Michael Frank's Blog

A few minutes  ago, my friend, who does not work in tech, but is interested in starting a business that would at minimum have a strong web presence, texted me:

“I wanna read a bit more about startups (particular web-based stuff). Give me some sweet reads”

Here is a slightly polished version of the email and reading list I sent back:

Even though I spend all day reading about tech startup stuff on the internet, I would say it is not extremely useful in terms of helping with execution. It is way more important to just try to execute your idea and then search around when you have issues or hurdles or you want another perspective about how to complete a task. (Most of the time when you run into an issue, someone has already built a solution or at least written about the issue before)

Having said that, here are the best resources I know to learn basic startup internet business stuff:

• http://platformsandnetworks.blogspot.com/2012/12/managing-startups-best-posts-of-2012.html (a well vetted list of quality posts)
• http://platformsandnetworks.blogspot.com/2011/01/launching-tech-ventures-part-iii-tools.html (a bunch of good skills for a business side guy to know)
• http://en.wikipedia.org/wiki/Lean_Startup (the lean startup movement, also paraphrased as do small experiments to validate your assumptions, is extremely impactful)
• If you have questions, Quora.com is google for most things startup related
• The best blogs to browse: cdixon.org, abovethecrowd.com, bothsidesofthetable.com
• Good tool kit of startup resources online: http://steveblank.com/tools-and-blogs-for-entrepreneurs/

I didn’t credit them in the email but many thanks to Tom Eisenmann, Eric Ries, everyone who contributes to Quora, Chris Dixon, Bill Gurley, Mark Suster and Steve Blank for the great content and curation above.

I am a major fan of the emerging “quantified-self” space – that is companies that help you measure your own activity. I definitely subscribe to the idea that: “if you can measure it, you can manage it.” With that said, I have been very excited to buy one of the wearable electronics wristbands that have recently come out. The two I settled on were either the Jawbone Up or the Nike FuelBand.

The major differences in my mind:

• The Jawbone Up measures your sleep (++ Jawbone)
• The Nike Fuel Band syncs the data with your phone via blue tooth (++Nike)

The minor differences:

• I like the aesthetic look of the Jawbone Up better (+Jawbone)
• The Jawbone Up supposedly has a better battery life (+Jawbone)
• The Jawbone is waterproof while the Nike is water resistant (+Jawbone)
• The Jawbone is $20 less than the Nike (+Jawbone)

By all accounts the Jawbone seemed like a better device for me until I found out one final thing that had nothing to do with the product:

• My younger brother just got a Nike FuelBand (++++Nike ->Game Over)

For those of you who don’t know me, my brother and I are very competitive and do a lot of the same sports and athletics. (My brother is also a freak who runs like seven miles every morning). Anyway, being able to easily compare my data with my brother so I can push myself vs. him is a huge difference maker and ultimately why I will buy the Nike product despite the above comparison of the specs. 

In tech, we often talk about getting critical mass so network effects kick in and in the case of this product, a network effect of one was all that was necessary. 

UPDATE: SURPRISE ENDING

So when I went to buy the Nike FuelBand in San Francisco’s Union Square, they didn’t have it available in my desired color and size at either Nike Town or at the Apple Store and they said ordering it would be weeks to months (which is hard to believe). They did have the Jawbone Up at the Apple Store though so I went with the Up. I guess the unexpected lesson here is that while network effects are important, having your product available when the customer wants it matters a lot as well. 

Last night, I got trained and took my first scooter ride with my monthly membership from Scoot Networks, a company that can most easily be described as Zip Car for Scooters. The process of getting signed up to use the scooters could generously be described as clunky and friction filled. You have to fill out like three different sign up forms to get signed up for the service, sign up for a training time, wait a week, and do the 45 minute training. And during my first training, I found out that the garage storing scooters closest to my house is closed on Sundays (Ugh) and the connector / phone charger on the Scooter was having problems accepting my standard issue iPhone. Despite these annoyances, riding a scooter in SF was a blast! and the service seems like it will be highly practical given its low price point – at $10 / day, its less than most single taxi rides within SF.

