Kyle Mulka's Blog

Dear family, friends, colleagues, blog subscribers, twitter followers, facebook friends, acquaintances, Jeff Bezos, and random internet stalkers,

Here’s the short version of this for people who don’t like to read a lot, and just want the facts:

• I’m quitting Amazon.
• April 17th is my last day
• May 12th I will be moving to Ann Arbor, Michigan
• I’m starting a software company initially employing me, myself, and I
• I plan to launch a web app at the first Twitter conference on May 26th
• I estimate I have more than a year’s worth of personal runway to figure things out

I’d like to tell you all about the start of the next chapter in my life. But, before I do that, a condensed history. I don’t want to spend too much time on the past, so I’ll just say this… I was born, raised, and went to college in Michigan. Go Blue! I studied computer science, and one summer decided to learn web development. That led me to an internship and then a full time position at Amazon.com in Seattle, WA. The thing is though, I want to build my own software products. Basically, I was doing the type of work on wanted to do (namely web development) but not building the type of software that I wanted to see exist in the world.

Amazon is a great company. They are going to do just fine. I’m actually going to keep my stock, and maybe even buy more. Although, if I’m going to do serious investing in the stock market, I’ll have to look at the companies and pick ones I really believe in. Palm may actually be one of those. Their new Web OS on the Palm Pre looks pretty darn impressive.

Amazon is no longer a start up. Too much friction to get certain things done. My guess is that this problem is not unique at all to Amazon… any big company has issues like this to deal with. Once you have a lot of revenue, any change could have a big impact in the business. Since a startup, by definition, doesn’t make a lot of money, changes to the business or the software won’t have as big of impact in absolute terms. It’s all about risk verses reward. In a large company, you can spread risk across many different projects and people in the company. This is one reason why Amazon is able to do so many innovative things. They can take risks in certain areas like Kindle, AWS, AmazonFresh, etc, and even if those projects fail (which I actually doubt they will) Amazon itself will be fine as a company.

When I was an intern at Amazon, Jeff Bezos (the CEO) gave a short talk, and then left most of the time for Q&A. The question I asked was why he thinks Amazon was able to survive the dot com bust. The answer he gave was basically that the business metric graphs had very little relationship to the stock price. Even though the stock price tanked, Amazon as a business was still growing. Also, Amazon has very long term view of things. There is little to no change in strategy in tough times. It’s always lower prices, more selection, and relentless focus on the customer. I think that is as much true today as we go through this recession, as it was during the dot com bust.

I’ve wanted to do a startup for a long time, but didn’t want to spend the time looking for funding. I’d rather spend the time building the software instead. People have told me that it is generally a good idea to build it first, before you start looking for funding. I still would rather not have to spend the time seeking out funding for a startup. So, I’m going to start by bootstrapping instead. This means that I will have to have revenue very early on, ideally on day one. So, I’ll be doing some experiementing with that model over the next few months.

Why am I moving back to Michigan? Well… my family lives there. I have some friends there. There is a small, but growing tech community in Ann Arbor. It is much cheaper than Seattle.

I’ve just jumped off a cliff, and I’ll be building the airplane on the way down. I estimate that I have more than a year’s worth of falling before I hit the ground. Wish me luck!

I am officially announcing my retirement from mapping projects as of today. Today coincides with the start of Where 2.0 which I attended last year. Some may know that I decided not to go this year. Along with Where 2.0, there were also two related events, Google Developer Day and WhereCamp which I attended. Google Developer Day which was free at the time has turned into Google I/O which now charges the equivalent of about 1 iPhone. WhereCamp kind of rode on the shoulders of Where 2.0 as a smaller free unconference for people really interested in geo stuff.

I’ve always been interested in maps. When I was a kid, I would always want to be the navigator with map in hand while riding in the car up north with my family. In boy scouts, I was very interested in orienteering… starting from a given point, navigating to other points using a compass and a map, or sometimes only a compass. Once GPS devices got cheap enough I bought a simple, but proven Garmin eTrex and went geocaching with it. When the Google Maps API came out I got really excited and started developing all kinds of different things with it. I even created a maps subdomain and titled the page Kyle Mulka’s Google Maps. While in college, I worked with my good friend Dan Feldman on a site called liveUgli, which was a real-time study buddy finder. We used the Google Maps API, but instead of the typical world map, we used floor plans of the major study locations on campus. In order to make the site more extensible, I developed two offshoot projects. The first, Gmap Uploader, had the goal of making it really easy to get floor plans into the system. The second, Cartiki, which actually uses the Gmap Uploader, was designed to make it really easy for users to edit the locations in the system.

