Marcelo Calbucci

 

    I just published on the Seattle 2.0 blog the new Seattle Startup Index for July.

 

    Sampa moved 1 position up to #24 despite the fact that we lost 1,655 positions on the overall Alexa Ranking. The reason for the loss? WDN (We don't know).

 

    It could be because we didn't send the monthly newsletter this month, it could be that people using the Alexa toolbar didn't visit the site as much, it could be that a lot of families are travelling during the summer months, which is our target audience by the way.

 

    Since Alexa is a sample-based (without demographic normalization) system, a change of plus or minus 10% can be considered quite inside the margin of error. Our internal stats continue to show a strong growth for this month. We are expecting a growth of 35%+ on unique visitors and 29%+ on Page Views according to the latest projections. Tomorrow we'll know for sure.

 

 

    I guess I forgot to blog about it, but we just filed three provisional patents for Sampa. Two about our intelligent privacy system and one about our link storage system.

 

    For the longest time I've been reluctant at filing any patents. Sampa is not a "research" play, but a technology execution and consumer branding play, so patents are not worth as much to us as they would for a bio-tech company for example.

 

    However, provisional patents are easy, cheap and quick to file. They also give you the right of using "patent pending" terminology on your product and they give a bit of extra confidence for investors about the uniqueness of your product.

 

   

 

    If you think of words like emotional, romantic and beautiful your are not talking about me or my work. What defines me and my work is functional, accurate, comprehensive and analytical.

 

    That is the reality for all engineers. Some know that and some are in denial.

 

    I'm working on one of the biggest features for Sampa yet. Just a handful of people know about it so I can tell you yet (come back next week).

 

    A few days ago another entrepreneur friend of mine stopped by to chat and I told him about the feature, and how cool and "Ajaxy" is going to be and all that.

 

    Despite the fact that he is an engineer as well, he's very good at asking the right questions. A few of the questions and points he said made me realize I was missing the big picture. I've got what he said immediately, but only the next day the full power of what his points sunk in.

 

    As engineers, we think in terms of classes, protocols, input, output, how to cover border cases, how to make sure the input is valid, etc. We are very deterministic and analytical on our approach to solve problems. It's hard for us to create a system that solves 80% of the cases. We are instantaneously drawn to solve 100% of them. And that might work well if you are doing software that will be used by other engineers, or by businesses, but consumers are not like that.

 

    They don't care they can choose between 800 different font types, or have 16 MM colors to pick from. They care if the 2 fonts they want are available and the 3 colors they pick match each other. Better yet, they care if the system picks the fonts and colors for them. It took the Office team about 15 years to realize that, but the new sytlesheet system in Word 2007 is an example of that.

 

    Now, the good news for us about this new feature is that all the competition is making the same mistake I was about to make. I might not be able to execute perfectly on the new vision, but at least I'm enlightened by this new knowledge and it will have a great influence on how this feature will shape out to be.

 

 

 

    What an awesome event yesterday night. A lot of the usual suspects. Entrepreneurs, angels, VCs, journalists.

 

    I've gotta tell you. I think Madrona is positioning itself as the cool VC in town. There is not a single event that I go to that I don't bump into a few of Madrona team members. Geoff Entress specifically I think is on every single event in town.

 

    A big thanks for the Redfin team as well to put this together.

 

    About the panel it wasn't that good. John Cook, Tricia Duryee, Michael Arrington, Fred Vogelstein and Rebecca Buckman were very predictable. They didn't say anything that a good Marketer/PR/CEO/Entrepreneur didn't know already. They want scoop, they want exclusitivity, they want to pinpoint trends before somebody else does, they want stories that sell more papers, magazines or have more link backs.

 

    The after party was pretty awesome. I've met a ton of people that I knew already, and a few new faces.

 

    The question I've got the most from other entrepreneurs was about our Facebook integration and how was that going, which is a topic for another blog post.