The experience reminded me a lot of other collaborative consumption apps that I have used in their clunky early days that have gotten much more frictionless and widely adopted over time. I commuted down to a job on Sand Hill Road in Menlo Park for a few months before I bought a car by mostly taking Get Around and Relay Rides and I remember the “fun” experience of the text messages and key exchanges early in the morning with strangers. But then again, I used the service day after day because it was a genuine improvement over other rental companies where I would have paid more money and had to commute farther to get a car. Likewise, during my short stint working at Uber in its early days, lots of things lacked polish from having a black car that was confusingly white to ETAs (estimated time of arrivals) that weren’t always that accurate, but for almost anyone who was stranded in SOMA on a rainy day or in the Mission late at night not able to get a cab, the core product of a reliable on-demand ride was a great one.

San Francisco is a blast because it is the tech industry’s sandbox with many interesting ideas getting tried here first in their roughest forms. As an early tester, it is this core product value I look for much more than refinement of experience. There are lots of apps that I try that have really elegant user interfaces that just don’t provide enough value for me to use regularly.  Now creating an early stage collaborative consumption product myself, I understand that many parts of our initial product from our UI to retention and social hooks are not very polished / non-existent but I’m hoping that if we can get our core product right early, in our case online private Spanish tutoring sessions, everything else can be refined in time. 

So I follow both technology and sports pretty closely and one pattern that increasingly seems to lead to success in both sectors is a willingness to do experiments.

Within startups, one best practice for finding a successful business model is the set of principles often referred to as “the lean startup.” Basically, it means testing and validating ideas as cheaply and efficiently as possible. One of my favorite examples of a startup adhering to these lean principles is a story behind the founding of Netflix. Apparently, to test out that DVDs could safely, easily, and cheaply be mailed out, the founders just mailed a bunch of DVDs to themselves and played them after they arrived to make sure they still worked. The cost to test this part of their idea was probably a few dollars and took a few days! This past August, I went to a conference called Startup-Onomics where a number of luminary economists such as Dan Ariely and Hal Varian spoke about how to apply behavioral economic to improve operations. One of their key messages was also do lots of tests because often data can prove human behavior to be counter intuitive.

Fast forward to the Sloan Sports Analytics Conference that occurred in Boston last week and you can see the same pattern, running quick experiments is a recipe for success. Over the last few years, the Houston Rockets have been able to continuously get incredible production out of relatively unproven players (including Kyle Lowry, Chase Budinger, Chuck Hayes, Aaron Brooks, and Chandler Parsons) by using similar tests with their players. General Manager Daryl Morey explained how they continuously bring in unproven young players to fill out and play key roles on their team instead of veterans because the upside is much higher. That is not to say that every one of these young players works out or their testing process is perfect. (If it was, they never would have let Jeremy Lin go.) But being willing to take a chance on players has given Houston a major competitive advantage in finding undervalued talent compared to most other teams in the NBA.

It’s not surprising that academically anchored conferences preached the benefits as experimentation as such a process is core to academic research. However, the way that practitioners are applying this methodology is really exciting and will be the source of incredible benefit (just as the innovation of A/B testing was in optimizing websites).  In any venture, startup, sports team, etc, knowledge and data are tremendous competitive advantages and being willing to do cheap tests that may fail but will provide valuable data should be a best practice as often as possible.

People love classifying high growth venture categories whether it be “collaborative consumption” or “big data” or whatever the emerging trend is. Well, I would like to coin my own new term: “compromise products.” These are products tailored towards someone who is compromising a lot by participating in the main experience. Good examples of compromise products are salads at steak houses, the kids play land at Ikea, or the high end spa at a ski resort. Most people don’t go to steakhouses for the salads or ski resorts for the spas, but these compromise products might make the experience more pleasant for others involved. The most recent example I learned about is that IKEA has been testing a “manland” for weary husbands.  I don’t think the category of compromise products has been extended to close to enough industries.

When I’m thinking about the business viability of a product, a major issue is usually who is going to buy the product and what is their willingness to pay. Well, often when I convince someone to come with me to share an experience that they may not love but I will, then I would be willing to spend at least a few extra dollars if I can make my company happier to be there. One example of a potential compromise product would be:

• A mobile app during baseball games that is People magazine crossed with Gamecast (e.g. when Derek Jeter is up, it lists who he’s dating, past relationships, and gossip) for female companions who don’t love sports. Yes - sign me up if the two dollar app purchase makes the $100+ experience of going to a game a lot more pleasant for my guest and consequently me!

The idea is pretty simple.

So, next time you hear your date or friend complaining about an experience you enjoy - think about how their experience could be improved. I wouldn’t be surprised if there end up being a number of stories about such experiences being the inspiration for successful companies.