Marcelo Calbucci

I think I published my email address a dozen times. This time I'll publish as an image to avoid email-scrappers from finding it and sending me tons of emails, but here it is...

 

    The week we were announcing Sampa closing I wrote and published a blog post telling the story from a personal point of view, meaning, how it affected my life. One thing to keep in mind when reading my blog posts, is that I write to keep it to myself. It's very powerful to go back 3-4 years ago into your blog and read what were you thinking back then.

    Three days after we announced, I spent 4 hours sitting in front of the computer writing 8 blog posts on the business aspect of Sampa. The title is misleading because I could never include everything there is to know, but I wrote things I think people would most interested in.

    Here is the list of the blog posts:

• Sampa: From Birth to Death (long version)
• Sampa: My first failed startup (short version)
• Part 1: You Don't Learn This At Microsoft
• Part 2: Early Mistakes at Sampa
• Part 3: The Market Noise & Facebook
• Part 4: Things We Did Right
• Part 5: Advice to Family & Baby Sites
• Part 6: Things I Won't Miss At Sampa
• Part 7: The Little Secret of Web Startups (guest post on TechCrunch)
• Part 8: Would I Do Sampa Again?

    And just to be extra clear, there are two important points: First, everything you are reading is from a guy that failed to build a business. Second, even if I had done everything that I thought was wrong right I still could not guarantee either a successful business (profit) or a successful exit for Sampa.

    I'll probably do a few more posts about Sampa on this blog, then move on. There is just so much dissecting you can do.

First of all, the original Sampa was meant to be a website builder solution primarily for small business. I even considered being a white-label solution for hosting companies. I think the small business solution for website is still somewhat broken. When people say they will build their website on top of Wordpress it gives me chills. Not because Wordpress isn’t a descent/simple solution to build websites, but because Wordpress is being considered a simple and effective way to build websites. Seriously, can’t we do better than that?

 

On the Family/Baby space, I also think there is lots of room to create a new and appealing solution. I don’t think anyone “owns” that space in terms of brand. If I tell you to give the first brand name that comes to mind when I say “baby website” or “family website”, you’ll probably won’t know, unless you are working on this space.

 

The problem with Family/Baby websites is not so much of a technology problem as it is of a marketing and customer acquisition problem.

 

There are 4 million new babies being born in the US every year, yet, there are just a few hundred thousand queries on Google to “create baby website” (and variants). That means people don’t even know they can have a baby website. Unless someone finds an effective way to reach those customers, you won’t have a winner. There are only three large companies, above them all, that have a finger on the pulse of the baby market, including names, emails and addresses. If someone figures out how to partner with them, this startup will likely emerge as the winner. What are those 3 companies? Toys-R-Us, Target and Wal-Mart. Together they probably reach half of all the new moms in the US through their baby registry, and if you own the “baby site” market, it’s just a matter of a year or two for you to own the “family site” space.

 

Some people will say that you should start at the Wedding site space, but I disagree. The wedding site builder industry is very crowded. There is no “central distribution point” you can reach many customers at once, and between a Wedding and a baby might be years.

 

As to answer the title of this post, I don’t know. Given the right pieces of the puzzle being in place I might do something like this again. I still think this is a good idea and the solutions out there are not good enough. It’s just a tough market to crack.

 

The part 7 of my series of 8 posts about Sampa just got published on TechCrunch. Check it out.

There are lots of good things I’ll miss at Sampa. Even my own family website which we used intensively for 3+ years (Site id 15, where the first 12 sites didn’t exist, 13 was Sampa’s blog, 14 was my personal blog). But running a startup is very stressful at many fronts and I can’t be happy to think I won’t have to deal with some things:

 

The Code

 

When I started Sampa AJAX wasn’t a known technology, there was no Cloud Computing, no JQuery, Firefox 1 Beta was just coming out. I had to build a lot of stuff from scratch. Over time, the code grew bigger and more complex. If we count today, Sampa probably has 500,000 lines of code, but over the last 4 years I probably wrote more than 1,000,000 lines. As you add, remove, change, fix, improve code it becomes complex and the architecture you designed 4 years ago it’s not ideal for the new needs of the system. I won’t miss adding features to the Sampa platform.

 

Systems Operations

 

I’ve been managing live services for about 7 years now. It sucks. Server goes down on a Saturday night and you have to figure out what the hell is happening, otherwise tens of thousands of users cannot read or update their websites. It’s very stressful. We never grew to a size I could justify hiring an Ops Manager/Systems Engineer. I just hate babysitting servers. It’s not like we had many issues, but once a week or once every two weeks. But is the stress of being on alert all the time. Only if you done that you know what I am saying.

