New ventures

TL;DR

Starting a new venture is hard work and success is unlikely So do it for the right reasons, because of your values, your interests, and because it is something you’re driven to do. It’s about the journey, not the destination.

Be careful who you bring along. You will most certainly find your share of naysayers, assholes, abusers, and false prophets. Don’t let these people take you down, you’re worth a lot more than any of them and in the end this is your journey not theirs.

The market (or lack thereof) will give you feedback that you don’t want to hear. Pay attention and make sure you are solving a valuable problem for a paying market. If you can scratch an itch, do something you love and make a good living from it, what more do you want?

Don’t discount the value of luck and timing in the wild success stories you read about. If what you’re working on becomes a massive success be thankful and give back to your community.

What is a New Venture?

A new venture is not necessarily a funded startup. New ventures are the development of products or services that are new to a business.

A startup by definition does not have a per-existing product or service or a functioning business model. It is all about finding the market, finding product market fit, developing the product, finding a product channel fit, and ensuring that it is a profitable or growth business.

For existing businesses, developing new products or services for their current markets, adjacent markets, or new markets is also a new venture albeit much more constrained. There are also added barriers in dealing with the corporate antibodies and competing business priorities. Some or all of the target market, business model and market channel may be predefined.

I believe that the challenges whether you are are in a startup or in a new part of an existing business are the same.

Past Experiences

I have had the fortune to work across a lot of different technologies and in many different industries over the years. I was also lucky enough to work in all size businesses from startups, small businesses to large multinationals. I lot of what I’ve worked on has not be commercially successful but I have learned a lot from all of it.

Here are some of things I’ve done in the past and some of the lessons learned:

• Car trip recording device (early 1980’s). This was small embedded controller module mounted inside vehicle. It used one of earliest removable solid state cards, I think the card apacity was on the order of megabytes and the device memory was about 64KB (if that). It suffered from static issues resulting in data loss , but it was my first exposure to C and running applications out of EEPROM. A great idea but much too early for the market and the technology.
• Brokerage applications (mid 1980’s). This was a small business that had a standalone MS-DOS application written in C for individual stock brokers. I was brought into convert this into a multi-user application running Unix. I spent a lot of time porting between the dozens of Unix variants and I learned about all the poor assumptions we make when we write code. The famous 1987 Wall Street stock market crash wiped out a large percentage of the customers and the business never really recovered fully.
• Trading application (late 1980’s). In the same small business, I managed to convince the management that we could creating an expert system that we could sell to in-house traders for the large financial institutions. We got some government grants that got me a Symbolics Lisp Machine and forward chaining inference tool. The idea was to create spreadsheet interface backed by an expert system with computational back end (Black Scholes and other models). The traders were going to be able to define their own rules and it was going to be deployed on a DEC VAX. Around the time I was getting started, the business ran into trouble and the AI Winter set in. Within a year the LISP machine and tool was repossessed. It turned out to be really hard to run complex real time pricing models on a VAX within an inference engine. A great idea but much to early with really negative business conditions.
• Medical Informatics (1990’s). I spent over seven years in three different companies building a nursing information system. I was one of the first software hires in startup that built a wireless handheld point of care device (POC) with a nursing workstation. We got this working in 1991 but the business ran into financial difficulties. It was acquired and shutdown a couple of more times. I had the chance to work with a lot of technology that was new and unfamiliar to me. The outer case of the POC was 3D printed using a SLA printer about the size of a couple of large refrigerators. The boards were made in house using SMT parts. It was amazing to see our technicians make the solder masks, place the parts and then put them in an oven to solder the boards. We wrote the application in C++ on Sun SPARC workstations. This was in the early days of C++ when the compiler was still being maintained by Bjarne Stroustrup. There was no standard library so I ended up writing my own data structure library and wrapped the Openview libraries. We went to a cross platform C++ library for Unix for a bit before we reworked the system for Windows. We had a domain expert who drove the design of the UI and the underlying data model. This turned out to have some significant problems which we managed to fix. We moved into supporting standardized care plans as well the nursing applications. Unfortunately for us, the medical practices were not ready for handheld devices and the clinicians were not ready to adopt standardized care plans.
• Telecommunications (late 1990s, early 2000’s). I was lucky to get to work for a new wireless broadband division for a large multinational vendor. It had been a startup that had been acquired by the vendor. It was during the time of massive acquisition sprees with the resulting chaos and confusion. I got to learn a lot about five nines reliability and the resultant system engineering issues when dealing with wireless. I then got into a venture capital financed optical network startup. I arrived in August 2001 just in time for 9/11 and the slow motion implosion of the company.
• Imaging (mid 2000’s to mid 2010’s). I managed to get into a R&D center for a large multinational working on imaging applications. I learned a lot about producing multi-lingual PC applications that shipped in the millions. Embedded systems have reliability as a cornerstone, but you face the same sort of problems with high volume software without the same practices. These were the days before updating over the Internet became standard practice. I was then very lucky to lead work on new applications for imaging but tied to print outputs. This was my introduction to cloud computing and the first time I had to deal with multinational distributed teams. Don’t let anyone tell you outsourcing is a cheap solution. There are costs associated with management, miscommunication, and cultural mismatches.
• Legal software (late 2010’s). I got into business with a lawyer who wanted to develop a legal management system. We worked together for some time but the core idea was that lawyers needed a system to manage their notes made on legal content such as court cases. These case notes would then be shared with other lawyers and the lawyers would pay a subscription fee to this service. I was assured that there was really no competition out there and that if we built it, the lawyers would come. The initial focus groups and market discussions supported this view. However I was told we could not test it unless we had a fully working platform. One thing that I had noticed was that none of the lawyers we talked to paid for any research support services. They relied on free legal resources. The lawyer that had started this business did not pay for any services either. I had held off for a long time in actively developing the software because I think in the back of mind, there was a fundamental doubt that we were solving a problem that our market would pay for. Most telling was that Evernote and Onenote were not being actively used. Some lawyers had Dropbox folders full of scanned marked up paper copies of legal cases and legislation. It also turned out that the lawyer had used Onenote for a while but abandoned it. In the end I left the business. The lesson that I learned here is that if you can’t find competition there probably is no business. It is fine to talk about niches and positioning but unless you can find evidence that your target market is spending money to solve your problem you don’t have a viable opportunity.