 

    If you are attending the event and you see me stop by to say hello. I expect to see a lot of familiar faces and get to met a few interesting folks as well.

 

    Twango has just been acquired by Nokia. You can read more on John Cook's blog.

    I talked a few times about Twango here. One of the co-founders is Philip Carmichael that worked with me on MSN Search. Even after we both left Microsoft we kept in touch and we even did some integration between Twango and Sampa.

    This is really cool for Philip and crew. Congrats guys.

 

    It takes 3 minutes and it can make $5 MM of difference for a lot of kids. Just go to DonorsChoose.org, and click on "Vote Now". Check out the instructions on how to vote.

 

    Oliver is a former colleague of mine from Microsoft and he's currently the CTO of DonorsChoose. So, please, help your friend help a friend and ask your friends to help a friend help a friend of a friend and ask your friend's friend... you got it.

 

 

    Glenn Kelman at Redfin is asking users to pick a design for their new homepage.

    My vote goes to #1 with some modifications and before I explain my choice, let me explain why I have not chosen #2, #3 or #4.

Why not #2:

• Pictures of people is hard to associate with yourself. If your black and see a white person, you might think is not for you. If you see somebody in glasses in front of a case full of books you might think this is too complicated (since the person must be smart).

• The pictures and the text don't correlate much. Showing a couple with the caption "Draft an offer online." is pretty disconnected.
• I'm not going to talk about that agent picture, but argh! Car salesman anyone?
• And to finish. The couple that draft the offer online is not the couple that save the money!!!

Why not #3:

• this would be my second choice (behind #1).
• Same problem as before. The couple looks too smart, making the process look more complicated than it has to be.
• "Redfin is changing the real estate industry in the consumer's favor" is very subjective and doesn't tell exactly what or how.
• I think it's a bit too much text to images ratio. Read any usability study and you'll see that people's eyes are automatically drawn to pictures (and their caption) and to numerals.

Why not #4:

• Same reasons as #3
• The bullets represent different things.
• The copy need some work. E.g. "We offer our clients a new way to buy or sell homes". How many people are selling multiple homes? Also, only your clients get that benefit? What do I do to become a client? The phrase could be replaced with "We offer you a new way to buy or sell your home".

Why I choose #1:

• The images and the text *do* correlate
• the cartoon images make it lighter and unexpected. Pictures of people/family is what you expect on *every* real estate website. Show you're different.
• Overall, it meets a lot of the criteria from "Made To Stick"
  

What would I change to improve #1:

• The phrase "Noe Valley or Boston or San Jose or 98105 or Baltimore" below the search box has two problems. First, it looks like you can enter only 5 options and that is conflicting with the watermark inside the search box. Second, it's missing the "e.g." or "example:" in front of it. People are not that dumb. Just pick a city name or zip code and it's enough. "Example: Seattle or 98105".
• The images should have number to them, to make sure they are not options but a sequence of events. I participated in a usability study about 4 years ago while working on MSN that showed how people can easily get confused if items like that are not numbered.
• Finally, the "Learn More" button is the call to action, it should be way more prominent. The problem here is that is red, and making it to big might blind someone.

    Now, if anyone from Redfin is reading this, ignore everything I said above and just do an A/B (/C/D) test and get the answer that you need.

 

    Finally, after a long, long wait Alexa has released it's toolbar for Firefox 2:

 

http://toolbar.alexa.com

 

 

 

 

    If you read any news website right now you probably already know of the accident in Sao Paulo, the city I was born.

 

    The airport the accident occurred is right in the middle of the city. This is probably one of the largest accidents in that airport history because it affected cars and people on the ground as well.

 

    About 12 years ago another accident happened when an airplane was taking off killing about 120 people and it hit many building and houses. Amazingly enough, nobody was killed or injured on the ground.

 

    The sad part of an accident where you live or used to live is that there is always a friend or a family member that lost a friend or a family member on the accident. The world is very small.