 

Entitled Customers

 

Most customers are nice. Most customers thank you for the free service. Some customers are just spoiled brats. But, by far, the worst customers were those that didn’t take the time to read anything and send an email to support for the easiest questions ever. It’s right there! In front of you! Can’t you read?

 

Bugs

 

There is a pile of bugs I’ll never have to fix. Yay!

 

Spam

 

My Sampa email has an enormous amount of incoming spam. Mostly because I monitor default addresses like webmaster@, sampa@, postmaster@, etc.

 

Opportunity Lost

 

This is a tough one to explain, but the fact is that when you build a very complex startup and you have a large user base to maintain, you can’t pursue new opportunities. I remember when Twitter spun off from Odeo, and how I felt at the time about taking my lessons from Sampa and running a more focused-service. At Sampa we had the concept of single-purpose vs. multi-purpose products. Flickr, YouTube, Twitter, Blogger were single-purpose products. They did one thing (mostly). Sampa, Facebook, Office were multi-purpose products. They did many things. Multi-purpose products complexity is not the sum of its parts, but also the sum of each connection between each of the parts. In other words, if building a blogging platform costs 1 and building a photo-sharing costs 1, building a blogging and photo-sharing service costs 3. Yes, multi-purpose products are more complex, hence harder to replicate in terms of technology, but while at Sampa I look at all the things I built and how I could have put that energy at creating a dozen single-purpose products, like: Blogging, Photo Sharing, Document Sharing, Page Builder, Web Layout Editor, Family Tree, Event Invitation, Baby Milestone Tracker, etc. And there were dozens of ideas that never made into the product because it wasn’t a fit for a “family site builder”.

 

It’s the story stupid! If I had just four words to describe about a strategy or vision for a family website, “It’s the story stupid” is how I would characterize it. First and foremost, people create a family or baby website to tell others about their kids life. Everything else is peripheral (and sometimes irrelevant). If you create a site that allows them to enter a picture, a paragraph of text and the emails of a dozen people to send it to, people would love it. Oh, wait. That exists already. It’s called email!

 

So, stop looking at the dozen of websites competing for the keyword “create a family website” and start tapping into how people behave naturally. For each family picture uploaded to the Web and for each blog post about kids posted, there are 10 times more pictures and stories sent via email – maybe 100 times more.

Email has the key elements necessary for sharing a story: You can attach anything to it (pictures, video, audio, links, documents), it’s fairly private, it’s very easy to use and it has your address book ready for you. All you need to do is write the content.

 

The obvious downside of email is that it's not easy to scroll through it or to organize it, either for you or for the recipients.

 

Unless you have a deep integration with email, you are asking too much of your users. Remember this is not about the author only, it’s about the readers. If the readers start complaining they can’t sign in, or “where is my password”, or “I forgot the link”, the author will just abandon your service. In other words: make it very easy for the readers to consume/contribute the content.

 

People Want a Mirror

 

Customization is also important. The problem is that startups tend to overdo it on customization, allowing users to pick any color, font, background, spacing, boxes styles, etc. When you do that, you’ll find out people can create really ugly things. Limit customization to a few of the graphical elements, but most importantly allow them to add a page Header that has their picture. People just love to see their own pictures on their website. If you had a completely plain website, black-and-white and allowed people to upload one picture of 900x200 to be their header and nothing else, people would tell you have a pretty good customization solution.

 

Public vs. Private

 

Privacy is important – for some. First of all, don’t assume people think like you. Some people like privacy, some don’t care. That’s the end of discussion. You might decide to take the everything-is-public route, but you have to do knowing you’ll lose some users. On the Family/Baby space, I think the ratio is 2/3 demand privacy, 1/3 don’t care. Another thing to keep in mind is that you might be listening to your customers on this topic and they were a self-selected group based on the features you already offered. Meaning if you ask them if they care about privacy you might get an overwhelming yes or no, but that's not a representation of the population in general.

 

Facebook Story

 

You must figure out Facebook integration. That’s not to say you should build your service inside Facebook, which I think you should not. But there is an element of easy-of-sharing and ease-of-connect on Facebook that you must leverage. The biggest problem with Facebook is that it cannot reach the younger than 13-year-old because of the law, and it cannot reach the older generation because of the technology barrier. So you have to create a solution that is seamless for those consuming the content on the website, consuming it on Facebook or consuming it via email.

