Home > Articles > Security > General Security and Privacy

Software [In]security: Startup Lessons

  • Print
  • + Share This
Gary McGraw discusses the seven lessons he's learned through his startup years at Cigital.
Like this article? We recommend

Seven Things I've Learned About Running Companies

Interacting with academia is an important part of what I do as CTO of Cigital. Though I have been known to lecture at Stanford, CMU, Cornell, Harvard, NC State, Purdue, and a bunch of other places, I have a special place in my heart for the University of Virginia (where I studied Philosophy as an undergraduate) and Indiana University (where I earned a dual Ph.D. in computer science and cognitive science).

Alf Weaver, a CS professor at UVa recently asked me to lecture to his Electronic Commerce Technologies course. I was happy to oblige. When I asked what I should lecture about, I got back a one word answer — startups.

Not quite sure of what to do, I decided to draw on my own experience at Cigital. In 1995 when I joined Cigital, it was known as Reliable Software Technologies (or RST) and had a grand total of seven employees. I'm proud to say that today Cigital has over 120 employees and offices in Virginia, NY, Boston, Silicon Valley, India, and Amsterdam.

Helping Cigital evolve has been both hard work and a joy. Here is a list of seven lessons I've learned through my own startup years. For the record, my startup journey is not complete, and there have been many ups and downs, so take these lessons with plenty of salt.

Learn to Think and Write. Focus on Communication

Now that my friends' kids are reaching college age, I find myself in many conversations about universities, computer science, software security, and so on. I always offer the same advice regarding school choice — learning to think and write is much more important than choosing a major, especially in undergrad school.

Back in 1984 when I attended UVa, I majored in (the lucrative field of) Philosophy. If there's one thing that Philosophy courses will teach you, it's how to write…and write…and write. The ability to analyze a text, synthesize an argument, and then get the point across clearly lies at the very heart of Philosophy.

When you're working on a startup, communications skills are paramount. So learning to think and write is essential.

In the same vein, a great course can change your life. One of the best courses I have ever had was Paul Humphreys' course Computers, Minds, and Brains. The course was about Artificial Intelligence, and focused lots of attention on Searle. As an antidote, my friends and I brought in Hofstadter and Dennett's book The Mind's I. As my studies progressed, I found myself at Indiana University with Doug Hofstadter as my advisor.

Perhaps Doug's sparkly creativity and counter-culture approach to academia rubbed off on me. In any case, I did learn an important lesson about science and how fields progress — it's more important to be intellectually honest than to be popular among the scientific elite. This view set me up nicely to play the "little boy" in the Emperor's New Clothes as applied to computer security.

A decade into the field of software security, it's gratifying to see (and study) large-scale software security initiatives. The once-heretical notion of shifting our attention from "protecting broken stuff from bad people" to "building stuff that's not broken in the first place" now seems just like common sense. Thinking, writing, communicating, and painstakingly pointing out the obvious all continue to play important roles.

Build a Network (Expect to Travel the World and Repeat Yourself)

Being outgoing and gregarious certainly helps when it comes to starting up a company. Attend as many group functions as you can, from academic conferences to tradeshows. Reach out and don't be shy. Meet people. Most importantly, understand that meeting people means meeting them in person. Virtual communications may play a major role in the modern business world, but they are no substitute for having a glass of wine and a chat together.

A big part of business is getting to know people. One great way to do that is to attend conferences, especially as a speaker. You may be surprised to learn that the biggest computer security tradeshow, RSA, is not known for its content as much as for the fact that everybody goes. A majority of my colleagues spend much more time outside of the conference talks meeting with others than they do attending (or giving) talks; as do I.

To give you some idea of how much time all this yammering can take up, consider that I gave 35 talks in 2007, 35 in 2008, and 20 in 2009 (so far). Repeating yourself seems to help. I've given the Exploiting Online Games talk at least 50 times, and I've lost track of how many talks about Software Security I've done.

Doing lots of public speaking turns out to have plenty in common with teaching. Learning the sometimes-surprising ways that other people think about what you've just said helps to expand your understanding of your own subject. Instead of being boring, giving the same talk for the hundredth time can be exhilarating as you watch others gain an understanding of what you're saying.

If you want to do a startup that involves a new idea (which every good startup should), expect to do plenty of teaching. As a reward for the too-much-travel this involves, allow yourself the occasional boondoggle (maybe one per year). This year I plan to attend the OWASP conference in Brazil.