 

    Yesterday morning I had my first golf lesson from Paul. We went to the range to get some of the basics. The biggest lesson for me was that I was seriously underestimating the complexity of the game. There are so many moving parts in a single hit that is pretty hard to play it well until you get most of those into the automatic mode.

 

    Next time we will probably do a 9-hole course. I think it'll take 2-4 more attempts before either gets interesting for me and I fall in love with the game or for me to give up being any good at it.

 

 

    Last week I put out a call for Guest Bloggers on the Seattle 2.0 site. It was more than well received and we have more than 10 people lined up to write relevant blog entries for entrepreneurs and startups.

 

    Our first entry was just posted today: Naming 101 by Christopher Johnson.

 

    Christopher is the founder of Phrasetrain as well as the blogger behind The Naming Inspector. Recently, I saw a great talk he gave at Ignite Seattle, check it out as well.

    Before I even started Sampa, one of the hardest to forecast line item on my projections was support cost. What is the bell curve of customer support (emails+calls/month)? How long does it take to handle each request? What is the average distribution based on the number of months a customer has been with us?

    I'm only talking about Web 2.0 services focused on individuals, since this is what Sampa is about. There is also a big difference between paid and free services. For now, Sampa is free so the customer's expectations for support are lower.

    On average, we are getting about 100-150 emails per week. A rough distribution of the emails is this:

• 5% praising the service
• 20% asking how to cancel the service (this is specific problem to us due to some UX confusion)
• 25% asking good question, usually indicating a problem on the system or something that is not easily found.
  
• 25% asking very very basic question (how do people find my blog? how do I upload a picture?)
• 20% are questions or comments that are incomprehensible. Seriously. E.g. "How do I add a site to my site?", "How do I upload my blog to my photo album?"
• 1-3% people ranting how we suck.

    Overall, I think the number of emails we receive is well below what I expected when I started. Right now, we sign up more than 1,000 new customers per week. Getting 100-150 emails per week is pretty low.

    Now, there is a very interesting distribution of customer emails. Some customers send 5-6 emails in one week, and those are the "expensive" customers for us.

    So far, it hasn't been that bad, but now it's start to becoming noticeable. It's taking me 30-60 minutes per day handling customer support. I think the moment it takes 2 hours per day, then it'll be a huge cost of my time.

   

 

    Great event coming in two weeks. Hurry to sign up since they only have 200 spots and 50 are already filled up just a few hours after the announcement.

 

    You can read it about at John Cook's Venture Blog and Redfin Corporate blog.

 

    Sign up: http://thenakedtruth.wetpaint.com/

 

    Seattle 2.0 is looking for people to write guest blog posts on topics from Marketing, Business Law, Recruiting, Design, etc.

 

    http://seattle20.sampasite.com/blog/Do-you-know-anything-about-start.htm

 

 

 

    We are already in 2007 and Sampa databases are running on SQL Server 2000, a 7 year old product!!! The rationale is the one that every business should use: It works, so why bother upgrading?

 

    The answer came in two batches. First, about 6 months ago while working on our Geo-location database, I hit some walls with the size of the database (25 GB). But the final blow came last week with our stats database.

 

    We collect about a gazillion datapoints from each website and each visitor of each website. That generates almost 1 GB of data everyday. Of couse this is raw data, useless unless processed. We have our own batch processing system that parse all this data, aggregates and save a portion of it to files and another on the database.

 

    Every once in a while, the processed data either gets corrupted (because the logs were not copied properly overnight), or we decide to change our definitions of Unique User, Page View, etc. In both bases we need to reprocess all data back to May/2006, when we started collecting data. That is about 25GB of raw logs.

 

    All the logs must be processed in less than 24 hours otherwise the day rollover causes data corruption. Currently, reprocessing all stats is taking about 10 hours. That is a pain because if anything goes wrong, I pretty much have to wait until the next day to apply the fix and try again, since the job cannot start between 7AM and 5PM because it will hit the UTC midnight (5PM PDT).