 

The only way to succeed with a Family or Baby web presence solution is to be multi-generational friendly.

 

I could share another dozen less important elements on being successful on this space, but it's all a guesswork because I wasn't successful.

This is a series of 8 blog posts that will describe more about my experiences building Sampa, my mistakes and what would I have done differently.

Sampa failed, but it didn’t fail because we did everything wrong. Some things we did right and I’m very proud of that. Some of those was on my blood, some I learned along the way.

Just Ship It!

The primary and most important thing that I did was to SHIP IT! Not a single person that I worked with directly or indirectly, can tell me we failed because we didn’t ship our product. I shipped good software, I shipped great software, I shipped bad software. The only reason we would abandon a feature was because we made a clear inexcusable strategic decision that feature was not on our best interest or should be traded-off by another one. That rarely happened.

Shipping is the result, and the two factors that led us to ship so often were: Focus and “no”. Focus meant when I was writing code, spec’ing, testing it or deploying I couldn’t care less about anything else. I was also good at saying “no” to myself and to others. I was able to quickly understand the difference between doing something “good enough” in 2 days or doing “the right thing” in 7 days. “Good enough” always wins on my book.

It’s Not The Code

For about 6-9 months into Sampa I was very much in love with the code and the platform I’ve built. But after the first Alpha and getting users asking some basic “how do I do this?” kind of question I quickly learned that your consumers and your partners couldn’t care less about your code. They didn’t care if it had 10,000 or 300,000 lines of code, if it was open source, if it complied w/ XHTML standards, if it used Tables or not on the HTML. They didn’t give a shit. No one gives a shit about this except developers. But we pat ourselves on the back every time we write well-documented, well-architectured and standard-compliant code.

So after that 6-9 months I just abandon my code and focused on how can I do feature X in the most efficient manner and please users. That takes a 90-degree turn in your thinking. When you care about users, you stop thinking like a geek and you start thinking about short and long term value. Don’t take me the wrong way, I still wrote shared libraries and well-compartmentalized code when I felt we would reuse it in a short period of time, otherwise I would just make the thing work. Later I learned that a lot of those concepts are called “Agile Development”.

Understand Your Limitations and Accept Them

If you really, really want to succeed you have to give away control to others. That involves everything you do today because there always will be someone better than you. Some things are easy choices. I quickly gave away all my control at being a CEO of Sampa. I also gave away all my ability at designing templates, homepages, etc., and many other things.

What we did really well was to learn that our team didn’t have the marketing branding skills to create a compelling user message. Like most tech startups, we all come from a tech background and that makes us a different beast. For example, for any geek “URL” is a simple term. When you tell a user that word they might freak out. I can see it right now someone reading this post and saying “that’s not true, nowadays everyone knows what URL is”. Alright, but if you are not like that go search your entire code for “URL” in a user message and immediately replace it with “link” or “web address”. Actually, when in doubt use whatever terminology Microsoft or Amazon uses (they test this stuff).

We Were Scrappy

Yes, we spent almost all of $1.35 million we raised (there are still some left). But that took about 24 months, which translate into a burn of about $50K a month. At one point we have 4 full-time and 5 part-time people. I won’t say that every penny we spent was worth spending, but if we did spend we thought at the time it was the right thing do to. Ok, maybe that’s an obvious and empty statement. But we didn’t have parties or morale events (which I regret a bit), we didn’t have a Wii, a Flat Screen TV, expensive chairs or desks. We didn’t sit in nice buildings with a view of the lake. We spent on things we felt would move the needle the most. Sometimes they didn’t, but unless we could plot a direct correlation between cash spent and a benefit, we would not spend the money.

Next post: The Little Secret of Web Startups

This is a series of 8 blog posts that will describe more about my experiences building Sampa, my mistakes and what would I have done differently.

I already talked about on Part 2: Early Mistakes at Sampa about being 12-months late to market. The problem was when I was launching Alpha and Beta versions of Sampa the number of startups launching Web 2.0 products was nearly doubling every couple of months. In 2005 and early 2006 there was one or two new startups launching everyday and you could read about them on TechCrunch, ReadWriteWeb and GigaOm. By late 2006 and all of 2007 there were half-dozen to a dozen new startups launching each and every day!