So, why am I doing this you might ask. Well… there’s a few reasons. The mapping projects that I have worked on so far haven’t gotten much usage. When I was playing around with the Google Maps API, I wasn’t really interested in web mapping itself, I think I was more interested in just playing around with new and cool technology. With the liveUgli, I was interested in making the life of college students better. With the Gmap Uploader and Cartiki, I was interested in building reusable components that could not only be used for liveUgli, but could be used by other developers for their projects. So, I get the feeling, I wasn’t really into mapping so much as the technology behind it, and then the applications of it. Now, I feel as though I have exhausted the interesting and useful stuff (at least to me) in the mapping space, I’m going to move on to other, more interesting things. Hopefully nothing too revolutionary (like the Google Maps API when it came out) gets released at Where 2.0 that will cause me to change my mind.

What am I going to do if I’m not going to be working on mapping projects? Well… I’ve got a few other things up my sleeve, which I will tell you about in more about in upcoming blog entries. Also, don’t forget that my day job at Amazon doesn’t have anything to do with maps, and they disbanded A9 Maps, the only mapping project they had, a while back.

So, if the only reason you are subscribed to this blog is because of your interest in Google Maps, maybe now is a good time to leave. Or, you could stay and read about my new adventures in the next phase of my life.

It really shouldn’t be this cold in Seattle this time of year. I guess I wouldn’t know since I’ve never been in Seattle this time of year… but it has only been this cold a couple times this winter. It even snowed a little the other day.

Last weekend I participated in this thing called Startup Weekend in Seattle. Basically, there were over 100 people attempting to create a startup company (basically a web app) going from concept to launch in just over 50 hours. It was pretty intense.

The weekend started Friday evening with everyone sitting at tables of 5-6 people. Each person shared their idea with the table. They were then written down and one person took the ideas and pitched some of them to another table. The other table decided which 1 or 2 ideas were worthy. These ideas were collected at the front in paper form then randomly distributed to different tables. Each table read the idea on the paper separately and voted either yes or no to the idea. Only about 5 ideas made it past this stage. My theory is that ideas on paper in a sentence or two don’t get people nearly as excited as if someone else was pitching it to them. We ran through that again to get a different set of about 5. So, 10 ideas total. Each of these 10 was pitched to the entire group by the person who originally suggested it. We took a vote on all 10. You could vote either yes or no to each one by either raising or not raising your hand. There were about 4 that made it past this round. Next, there was discussion about each of these 4, and different people talked about the merits of each. Then, a final yes or no vote from everyone and we came out with a clear winner. We were going to build a RideShare that actually worked and that people actually used.

So… now to figure out exactly what we are building, and who else is building similar stuff. An hour or so into this, and the business development people come back with their findings on the competition. They determine that there are some strong competitors in our area, namely Goose Networks. They are doing a lot of the stuff we would be doing and it would be difficult to differentiate ourselves. So, we took a vote if with this new information we should switch ideas. Most people thought we should switch, and so we decided on the “Match.com for Enterprise” idea.

Basically we would create a web application that would allow small to medium sized businesses get a better sense of what skills their employees had to better utilize these. Each employee in the business gets a profile which has basic information plus their answers to certain questions the boss had asked. So, the boss could ask something like “What languages are you fluent in?”. Each employee would answer this question and their answer would show up in their profile. Then, later you could search through the system to find an employee fluent in Spanish to talk with a new client. This is just an example. I’m sure there are plenty of questions you could ask to get useful answers to help in finding skills in your organization. What are we going to call this? We call it SkillBit. The skill part is easy to understand. The bit part represents the simplicity of the software and lack of features I suppose. As with any good web application, it does one thing, and does it well… hopefully.

We have yet to launch SkillBit publicly, but you can sign up on a waiting list to get an announcement of when it launches.  If you want more details about how the weekend went, you can read the Seattle Startup Weekend blog which was posted two by two people throughout the weekend.

A lot of stuff has happened since I last posted, both in the world and in my life. I’ve been hesitant to blog about it mostly because I have to spend time doing it for, what seems to be of little benefit. I’m not even sure who really reads this. I know my family reads it, some of my friends might read it, other folks out there interested in Google Maps stuff might read it, and maybe some other random people out there might read it. I’m going to try to post more often, and am going to do a few quick posts right now.

After a few hours of playing with the Amazon Kindle, here’s what I’ve found so far.

It’s way too easy to accidentally hit the huge side buttons. You have to try not to hit them by always holding the device at the bottom and being careful to put your fingers somewhere where there isn’t a button.

Getting free 2 week trial subscriptions to newspapers and blogs is cool, but it looks like I have to remember to go and cancel them if I don’t want them. Getting previews of books is cool too. At least those don’t automatically get bought after 2 weeks.