 

    All that to say that we attacked the problem in three fronts:

• Optimize the code to use multi-threading to process;
• Bought a new server w/ 2-procs (dual core) and RAID 1 (mirroring) that hopefully will improve read speed.
• Upgraded to SQL Server 2005

    SQL 2000 is great, but is old. The UI is old, the functionality is old, the SQL Query Analyzer is the Notepad of database management. I'm hoping that Microsoft did a great job on the UI and perf of SQL 2005.

 

    But most of all, I hopde the migration is smooth. This would be a really bad time for a painful upgrade experience.

 

 

 

 

 

    No question there are quite a few smart people on Google. But it looks like they skipped the 6th Grade geography class since they think Mexico is in Central America.

 

    From Google Analytics:

 

 

    I'm in desperate need of a new cell phone. My Sprint PCS HTC 6700 is dying. Some people say it's dead already and I don't want to admit to it. I lost the touch-screen pen and the back cover for the battery. The phone doesn't ring anymore, only vibrates and the battery doesn't last more than 8 hours -- in standby!

 

    My contract only expires in September, so I have two more months before being free again or pay an extra $200.

 

    This 4th of July I played with the iPhone of a friend. Oh... my... God... That thing is incredible. It really redefines the personal assistant realm. I'm talking about the pipe dream we had on the 80s about a single device that would do everything for you and would become an extension of your life. The iPhone is it.

 

     But I'm not going to buy one now, and it boils down to some pretty serious limitation of the current version, which probably will be addressed in the future:

 

1. AT&T: I used it before. Coverage is not great on my house, the Internet speed sucks and the customer service of AT&T...
   
2. No Exchange Server synchronization: I don't like to install software on my computer to synchronize with my device. I like all synchronization to happen over the air, after all, why do you have a wireless connection on that thing? More importantly, I want calendar and contact synchronization.
   
3. No Terminal Server: Ok, I'm a freakin' power user. I guess this is well beyond power-user, but from time to time I need to connect to one of our servers to perform some small maintenance. It's not like I can't find a PC nearby, is just that I don't want to have to leave the "Children's Museum" when I'm there on a Sunday with my kid to make sure a server is running ok.
   
4. Version 1: This is a first generation product. It will have problems that Apple has not foreseen. It will have hardware malfunctions, it will have software issues. I'm not so much interested in testing the product for Apple. I just want it to work.

    I guess I'm going to wait until 2009 for the iPhone 2, or the iPhone Nano (much smaller) or the iPhone Shuffle (makes random calls for you).

 

 

 

    So, you’ve decided to join a Startup and live the life of high risk, high reward. The biggest problem of leaving a large and established company to a small, starting, still being defined company is the lack of understanding the basic differences. Yes, a salary of $100,000 is more than a salary of $90,000, but numbers are very misleading when a lot of other things are taken into consideration.

 

    This is a list of 10 things you need to know before you join a Startup.

 

 

1) Do you love the product?
    A startup is almost like a cult. You have to believe in it. There are many things to believe in, but four are the most important: The problem exists, the problem matters to people, the product/service is a good solution for that problem, and, last but not least, this startup is uniquely positioning to create the product/service (either because it’s early to market or because it has lots of good people with industry expertise)

2) Do you understand the history of the company?
    Who founded? Why? When? What happen between then and now? This is very important. The vast majority of businesses are legit and done by good people, but some are business built by con man or woman. Make sure the founder or the CEO doesn’t have a problem with the law, he isn’t being sued by his former employer or by a federal prosecutor on some SEC fraud case. Something like this would be the kiss of death for a startup.