Building Websites, either for consumers or business is a pretty broken industry dominated by 1&1, GoDaddy and the tens of thousands of consultants out there. So, I’m sure I wasn’t the first to think about fixing this industry and at one point I had a list of 100 (!!!) startups playing on that space. From vertical solutions (for Homeowners Associations, Dentists, Families, Baby, etc.) to very horizontal plays like substitutes for FrontPage and Geocities. To top it off, there was the rise of blogging and photo-sharing services like Flickr. That’s a hell lot of noise.

Seriously, no matter how good your solution was, with this many players every time we talked about Sampa people would say “you are just like X”. And we were nothing like X – on our minds. That didn’t matter. What matters is the consumer perception. And the press/blogosphere perception is just representative of the consumer perception.

We looked deep into our strength and weaknesses, in what we had and what others didn’t and we had a feature that we implemented in January/2007 (and we patented it as well) that allowed users to add friends and family to a private site and when new updates were sent to these users they would just click on the link on the email and be able to see the photo-album, or blog post, or whatever, without ever have to enter a password (over-simplifying for the sake of brevity). It was very powerful! Users loved it. We loved it.

Over 2007 we changed our brand to be around Privacy and Safe Sharing. Although that branding should have been with us from the get go, it wasn’t easy to explain and people would go “but I can have a private site on Blogger”, or “I can share privately on Flickr”. Argh!

Then, Facebook came. At first we didn’t perceived Facebook as an immediate threat. It was mostly college (or recently graduated 20-something) and the geeks that used it. But they started to grow, and grow and grow and mainstream was using it. Because Facebook has most of your friends and quite a few of your family members on it, and sharing was by default private, it kind make sense for you to upload the latest picture of little Jimmy playing on the beach and everyone could go “oh, cute”.

Facebook is awful at creating a family place or a baby place on the web because it doesn’t give you *any* customization (I’ll talk more about why this is important in another blog post) and it’s cumbersome to tell people about new content you want to share.

Next post: Things We Did Right

This is a series of 8 blog posts that will describe more about my experiences building Sampa, my mistakes and what would I have done differently.

About half-way through Sampa existence I’ve made a blog post named Top 7 Mistakes At Sampa. It’s pretty hard when you look back at 4+ years of work and not to find hundreds of mistakes you’ve made, small ones, big ones and everything in between. At the end of the day, the only mistakes that matter are those where you believe would have made a difference on the outcome of the company or made the journey better. Here are the key mistakes I believe I've made:

You Need A Co-Founder & Tell Friends

The two mistakes I’ve made out of the gate (i.e., even before I started working on Sampa) were to not look for a co-founder and to be stealth about the product. I actually believe we could have a better Sampa, shipped faster and be more customer focused if more people were in-the-know and if I had someone to split the work. I needed someone that was either not a developer, but would have good business, marketing and design skills, or someone that was a better developer than I was so I could shift at learning the business or marketing (and hire a designer). And the Stealth part is a no-brainer. After I told people what I was doing around September/2005 (9 months into Sampa) instead of having a lot of spies looking into Sampa, I’ve got my own army of friends-spies looking at what else was on the web and sending me a lot of good info, connecting me with key people and sending critical early feedback on the business.

Timing Is Important

I will write about the noise on the market in another post and how that affected us, and a lot of that could have been avoided if I had started Sampa 12 months earlier. Startup is all about timing. Some people think if they ship a month earlier they will “win”, and that’s BS. But a year can make a huge difference. If you are too early, the customers and the partners are not prepared to absorb that technology so you better have a lot of cash to keep fighting. If you are late, you have dozens of other companies doing very similar (or substitute) solutions. And that’s a tough marketing/branding battle.

Paid Is Different Than Free

When I started Sampa the business plan didn’t include any free offerings. As time went by, and I was about to ship Alpha 1 (Sep/05), the Web 2.0 wave was starting to hit and everyone was offering everything free. Add to the fact that doing a paid offering requires quite a bit of work and would limit the growth rate of Sampa, I changed the business to be free. I somewhat regret that decision. When I was on my early-20s my boss and I were talking about Pizzerias in Sao Paulo (there is about 1 every 3 blocks in the city) and he said “they only way someone can open a new pizzeria and be successful is if he charges 3 times as much”. Sorry to say this to the TechCrunch-reading crowd or to Chris Anderson but there is a correlation between price and perceived value. Free is perceived as “not as good”. Even on the Internet!

Create A Healthy Win-Win Ecosystem

Now, probably the biggest mistake of them all was not to find partners early enough, primarily companies that would be the distribution channels for Sampa, the companies providing design templates for our users and much more. First of all, when I left Microsoft I swear I didn’t even know what “business development” was. When you work at the Mothership, you don’t do partnerships. You seat and wait for all kinds of companies to present you with offerings and you just give thumbs up or down. On Startups you have to work long and hard.

