Engineer Entrepreneur: Peter Sichel April 2001
My first interview is with Peter Sichel, founder of Sustainable Softworks.
Rentzsch: | When we first met at MacHack (1999, I believe), you told me a little about how you started Sustainable Softworks. If I remember correctly, you used to work at DEC on networking. One day, you resigned and started learning Macintosh programming with the intent to write IPNetMonitor. After six months of almost nothing but reading, you started Mac programming in earnest and published your software. Could you fill in the details and offer any corrections to my recollection? |
Sichel: | I think we met at MacHack 2000. Your recollection is quite good. As you remembered, I worked at Digital but in Video Terminals Architecture. "Architecture" is a funny name for people who understand the big picture and help an organization document and remember what it learns from one product to the next. Video terminals include real time operating systems that deal with communications, so it was a fine place to learn about software design and networking.
I fell in love with the Mac in 1986 and saw right away that the User Interface was how software should be done. I used Macs extensively at Digital and became part of a small group of Mac fans in an increasingly PC dominated company. As Digital shook itself apart for having repeatedly misunderstand the Personal Computer revolution, I needed something else to do. Talking one day with my fellow Mac heads, we lamented the lack of a software IP router for the Mac. We all had several machines at home and increasingly needed them to be on the Internet. With Apple adopting Open Transport, a way to do it became clear.
I resigned from Digital in January 1996 and set out to build a software IP router for the Mac. I was an experienced software developer, but had never seriously developed for the Mac. It was a huge leap. My wife and I had agreed I needed to have at least one consulting client and meet some minimum financial targets each year ($10K, $25K, and $50K), otherwise I'd have to get a real job.
In my own mind, I had reasoned that doing a software IP router was too hard for a casual effort, and that there wasn't enough money in it to attract much larger competition. I also knew from Apple that more than half of all Mac users owned more than one Mac, and that TidBITS was being sent weekly to over 50,000 subscribers. It wasn't hard to see a market for at least a few thousand software routers, enough to recover my investment.
In addition to consulting on the side, I set out to learn TCP/IP, Open Transport, Mac programming, and PowerPlant. I spent the first 6 months mostly studying in a spare bedroom I had claimed as my office. My wife wondered if I was ever going to "do" anything. Later she sometimes joked saying "show me the money".
My first Mac program was the Subnet Calculator (now part of IPNetMonitor). Gradually I started filling in more of the pieces. I knew I needed a module below IP to do Network Address Translation, so I built the Monitor tool to test it.
By Macworld Boston (August 96), I set out to build some relationships at Apple. I found the networking folks in the Apple booth, and demonstrated my little tools from my PB520c. I told them I was writing a software IP router for the Mac, was serious, and could use some technical help with Open Transport. Apple's network evangelist at the time was stunned to discover a guy who wanted to solve this problem for them, and wasn't asking for any money. He put me in touch with the Product Manager and tech lead for Open Transport.
Over the next 8 months, Apple's plan to release OT1.5 with IP multihoming kept getting pushed out. The OT Product Manager suggested I release the tools I had built as a product to start earning some money (Subnet Calc, Ping, Monitor, and TraceRoute), thus IPNetMonitor was born.
In March of 1997, Open Transport was "cancelled". I wrote to everyone I knew at Apple pointing out that they still needed IP networking and since they were not planning to do it, to please let me. All I needed was a little information. With higher management otherwise distracted, the OT Product Manager thought this was a reasonable idea and agreed to help.
I learned from Apple that Open Transport used the same network routing APIs as BSD and wrote a bunch of code accordingly, but I couldn't quite get it to work. I could see there were some include files missing. Once again I pleaded with the OT Product Manager for help. Within a few hours of receiving the missing include files, my code suddenly came to life. I could do IP multihoming and forwarding. The rest followed.
Having come to the Mac platform fairly recently, I started right away with PowerPlant. I'm still embarrassed sometimes by how little I know about the Toolbox and other traditional areas of Mac programming. I now believe one of the factors in my success was figuring out early on there were a lot of things I didn't need to learn right away. I started building a real product, and would study and learn what I needed along the way.
|
Rentzsch: | How did you come up with your company's name? |
Sichel: | In the 1980s I became involved with a volunteer group looking at what was needed on the planet to address some of our most pressing and intractable problems. As I mentioned above I'm a "big picture" person and naturally allured by understanding big problems and what can be done about them. Having seen how DEC mortgaged its future and the need for sustainability in many areas of the economy, I chose Sustainable Softworks as part of my response to what is needed. Sustainability and new ways of thinking (soft as opposed to hard technology). The pun with "software" made it more appealing. To this day, all our employees work from home (no commuting) and we have no retail packaging. I have tried to do business with a sense of generosity or "Earth spirit" as I understand it.