3) Do you add value to the company?
    Desperate recruiters/managers eager to fill in open positions will tell you whatever you want to hear about how valuable you’ll be to the company. Are you really? Understand how much of your knowledge and skills can be used on the company. Be careful with “learning on the job” because that only works on large companies that can afford to have you learn how things work for months on. On a startup you have to be productive on week 2, otherwise you’ll be a short term liability and startups cannot afford that.

 

4) Industry experience or technical experience?
    Same topic as #3, you need to know why they are hiring you. Is it because you have experience on the market/segment they are on, or is it because you have technical experience that can be applied across segments? In the rare case it is because of both, you have a strong negotiation power with the hiring company. If it’s neither…

 

5) Why are you going to a startup?
    The differences between working on a large company and a startup are tremendous. I could create a table comparing 20+ criteria, but it boils down to one thing: Your personality. Some people like the stability, the internal support, the predictability of a large company. Some people like the surprises, the struggle and the impact they have on the product on a startup. If you don’t know which one you like best, you should try both. And, in both cases, each one (the startup and the big company) really value your experience on the other side.

 

6) What are the benefits?
    Don’t expect BigCo benefits at your startup. No matter how you look at it, the cash and non-cash benefits will be less. You don’t join a startup to have a higher salary, better health plan and more options on your 401(k). The benefit is the equity (a.k.a. stocks). The risk-reward is much higher a startup. Nobody expects Microsoft stocks to quadruple in price over the next 3 years, but your startup can be worth 25 times more in 3 years than it is now.

 

7) How much cash in the bank?
    It is a bit awkward asking how much cash a startup has on the bank, but is a fair question. A fairer question would be to ask how much runaway (or how much time left) the startup has before need to raise another round of money or becoming cash-flow positive. You probably don’t want to join a startup that has only 8 months of cash in the bank, even if they say they are just about to sign a new round of investment. It’s also unreasonable to expect a startup to have 3 years of runaway.

 

8) What is my equity stake?
    Now you’ve got this offer to a startup and everything looks great and they tell you’ll get 200,000 stock options. Holy s**t! Two-hundred thousand! Can you believe it? But wait… What 200,000 means? Does it mean 1% or 0.001% of the company? That is the answer you should know. How many outstanding shares and granted options there are, and how many you have? From that you’ll know if you got a great deal or a lame deal. Just notice that the higher the number of people and further along a startup is smaller your equity stake will be. Don’t expect to get 1% on a 100-person startup (too high) or 0.1% on a 3-person startup (too low).

 

9) What is the most likely exit?
    What is an “exit” you might ask? Thank you for asking. An “exit” is startup-speak to when you convert your equity (stock) into cash. There are basically two exits: IPO or acquisition. You probably know what both mean so I won’t explain. The key is to know the value of the exit -- the market cap of the company at a time that you can convert your stock into cash. If anyone figures out how to compute that he would be a millionaire since there is more art than science on the case of startups. But think like this: if you are working on a gadget manufacturing startup, you should look at similar companies that had exits on the last few years. If the range was between $20 and $100 million, there is no reason to think your startup will be $1 billion. Now, if you think the exit will be $100 million, and you’ve got 0.5% of stocks, you know the maximum you’ll make is $500,000 (if you were fully vested).

 

10) What are the escape routes?
    One of the biggest excuses my friends tell me about why they don’t join a startup is because it could be out of business in a year or so and they would be unemployed, to which I scientifically reply: Whaaat? If you work at Microsoft there is zero chance you wouldn’t be hired back. At a higher level and higher salary! Unless you are about to be fired from Microsoft, then you’re right, you’ll be unemployed and probably you are the one that sunk the startup. The fact is that unless you are working without pay, you’re getting a salary and benefits, and you are learning a significant amount about things you’d never learn at a big company. So, the only thing you are really risking is the equity, to which you either get nothing or quite a bit of dollars. BTW, how’s that Microsoft stock option you’ve got over the last 7 years treating you?

 

    Got any questions about joining a startup? Send it to me I'll augment the list.