The first couple of weeks of my sabbatical were full of new experiences that I was excited to write about. But as time went on, the new experiences became less frequent and my writing transitioned to feature updates on Much Finer’s Blog. But, now that I’ve finally reached the end of my 6 month timeline, I find that I have a lot to talk about: a working product, and a lot of expensively-acquired perspective.
So here’s my update, if this is too much to read, check out the TLDR at the bottom for just the bullets ;)
Much Finer Update
When I first started, back in April, I had a long list of projects I wanted to work on, from iPhone apps to an A/B testing platform. I didn’t know where to start. So, I thought back through my last couple years at Bing and realized that the most valuable thing we did in user experience was to adopt experimentation as part of an iterative design process. This let us measure how good our designs were, and deliver continued improvements with each release. We started with just simple A/B tests, which were really valuable, however they had a couple drawbacks. They required that we build production-quality features, they took about 2 months to get results, and the results were often mixed - with equal number of positives and negatives and no clear reason why.
Soon after, we adopted another technique that helped us overcome some of the A/B test issues - we started conducting experiments on design concepts and prototypes using a representative customer panel. Similar to an A/B test, we asked them to select the better design and tell us why it was better. This approach helped address many of the issues with A/B tests: we could now get feedback in 1 week instead of 2 months, we didn’t need to involve the development team to build production code, it is cheap enough that we could let anyone run experiments, and we would get written customer feedback to better understand the results. (We calibrated these experiments against the A/B test results from a prior release to ensure they were accurate)
So, I figured I should start with what I knew, and start with what had been the most valuable part. There are many, many great A/B testing platforms, but there weren’t great options for the second experimentation approach. After 5 months of development, I now have a working version that will conduct 3 different types of user experiments on design concepts and ‘do the math’ to determine which one is the best, and why. We have a (very) few earlier adopters and a plan for how to take it to market. I had originally expected the whole project to last 6 months, but I ended up taking Aug off, and development took a bit longer than I expected. So I’m going to add 2 more months to my plan and focus entirely on selling and customer development to see what kind of traction I can get.
One of the big things I’ve learned over the past couple months, is that Minimum Viable Product is not a milestone, it is a phase. Each release I completed I triumphantly declared that I had completed an MVP, but… the results were difficult to analyze, …but we couldn’t tell you anything about the customer panel, …but, we were running the experiments in the most expensive way possible… With an iterative development model, I don’t think it can be any other way, because you don’t really know what the limitations will be until you get there. I would now say that an MVP starts when you have a craptastic prototype working end-to-end, and it is done when your target customer can use it without you in the room. From then on you’re in the beta stage.
Everyone wants to know about the finances. While I only expected to take 6 months off, I budgeted for a whole year just in case, and then added a $10k buffer. From a personal standpoint, there has been very little difference in my life. I don’t feel poorer even though I have put myself on a pretty strict budget.
The unexpected part is that I actually think that I’m happier now than when I had more money in my budget. I know that sounds wrong, but the first big change is that I don’t buy dumb things anymore, e.g. no new cameras, iphones, gadgets, cars, etc. For whatever reason, I don’t seem to want those things they way I used to. (I think this also has to do with lower stress levels and more free time as well, which I’ll get to). Now, I’m much more diligent about where I spend money, and I tend to focus it on cooking, friends/family, traveling and home improvements.
As part of planning for this sabbatical, I did the math to figure out how much I would have to make through a successful company for this to be a good ROI. The results are pretty clear that taking such a long sabbatical doesn’t really make financial sense - unless I beat the odds and start a really successful company. The time-value of a good tech salary, well invested, is pretty hard to beat. Add in health insurance and the complexity of tax & labor laws and you are almost crazy to do this. Or at least a little delusional - which describes every CEO I’ve ever met. Therefore, you might wonder why take such a risk? For me, it is much more about the chance to create my dream job, to have more flexibility and to be in charge of my own career. I know a few people that have been able to make it happen and they have an amazing amount of freedom in their lives to run interesting companies, incubate new ideas, and travel.
How much does it cost to incubate a software product? It is a lot less expensive than it use to be, and we are rapidly moving towards a world where prototyping is free and your costs scale with the successful growth of your business. Here’s a summary of our costs, amortized over a year:
|Corporate Entity||$400 / yr||Incorporating can be done almost completely online these days, even in Delaware. Includes registration and filing fees. Legal documents were provided by an online legal service that specializes in tech companies. I also bought an ebook written by a WA lawyer describing how incorporation works in the state.|
|Office Space||$4,300 / yr||My highest expense has been shared office space. I thought I could work from coffee shops, but nothing beats a great workspace with interesting people.|
|Website Hosting||$600 / yr||Usage fees for various hosting services (AWS, Heroku, SSL Cert, DNS). Most services are free until you hit significant traffic and then costs scale up.|
|Software||$70||I am using a lot of 3rd party software, from Google Apps for Business, to Google Analytics, BitBucket, to dozens of libraries on GitHub. Everything is free except for Sublime Text 2. I expected to use a lot more Microsoft tech, but for a variety of reasons it didn't work out, I'll write about that separately.|
$1,000 for Analytics
|Almost all of the code is either UX look & feel, plumbing, or libraries taken from GitHub. The only thing I paid for is a bit of the data analysis code that is written in R by a local statistician.|
|Marketing||$400||My only marketing expense so far has been paying for the Research Panel costs for beta customer's experiments. I think of this as free trials.|
Up until now, what has been really fulfilling about this project has been getting deep into the code again and building something with my own two hands. As a program manager/ middle-manager, my primary contribution over the past 4 years has been in all the soft-spaces between the big deliverables - prioritization, schedules, balancing personalities, culture, staffing, communication, frameworks, etc. All of the actual creation happened at the IC level, and a lot of it specifically with the devs and the designers. I highly value the role of PM and middle management, and I expect I will do them again. But it is always fun to get your hands dirty again, even if just for a little while.
