't feel the need to do one in school. I didn't want to become to attached to one; I wanted to finish school, probably because I wanted to do things the "normal" way. Plus at school, I chose to focus on classes after not doing so hot and knowing that there will always be ideas for a start up, but only 8 semesters of school. I also wanted to try something different (a different industry, size company, etc) every time period.
Also I think big companies provide big platforms to build on; like what I did at Deutsche Bahn http://wiki.theplaz.com/Touch%26Travel They also provide big puzzles to solve. I've always been more interested in the people/business applications of technology (I'm not interested in inventing the next ZIP algorithm). And I think office politics are a challenge, not irksome. I think companies can provide built-in support networks. They've learned a lot and can teach you a lot - especially for internships. They provide a stable base - you're not worried about your salary, raising money, where you will live, etc. (I am a very plan ahead type person.... I need to research fully, plan out the most efficient way to do something, and then execute a plan. You might be interested in my post how I approach school http://wiki.theplaz.com/Working_out_a_System)
But as a Product manager intern at Disney I still think I have discretion. No one is checking in where I use my time. Now I am in a special situation. I report directly to the VP who oversees ~200 people and ~$120 million in projects. I've been assigned tasks to do and as long as I make suitable progress on them people don't care what you are doing every minute.
On the flip side, I am in the office for pretty much the same times every day. I'm paid hourly which is a real puzzle to me - my first time doing that for an internship. But being in the office makes you accessible for meetings and provides some structure. It also prevents work from taking over your life.
Perhaps this mirrors the company structure. At Disney there is no possibility of a big IPO payoff. Disney makes a steady return (and experiences steady growth) day in day out. We know we’re here for the long haul.
I don't think traditional management techniques are worthless. You also need some idea of planning. If you are interfacing with outside parties you need to be more scheduled. One of our projects has hundreds of people involved. So many things need to align that you just have to plan. If we are going to roll out physical infrastructure, train people, commit to a release day, etc you need to know when stuff gets finished! To know that, we need to understand what resources you have, how fast people are working, etc.
Any big business works by giving responsibility for a small section of the business to individuals/small teams. Each group owns a part of the billion dollar business. Often these parts all have to work together. For example, if our ~10 person ride engineering team at Disneyland stopped doing their job, the whole park would grind to a halt. It would block billions of dollars of revenue and make national news. But inside the team, it is fairly free to organize how it wants. Sometimes things are different: a team has incremental improvements to the business. For example, Disney Weddings. It adds some revenue by attracting new guests who would not normally come. But they work with the other groups, like hotels, to rent conference space. They compete with other wedding locations. Entire lines of business are like this. Disney Cruise Lines leverages the Disney brand – building on our characters and reputation for quality, but they can implement programs on their own to react to the cruise market. The division builds on synergy, but must make a profit themselves at the end of the day.
Often times management comes from the need to present a career progression. People want to know that they can advance and not just do the same job their entire lives. Start ups are not thinking about this yet.
Plus, we're a public company so we need to provide a clear ROI. (Google is moving more towards this, and Facebook will probably soon). These estimations are often wildly estimated. But they should still be helpful tools. I worked in management consulting where we used some of these tools. They break large problems into smaller problems where it is easier to guess and make assumptions. You do have to be vigilant to avoid confirmation bias. Now like at universities, project money funds more than just the project - it keeps people employed here for decades which builds an institutional memory.
Our parks have lots of opportunities for investing cash. The management needs to compare all of the opportunities and pick which ones look best. As a large public company we can deploy a little over a billion dollars if we have a good idea. A start up first has to convince a VC to give them cash and even then no one will give you a billion dollars all at once. A certain amount of work needs to go into showing where that billion dollars will be spent and that it will impact revenue by more than the project costs.
It has to work. Start ups are inherently risky and can take some time to show a profit, but then must profit big. Disney investors, on the other hand, know that we will provide a stable return. It’s not that risky to invest in Disney versus a start up. This leads to pressure to show short term returns and could prevent ideas that eventually turn out to be big – but this is what the investors want.
ROI planning is also a way to pick which project among many. The executives who have all the information are the best able, as opposed to each employee. Sometimes getting a big project done is more important that random exploration. (Since moving to “agile” teams can self-organize to work on their own sections – but this isn’t really the exploration talked about in the Valve article). Again, completing a project could be a better choice than taking a risky bet on R&D – based on the goals of a company.
I like the idea of 20% time at Google. Could that work here?. The question here is what will make a business impact? 20% time worked at Google because they are sought new revenue however they can. For example, one of their biggest ideas was extending Ad Sense from Google results pages to other sites. We grow through fairly standard means; while we do grow with new ideas (cruise ships, guided tours, etc) those ideas have to big enough to purse, so in some respects they are fairly obvious. It's not cheap to churn out a new Disney Parks and Resorts product. The marginal cost of a website is near zero. Bricks and steel are expensive. Perhaps it’s a good idea for the Consumer Products division who is always churning out new $10-15 products. Perhaps I just can’t see what ideas *could* come out of 20% time at WDPRO.