|
Rentzsch: | You mentioned that you and your wife agreed to have at least one consulting client when you started out in 1996. Do you still do consulting today? |
Sichel: | Yes, but I'm much more selective about accepting new clients. Since there's plenty of work for me at Sustainable Softworks, I mostly accept clients who are looking for something I'm already working on or that fits with our current business. |
Rentzsch: | I see you used Apple's statistics and the TidBITS mailing list size to estimate the size of your market. Looking back, how close are the actual numbers compared to your initial estimates? If you knew then what you know now, would you have used the same sources to estimate your market size or would have used others? If so, what others? |
Sichel: | My initial estimate was intended to show there was a real market for a software IP router. A lower bound. I figured I needed to sell at least 2000 copies to recover my investment. The actual market is much larger. IPNetRouter has over 13,000 registered users not including licensing deals with a few large customers. The return on investment has exceeded my expectations. If you remember back to 1995, the Internet was not as prominent as it is today. Netscape had recently introduced the point-and-click HyperCard user experience to distributed information on the Internet, but it was still too hard for average home users to get connected.
I think the numbers I used were good because they were based on real information I could see and understand, not abstract projections of how fast the Internet would grow.
|
Rentzsch: | After the initial software purchase you don't charge for support, bug fixes or minor upgrades. |
Sichel: | That's correct, we haven't needed to. |
Rentzsch: | Unless I'm missing something, that only leaves major upgrades as a post-sale revenue source. What's your best source of revenue after the initial sale? Cross-selling, where you first sell them, say IPNetMonitor and later on IPNetRouter? |
Sichel: | Our best by far is word-of-mouth referrals. Satisfied customers refer other customers. Cross selling has also been a strong source of revenue. We don't offer just one product, but four, each with compelling features that are hard to find anywhere else.
|
Rentzsch: | I have friends who believe charging for software and "giving away" documentation and support isn't a sustainable business model. They advocate giving away the software (usually including opening the source code) and selling documentation and support. What's your take on this model? |
Sichel: | What an interesting question! I think people arrive at different economic conclusions based on different assumptions. At MacHack 2000, Eric Raymond talked about the "Cathedral and the Bazaar" as metaphors for traditional software development versus Open Source. There are several important ideas in the Open Source model, a few of which I will review below. In the final analysis however, I don't think it is black and white at this time. I see room for elements of both approaches.
Let's review some of the key findings of the Open Source movement:
- Software is labor intensive, expensive to produce, and often poor quality (buggy) despite the substantial efforts of talented people.
- One of the largest cost factors is finding and removing bugs.
- All bugs become shallow (easy to resolve) when enough (of the right) eyes are looking at the problem.
Open Source is not so much a business model as a collection of techniques and agreements for getting more of the right people involved in the software development and testing process. The goal is to radically reduce the cost of producing high quality software.
I have adopted some of these techniques already including: (1) Release early and often; (2) Develop and test new features incrementally; (3) Work aggressively with end users and other experts to resolve bugs as early as possible in the development cycle.
Customers are delighted to discover we can fix most bugs in a few days or weeks, and release new test versions regularly.
As for the sustainability of this model, currently I think it works in part because we are small and found ourselves in the right place at the right time to leverage existing Internet technology. My experience of working at DEC, a once proud Fortune 500 company with vast technology and talent that ultimately failed was very sobering. Allow me to offer two lessons I gained from this experience:
- The purpose of a business organization is to serve customers (not to maximize profits).
- The model for sustainability is the Earth life system. It operates through differentiation and mutually beneficial relationships so that nothing is wasted.
Sustainable Softworks has a very low overhead and responsive software business model that supports a few of us to make a decent living. We sell directly to customers and encourage them to participate in improving our products.
As we learn from experience the model will adapt, but our purpose is to serve customers and build mutually beneficial relationships, not to grow for its own sake or maximize shareholder value.
|
Rentzsch: | Your software is included in AirPort. Can you discuss what your software does for AirPort, how the deal went down and what it's like to waltz with an elephant (Apple)? Or are the details hush-hush? |
Sichel: | IPNetRouter is used to configure IP multilink multihoming and forwarding in Open Transport, and also provides the NAT and DHCP Server functions of Apple's AirPort Software Base Station. This is public information since some of the software components supplied with AirPort are so identified. In general Apple's technology licensing arrangements are covered under NDA so there's not much I can say about that.