Follow the Categorical Imperative

In case you're not familiar with Kant, the categorical imperative boils down (at least in my view) to the Golden Rule — do unto others as you expect them to do to you. Business is not the art of war, abundant business-related pointers to Sun Tzu to the contrary.

It turns out that the planet is really small. Expect to cross paths with the same people over and over. Be nice.

Deliver value and build a reputation for good work…even if this means backtracking to cover the occasional screw up at a financial loss. Be exacting and never lower your standards. Deliver only those things you would like to receive yourself.

If you're lucky, the people in your network (see above) will also follow the categorical imperative.

Know What it Means to Look Into the Pit. Achieve the Buddha Calm. (Leadership is Critical)

Business is not always easy. Sometimes developing a market is hard. Sometimes economic times are difficult. Sometimes doing the right thing is non-trivial. A good portion of leadership is avoiding panic (Douglas Adams had that right). Know what it means to "look into the pit," especially if you're responsible for the livelihoods of others. Be responsible for your employees and their families.

Adult supervision is always helpful when you're in startup mode. Seek advice and council from others who have done it before.

Set realistic goals with (not for) your employees. Harkening back to the categorical imperative, note that all successful individuals have goals as do all successful businesses. Getting these goals to align to a common purpose is the key to success.

Develop and use metrics. These may be financial metrics such as EBITDA (earnings before interest, taxes, depreciation and amortization…a proxy for cash), or they may be related to growing people or building things, but clarity and agreement on measurement makes setting and accomplishing goals possible.

Be Patient. Be Consistent. Develop a Rhythm

Contrary to popular mythology about startups, getting rich quick only happens with the lottery (which I believe should be called the "stupid tax"). Do not expect to become an overnight financial success just because you're associated with a startup.

In fact, expect just the opposite. Most startups fail. Of the ones that succeed, a vast majority require decades of hard work. When you're contemplating success, do so in terms of years, not months.

Patience and endurance are critical to success in the long term. One helpful technique I have hit on is to develop a "communications rhythm" to help with consistent and persistent delivery. Rhythm makes the endurance you'll need bearable.

I have several rhythms I am executing against now, including this monthly column for InformIT (a column in existence for five years), the Silver Bullet Security podcast which also comes out monthly and is on episode forty-two, and the Reality Check Security podcast, a relative newcomer with nine monthly episodes.

Larger scale rhythms are possible as well. I am a bit off my rhythm when it comes to books (the default rhythm seems to be one every 2 or 3 years), but hopefully I can give myself some slack after producing nine. Doing lots of writing is a relentless challenge, but it is important as a communications tactic.

Your rhythm may not involve writing, but the same technique can apply to any kind of creative activity.

Have Fun. Do What You're Passionate About

I must confess, I have no idea what to say when people complain about their jobs. I love my job, and if I didn't I would find one that I did. Since your job is one of the most important parts of your life, it's critical that you love it. To be pithy about it: love your job so you can live your job.

You can't fake passion. If you're truly passionate about what you do, other people will see your passion and align. Do not compromise your passion to make a buck — especially in a startup!

Utilize Your Network to Build Great Stuff

I used to believe that I would build my business network until I was forty, after which I would leverage my business network. Wrong. Now that I'm past forty, I have been disavowed of that belief! A network of people takes lots of care and feeding — in person. That means plenty of travel, and plenty of wine.

The good news is that it's true that many minds are much greater than one. If you develop a good network, it will be helpful in ways you may not anticipate.

Sharing your ideas generously and publishing things so that others can use them is important as well. Don't try to hoard ideas. Share them. And don't try to squirrel away opportunity in a jar; instead, exploit it together for the common good. Hopefully the BSIMM is a living example of how a network can work to everyone's advantage. The BSIMM was possible because I knew the people running software security initiatives all over the country, and we trusted each other. Publishing the BSIMM under the creative commons license (which allows free unrestricted use by all) was an important part of repaying that trust.

Two important caveats: Your mileage may vary with these lessons. Think of them as guidelines, not laws. Running a startup is just as much about the journey as it is about some particular destination. Your implementation will always be uniquely yours.

The original powerpoint from the CSCS 4753 "Electronic Commerce Technologies" lecture can be found on the Justice League blog.

  • + Share This
  • 🔖 Save To Your Account