The most dramatic change in my day-to-day has been in the work environment. Working from home was great for exactly 1 week, and then I started to go crazy. Coffee shops were also great for exactly 1 week, but you are still don’t have much interaction, and the caffeine makes you jitter and crazy. Shared office space has been a huge improvement, especially spaces like The Hub, where they also provide a community. At the end of the day, what I miss most about Microsoft are the people, even the high maintenance ones.
Another thing I’ve noticed is that I’m now able to get everything I need done in about an 8 hour day (or 10 hour days before a big deadline). I think the big change for me is less email. Email is unorganized, mostly irrelevant, but with a huge downside if you miss that one mail. I used to spend 2-4 hours a day on email, and it really, really drained me. I can’t believe how much more time I have in the day between this and the shorter commute. While I don’t expect this to last, I’m going to enjoy it while it does.
One thing I didn’t expect was how much of my own self identity was caught up in my role in Bing. I was a somebody in a hierarchy that I understood, and people outside of Microsoft valued. When I went to conferences people would line up to talk to the expert from microsoft! I recently went to a meetup and told them that I was a founder of a no-name startup, and I could see people immediately writing me off. It can be really humbling. But, I think there can be goodness in this too. It is a good reminder that titles aren’t what make you valuable, it is your ongoing contributions and your ability to connect with people. If I can work through this successfully, I will be a lot more valuable than if I had simply continued to improve my title at Bing.
My life before the Sabbatical was great, and life in the Sabbatical is still great, although with a different flavor. The biggest change is that I feel much more part of a community than I did before. Many times I felt like my life was a series of work meetings and social functions divided by a sea of highways. Sit in traffic and end up at work. Sit in traffic and end up at the gym. Leave the gym and arrive home or at a social function. Repeat the next day. Having spent nearly 2 months in Europe this year, I now see a stark contrast. Walking everywhere, causal meet ups at bars, knowing the people on your block, at the market and getting home early enough to know your neighbors and develop deeper friendships. These things are nearly impossible to do working 10+ hour days with a commute across the lake. But the added enjoyment they bring to my life is huge, and I hope to be able to keep it up. (To be fair, plenty of folks in Europe had this problem too, especially those consulting for U.S. companies.)
The other big change is that the nature of the stress in my life has changed from almost entirely external (from a boss or a team) to entirely internal (can I build this, can I sell it, am I meeting my budget?). It isn’t more or less stress, but since it comes from myself, I have a lot more control over it. I’m sure this will change over time, especially as I gain customers, but right now I’m enjoying where I’m at.
Other than that, there really aren’t that many changes. I expected to spend more time exercising, more time on hobbies and stuff like that. But, it turns out that working at Bing wasn’t holding me back from any of those things. The good news is that I don’t exercise less, but I certainly don’t have other coding projects on the side, or other types of hobbies. Now I work on Much Finer and spend time with friends and family. Not a bad life at all!
The next phase for Much Finer is all about selling, selling, selling. This is where I get to see if there really is a market for my product, and I expect to go through a lot of rejection before developing a customer base. This is something I had always taken for granted at Microsoft - we had a gigantic channel and world-wide go-to-market teams eager to pickup any products we built and sell them to a large pool of customers. I’m both very excited and very intimidated to spend time selling (I hate rejection), but I can’t think of a more useful experience to go through. I really believe that the big challenge for companies over the next 10 years is competing for customers’s attention and wallet. The number of companies and products are growing exponentially, but customers' time and attention are essentially fixed resources, so the competition will be fierce and very sophisticated. I’ve got 2 months, I can’t wait to see what happens.
The other thing I’m spending time on is taking a class on statistics, and another on data analysis. I’m quickly becoming a believer that the Data Scientist position will be what the Computer Scientist position was 10 years ago. Data analysis will be the next frontier for value creation, which automation has been for the last 10. I think these data scientists will be just as scarce with so few people graduating with the comp sci and mathematical backgrounds to make it happen.
- Much Finer now has an MVP done and we’re moving into a focus on selling.
- Minimum Viable Product isn’t a milestone like I thought it was, it is a phase in the iterative development lifecycle.
- I’m on a tighter budget now, and it turns out that I’m actually happier. No longer spending money on stupid stuff like gadgets, bank fees and eating out every meal. I also am focusing more on family, friends and the simple things in life.
- Financially it doesn’t make sense to take a year off. Do a paid-leave for 3 months instead. Be really certain, and a little delusional before you start a company.
- Starting a company is dirt cheap these days. All code exists, everything is a service for free (prototype). Product design and sales/marketing are the challenges.
- At Microsoft I had a place in a well-defined hierarchy, I was a somebody! Outside, I didn’t have anything to define me. It takes a while to get used to that, but it is a really powerful thing when you do.
- I feel like I live in a community, I like that. Before I had friends and work, all separated by a sea of highways. It kind of felt like I lived in a hotel.
- What’s next: Selling, selling, selling! And taking two classes: Statistics, and Data Analysis.