For me, Apple was a very demanding client, but also had a lot to offer. I hope we collaborate more in the future as appropriate.
|
Rentzsch: | If you could travel back in time and meet yourself before you started Sustainable Softworks, what are the most important lessons you'd pass along? |
Sichel: | Starting Sustainable Softworks was a dream that I am still living. I'm amazed sometimes I found the courage to do it. A lot of people helped along the way. I think my most important lesson is don't be afraid to try living your dream. Even if you start small and it doesn't turn out the way you expected, the experience will be worth it. Most people who start a small business are very good at what they do. They wouldn't have started otherwise. The next challenge is learning enough about how to market and run your business to pay the bills.
Some ideas that have helped me:
- Don't spend money you don't have;
- Go the extra distance to treat your customers well;
- Ask for help when you need it.
Some of our early customers became my best teachers.
|
Rentzsch: | What mistakes have you made? |
Sichel: | Lots, that's how I learned. One of my most memorable was negotiating a site license fee with a large telecommunications company I'll call XYZ. I suggested a license fee of $500. The person on the other end of the phone said, "You don't understand, this is XYZ you are dealing with, if you say $500, they won't take you seriously. Let's make it $1000." Another mistake was waiting too long to get help. As the business grew, I kept trying to do everything myself and almost burned out. My wife pushed me to hire my first employee.
|
Rentzsch: | What successes do you look back upon most fondly? |
Sichel: | Many delighted customers. Receiving a MACWORLD Editors' Choice Award. |
Rentzsch: | What's the corporate makeup of Sustainable Softworks? Do you have an office or do you still work from your home? Last I heard, you hired an employee. What's his/her job? Does he/she work with you in the same physical location, or remotely? Do you anticipate hiring more employees? |
Sichel: | I incorporated Sustainable Softworks in 1998 to simplify hiring employees and contracting with other companies. Before this, I was a sole proprietor. Sustworks currently has four employees. Three full time (including me), and a part time Administrator. We all work from our homes. Geographically, we're in four different states (RI, MA, FL, and CA) so we only meet in person a few times a year (like at MACWORLD).
I've done most of the actual software development and manage the business. The other employees handle our web automation and business development, tech support, marketing, customer database, and increasingly some software development.
As for hiring more employees, I would say yes and no. I expect we'll hire more people if it makes sense, but I'm not looking to manage a large organization.
|
Rentzsch: | IPNetRouter alone has over 13,000 users. What's the technical support burden like? Each time I've mailed you, you've been quick to respond, even during when I'd assume you're under a heavy email load, like during the "MacAttack" DDoS scare of late 1999. I note you've set up a user mailing list -- does that help? What software do you use to run the list? |
Sichel: | With thousands of customers, the technical support component is substantial and one we've worked hard to manage. I'm very comfortable with Email and started out answering every question myself. As the business keeps growing, I've had to put more thought into how we can continue to offer the personal response people deserve when they have a serious question. Our system is still evolving and far from perfect, but I think it works pretty well for a lot of people. Our NetTalk mailing list has been a huge success allowing customers to help each other and us as well. I got the original idea from a comment by Guy Kawasaki who said: "Any software developer who doesn't run a mailing list is a bozo" (or something similar). The list server uses LetterRip Pro, a wonderful product that makes running a mailing list surprisingly easy. I've used whatever spare Mac I had to run the list and serve as my Internet gateway including a PM6100, PBG3, iMac, and now a PM7300. Connectivity has ranged from a 28.8K dialup to ISDN and now a cable modem.
Over time we've also put a lot of work into providing information on our website including step-by-step instructions, FAQs, searchable list archives, and white papers. We try to help customers help themselves as much as possible, so we still have time to assist those who really need personal attention. One obvious short coming from creating all this information incrementally is that we've been slow to organize it as a comprehensive manual. Like the Internet itself, information is scattered in many places.
We also developed our own web based on-line support system to track individual requests and share support cases between employees. One employee reads every support case submitted, answers the ones he can, and consults or reassigns others as appropriate. We handle support a little bit like a volleyball team where everyone hits the ball sometimes. Web based forms can be clunky compared to Email, so I'd still like to see a better front end to our support system.
Two other factors that have helped us manage support are the 21-day free trial so that people have a chance to see up front whether the products are right for them, and the nature of routers themselves. Once properly configured, routers pretty much just work until your needs change.