Then there is also the question of consistency. Should the site look the same between teams? That requires planning and collaboration.
Things may fall apart when you benchmark them. A new mobile site might look like a success if it has a 10% ROI, but what if you passed up a project that would have been 15% ROI. That is bad.
Perhaps one of the things is communication. Because of my role I have easy access to the VPs. If I had an idea (which actually has not happened - there are tons of smart people here who have thought of everything - or perhaps I'm too much on the inside and can't think of ideas anymore - or I too quickly discard ideas because I know the challenges they face). Idea collection sites (which I worked on for IFC in MIT's 15.571 - highly recommended if you are interested in business IT) can work well in companies.
Then there is always the question of quality. Everything we put out has to be top quality. It's the same with Google. Anything Google announces will have hundreds of thousands of people using it that day. You also have to get into all of the scalability/optimization mess BEFORE you launch an idea. You can't really do a cheap prototype all the time. (But we do do them in some circumstances.) Also everything we do is instantly blogged about.
Also many times you can't change things midship after deploying stuff. When we deploy a new tool - it has to work. Google can't break Gmail - too many people are relying on it. It's the same with our attractions. Guests pay good money to come to the park. The last thing they want to hear after waiting in line for an hour is the ride broke. That just cannot happen. Preventing this from happening increases the cost of testing, planning, prepping. Perhaps it doubles the cost, if not more. This is one of the reasons military stuff is so expensive.
This is one of the challenges Facebook is facing. Their culture is to actively rejecting everything I talked about above. They want to maintain the scrappy web start up feel. But when you are used by millions of people you run into huge problems when you change the site overnight. There is also the complexity of having to translate, check, QA lots of different versions. For example, Google has to check Gmail in 100s of languages, internalization things (month-day or day-month?), plus regular vs Google Apps versions. It increases exponentially! You must really think through all this, plus the privacy and cultural implications when you are a big, complicated site. Like it or not, Facebook has a huge influence on the culture. They just don’t want to recognize it.
You also have to have at least some controls in place around cash and customer data. You can't have a disgruntled employee steal the companies' customer data or cash. For private data, including credit card numbers, you need practices in place to avoid that. At some point you are just too big a target that it is just too tempting, too large of a reward, and too big a risk to your business (falling from a higher ledge) to not put in place controls.
Now we have a bunch of job titles here. We have Content Strategists, Information Architects, and Copy Writers. These are separate titles. Now especially with agile, they all communicate as a team and anyone can speak up, but we still have separate roles. I asked some of the managers about this and they think having the specialized roles allows each person to focus and advocate for their area. It also gets more eyes working on a problem - which generally leads to a better outcome.
In order to divide up these resources, PMs need to plan. A good PM knows that a copy writer will be needed in 2 weeks. They submit their request and then their functional manager assigns them to a project - thus the project has the right people at the right time.
WDPRO is much more professionally managed than NextJump where I worked last summer. We are not bouncing from one emergency to another. But we work regular hours and make regular, predictable progress. Turnover is also much, much lower than at NextJump. Compared with NextJump WDPRO produces quality work more consistently.
For example, today is the Cars Launch and its business as usual around the office. All of the work was done way before by each department. The launch runs smoothly because it was well planned and well managed – each person doing their share. No drama. High quality.
This consistency is also important regarding the planning again. Sometimes a higher mean $$ cost is worth it because of the decrease in variance. Again planning/collaboration.
Now I think there should be more inter-project (and especially inter-team collaboration at Disney Parks and Resorts). It would help collaboration which would lead to better projects. Now at some firms (banks, insurance companies) preventing the cash/privacy risk is more important than creativity. When you are chugging along making a steady return, especially a business that has built up trust with your customers, maintaining that trust is more important. Other times (like at my job at Disney) the benefits of more integration out weighs the loss of innovation.
At Disney, it’s all about secrecy protection. Disney does not want plans of the new Shanghai park to leak, so Disney accepts less collaboration to keep the number of people clued in (and thus able to leak info small). The US Government faces this challenge with regards to the Bradly Manning leak. How much info should an analyst have? One of the reasons the government failed to see 9/11 was too much separate data. But then one guy can leak it all!
What is more important? An XX% improvement in terrorism prevention or an XX% improvement in the prevention of leaks. It varies by employee, environment, and your goals. There are points when you want innovation (like when you are starting top plan), but there are points where stuff needs to get done or need to keep things under wraps. This is an important strategy to think about and set.
Now why do I write about all this - it's kinda my job! In general, I really want to know WHY. I don't want to just accept the established ways. I want to figure out myself what is best. I don't just want to fall into the Silicon Valley stereotype. I want to know what the best way to work is. Why do people plan? How do people plan? Why do people do certain things? (Management majors are not all that useless after all – lol)
Plus, you never know I might end up with my own start up one day. I just want to know what I am doing first so I have a better chance of success.