The Kindle is indeed lightweight and easy to read. Although its still possible to get glare on the screen, its far less than on other types of screens, and looks great even in direct sunlight.

The back cover is really hard to get off, but I guess it was designed that way. At least it doesn’t come off as easily as the side buttons get hit. The battery and memory card slot are behind the back cover.

It comes with a book cover that is pretty cool, but it falls out too easily. They should have made the locking mechanism tighter.

You can start using the Kindle as soon as you get it. It doesn’t come charged, but it works while its charging so that’s not too bad. The Kindle is tied to your Amazon account when you buy it. So, from the home page where a list of all your reading is there was a personalized letter from Jeff Bezos. In the upper right, it says Kyle’s Kindle. I would assume yours would have your name on it instead.

The Kindle saves your spot in every piece of reading you have and shows you how far you are in each item on the home screen.

Something that isn’t talked about as much is the Ask NowNow service. From the Kindle, you can type in a question, and someone will browse the internet to find an answer for you. So, if you are driving around looking for a place to eat you can ask your Kindle “what’s the best place for burgers in Seattle?” and it will give you answers from 3 different people. In this case, the answer is Dick’s Drive-In at 111 NE 45th St. The people answering also listed several other places and gave a link to the CitySearch page where they got the results. I’m assuming they just used Google to search for “burgers seattle”. Which, I guess I could have done because the Kindle has a web browser.

Yes, that’s right, the Kindle comes with a web browser. It displays text and images in black and white. So, Wikipedia does indeed work, but not that well in my opinion. It says it supports javascript in advanced mode, but regular Google Maps doesn’t work. The non-javascript version sorta works, but the Kindle cuts off images if they are at the bottom of the first screen of a web page with no option to scroll half way. You either go to the next screen full, or the previous screen full.

The Kindle has a cool ‘screen saver’ type thing which gets activated after a while of non-use. Alternatively, you can activate it manually. I don’t think its actually saving the screen at all, but it does lock the kindle so that you can’t accidentally press buttons.

There is a back button! This is great for when you want to look up something, go to another book, check something on the web, then want to go back to wherever you came from. This not only works in the web browser, but in books, newspapers, in the menus, etc. So, its actually not that difficult to go to different sections or articles in the newspaper. You don’t have to click ‘next page’ a bunch of times. You just click the name of the section or article, read a little into it, then click the back button and you will be taken back to the list of articles or sections.

Yup… that’s right, I shelled out $400 for an oversize PDA with a black and white screen that has a refresh rate of about once per second.

On the other hand, it has a free wireless internet connection wherever you go, replaces the weight of over 200 books, and has a screen which looks more like a book than a computer screen.

While reading Don’s blog (from the last post) I saw that SmugMug was having a contest this month for who could build the coolest application using the SmugMug API. The winner gets a free iPhone. So, I think I’m going to enter and make something cool. Wish me luck. As a side benefit to using their API, I get a free lifetime Pro account, so I’ve uploaded a bunch of photos I had sitting on my computer. You can view them here.

Today felt like a really productive day, even though I didn’t get a whole lot of sleep last night. Today I actually did some real development as opposed to bug fixing and small tweaks to our websites like I have been doing for a while now. I did quite a bit of refactoring today, which is fun if you understand what’s going on with the software. The best analogy I’ve heard so far for refactoring is reshaping a piece of clay. You take a glob of code here, and move it over there. Take some extra code out that you don’t need, and make the code look nicer. Nicer looking code is easier to maintain. Basically, if you have to fix or change something later it will be easier to understand and fix without too much work. Refactoring can be frustrating if the code is too ugly in the first place. If you don’t really understand what the code is doing, you can really screw it up by moving around those globs of clay, or code in this case. But, this code was pretty clean and simple. It just needed some minor refactoring.

After work I went to an entrepreneur networking event at Del Ray hosted by nPost. I met some cool people working on some really cool startups. Several people from Zillow were there, and I talked to one of them. I talked to a VC who was working on his own business plan for something in the mobile shopping arena. I met the two guys, John and Tom from Wishlisting.com. You should check out their site. It’s pretty nifty. I think I’ll start using it. I was directed towards another entrepreneur, Matt, who is working on a site for online video tours of real estate. We talked about how he might be able to use my Gmap Uploader for the floor plans on his site. Definitely check out his site called Cool Toors (a second o instead of the u) if you have property for sale or rent and want a good way to put up a virtual tour.

So, Ajay, my current roommate and I have a running challenge going: See who can run the most miles before he leaves on the 24th. Right now he is kicking my butt. He ran 12 miles in the past three days. I ran 4 miles one day, but I need to keep that up in order to beat him in the challenge. Wish me luck.