Overall I think we've provided a lot of support for such a small company, but the unmet need for Internet expertise and consulting is enormous and remains a challenge. There's a constant tension between our desire to help and our own limitations. We've chosen to discourage telephone support and frequently refer people to other resources.
I wish more people recognized that Email isn't perfect. Messages sometimes get lost, are sent to or from the wrong address, have the wrong date, etc. If someone doesn't get a response from us, I would hope they could be a little understanding and try again.
As I was first getting IPNetRouter to work, I approached some traditional Macintosh networking companies about licensing and reselling the product. One Marketing Manager responded they were not interested because providing support for such a technical product would be too expensive. From their point of view he was probably right, but the Internet has changed the rules.
|
Rentzsch: | Could you talk about your marketing strategies? You already mentioned that word-of-mouth referral plays a strong factor. And I bet that Macworld Award didn't hurt, either. What other forms of marketing do you use? Paid advertising like Magazine Ads or Banner Ads? Free evaluation copies for product reviews? Direct Mail? Affiliate Marketing? |
Sichel: | This is a great question that goes to the heart of any small business. I think some people confuse marketing with sales and advertising. To me, marketing is understanding customer needs and communicating what you have to offer effectively. There's a huge difference between quietly letting people know that you are out there solving this problem, and trying to sell what you have.
By far the most important first step is figuring out what problems you want to solve. I learned at DEC that doing simple things that are really needed pays off more than complex or even cool stuff that isn't as critical. So my first two marketing principles are:
(1) Try to create solutions that people really need and will use everyday.
(2) Try to solve problems that other people are not addressing.
I started by looking at my own experience and then talking to a lot of people to expand my perspective. Once I had a product that addressed a real need and let people know I was serious about incorporating their feedback, I was amazed at the number of really good ideas and suggestions people offered. Many of our best product features came from listening carefully to customers.
My next marketing principle is to understand relationships. People are naturally cautious about buying from an unknown entity so I've tried to tell people who we are and what we're about, our story so to speak.
(3) Take the time to build relationships and trust so that people are comfortable dealing with you.
My fourth marketing principle is:
(4) Price the products to offer a lot of value.
People are taking a risk when they buy from a small unknown company. We want people to feel good about buying from us by rewarding that risk.
A lot of our marketing follows from these basic ideas like our 21-day free trial and lots of helpful information on our website. Some other strategies:
- Our on-line registration process makes it quick and easy for people to register.
- I wanted Apple to keep supporting Open Transport, so I donated a site license for all our products to Apple Computer so that more Apple employees would see the power of OT.
- We offer a 50% educational or student discount to help customers who might be less able to afford our products.
As the business has grown, we've begun to experiment with advertising and other forms of sponsorship. Since we're just learning about this, we've tried to sponsor small informational websites or other activities that we believe contribute to the overall Mac community.
I know we can't shout as loud as the big guys, so instead we try to communicate differently or in places they are not.
Marketing is a huge topic, so let me finish by recommending a book that I found helpful: "Marketing High Technology" by William Davidow. Bill Davidow was an engineer who became marketer extraordinaire and put the young startup Intel on the Map. The story is fascinating and insightful.
|
Rentzsch: | Adam Engst, of TidBITS fame, wrote a series of articles called "Hacking the Press" where he outlined various techniques to get noticed by the press. Have you utilized any those techniques? If so, how did they work out? Any insights are welcome. |
Sichel: | Adam's article was excellent. I think his most important point is to build relationships with the media. It makes perfect sense. We've been slow to do this ourselves because I didn't understand the media very well and have been busy with everything else. Adam has been helpful and we're starting to see more coverage. Some lessons I've experienced are:
- Don't call it "shareware" if you are serious about selling software to make a living.
- If you set your prices too low, people may take you less seriously. Product reviews generally try to help readers with larger buying decisions.
I find this stuff fascinating and would be glad to compare notes with others who are interested. A big part of being an entrepreneur is just putting yourself on a collision course with learning.
|
Saturday, February 15, 2003 12:00 AM
| Focus of this site Contact Me Topics RSS Feed Linkblog TwitterAndy Finnell Bill Bumgarner Brent Simmons Daniel Jalkut Dave Dribin Eric Albert Eric Rescorla Eric Sink Greg Miller Gus Mueller Jeremy Zawodny John Gruber Mark Dalrymple Michael Tsai Peter Ammon Raymond Chen Ryan Wilcox Scott Stevenson Steven Frank The Daily WTF we hates software Wil Shipley
|