Reflection

When I look back at what I enjoyed and the lessons I learned over the years, it comes down to a few key observations

• Take risks. I have taken a lot of risks in terms of career directions, income, and opportunities. I regret none of these. What I do regret is not having taken other risks. For example, I have only recently decided to start my own business. I have always relied on choosing employment rather than choosing myself.
• Work with people you like and respect. Most people have a gut feeling when they first meet people or develop a gut feeling over time that something is wrong. Respect this feeling and don’t get involved with people that ping your radar. Extract yourself as soon as possible when you feel things are not right. Most of my worst mistakes have been when I’ve overruled my feelings or felt a misguided sense of loyalty. No amount of a gain is worth dealing with people you don’t like and cannot respect.
• Learn the problem domain. I have found that relying only on domain experts without developing your own knowledge about the domain leads to serious mistakes. Everyone brings their own experiences, biases, and point of view. Learn what those are and use them for your own purposes. Getting out there and talking to customers means being able to converse with them effectively. In order to do that you need to understand the language being used.
• Enjoy working in the domain. People are attracted so certain types of domains, if you don’t like the domain or the people working in the domain find another. Don’t fool yourself into finding something interesting about a domain you don’t care for. The legal domain is an anathema to me so I focused on the NLP and Recommender System problems and I missed the crucial issues until I had wasted a lot of my resources. Your time is you most valuable asset. Don’t waste it working in areas you don’t care for.
• Always check the market. Nothing is static in any market or business. In order to be successful over time you need to be checking with the market. Check your problem hypothesis, what competitors and new entrants are doing, what industry thought leaders are saying. Develop your sense of what the market is looking for through your own research and discussions. Be open to listening; even to things you’d rather not hear. When I worked in medical informatics it took me a long time to understand that the deepest problems could not be solved through the application of technology. Once I did I got out of the field. When I worked in the imaging R&D area, I learned a lot about why people took and shared photos. It was obvious that the future was not about cameras and printing but about being able to capture an moment in time and share it. Mobile phones were going to win the war and decimate the business but this was a highly unpopular opinion.
• Always keep learning. Whatever craft or trade you’ve learned, it is constantly changing. If you want to be effective keep learning. Sometimes what you’ve relied on in the past is no longer relevant but you can transfer your knowledge to the next new thing. In my field of software there is constant change but many of the new things are variations of things gone before. One mistake that I made is not practicing enough. I keep up in the reading side of things but didn’t do enough actual development during my 12 years in management. I’ve rectified that over the past couple of years but it has not been easy.

If you’re in a new venture or thinking about starting one, good luck and enjoy the journey.

Till next time.