Next post: The Market Noise And Facebook

This is a series of 8 blog posts that will describe more about my experiences building Sampa, my mistakes and what would I have done differently. I already wrote from a personal point of view and now I'm writing from a business POV.

I spent seven years at Microsoft learning how to build large scale software. And by large scale software I mean millions of lines of code and software that’s used by hundreds of millions of people. There are just a dozen or so companies on the planet where you can have that kind of experience. It’s priceless.

The more Microsoft grew and the more the group I was in grew as well, the smaller my scope of ownership, control and influence was. I’ve always been curious about more than the engineering part of building a product. I wanted to learn about Marketing, running a business, user experience, strategy, M&A, PR, Focus Groups, etc. I couldn’t do it at Microsoft. The few times I tried to got out of my “zone” I was either reprimanded or shunt.

The first thing you don’t learn at Microsoft is Customer Focus

Then I went to do a startup to feel like I had purpose in life. You think after 15+ years building software, that building a tech startup wouldn’t be hard. I know the hardware I need, I know the software, I can built whatever people through at me. I felt like I understood pretty well customer focus.

The reality is that Microsoft is not a customer-centric company. It’s a engineering-centric company. Most people inside Microsoft doesn’t even know that. Actually, they probably don’t even know what I’m talking about. They will argue they are customer-centric because they do focus groups, they interview customers, they listen to support and to feedback on the Internet, yada, yada yada.

The second thing you don’t learn at Microsoft is Customer Acquisition

Between Jul/2000 and Apr/2001 I worked on a project called MSN SmartTags. For those that don’t remember, it was a web augmentation plug-in that every time certain words would appear on a web page a squiggly line (like the Word spell correction line) would appear under that word. Once you passed the mouse over it, a menu would appear with contextual links. For example, if the word was “Seattle”, you could see “Weather in Seattle”, “Travel to Seattle”, etc. The project never shipped, but it was supposed to come integrated with Internet Explorer 6 and Windows XP. The day IE6 and XP shipped, this product I built would be on the hands of tens of millions of users. In a few months being used by hundreds of millions and in a few years by about 1 billion people.

Let me give another example: Products at Microsoft which fail to reach millions or tens of millions of people on the first few weeks are considered a complete failure. No one, not a single soul at Microsoft thinks anything out of the ordinary about that.

Microsoft (and Google, and Yahoo, and IBM…) has its foot on the door of so many people and so many companies that it just a matter of embedding, integrating, adding, up-selling a product to another to guarantee customer acquisition. Contrary to the Federal government, I believe that Microsoft earned this right and it’s good for them. But it shields anyone at Microsoft at understanding the fundamentals of customer acquisition.

Even upper management at Microsoft doesn’t get it. They are probably still scratching their heads trying to figure out why Windows Live Spaces is not as popular as MySpace or Facebook.

My Plea To People Leaving Microsoft

If you leave Microsoft today I suggest you have a deep and long conversation with people building startups to ask them what they consider customer focus, how they decide to build what they’ve built and how they go about acquiring customers. There is a lot out there to learn.

Next post: Early Mistakes at Sampa

[If you don't have time for the long version read the short version]

 

Sampa was first conceived on my mind around 2001. The story is one of those “personal pains” kind of story. I’ve been building websites for family and friends since 1996. Every time I built a new website I’ve got annoyed by how much pattern there was on building a website, yet I had to start each one from scratch every time. My mother has a PR company in Brazil. I’ve built her first website in 97, a new one in 2001 and a new one in 2003. Between 2001 and 2003 I’ve spent quite a bit of time looking at PR companies websites. They are all the same, except by the copy, logo and colors, they all contain fundamental website elements: An about page, a contact page, a list of clients, case-studies, a news page, press releases, etc. So I thought it would be an interesting business if I could automate the process of building a PR companies’ websites. Then I realized there wasn’t enough PR companies in the world to make this an interesting business opportunity.

 

Between 2001 and 2003 the idea of a “PR Company Site Builder” sat on the back of my head while I tried to figure out how can I make this an a larger business concept. By early 2003 I realized that like PR companies, each industry had their own website pattern (yes, it took 2 years for me to realize that, but I had a more than full-time job at Microsoft). Not only industries, but personal websites, family websites and group websites, also had patterns. So I started looking into Churches, Dentists, PTAs, Homeowner Associations, and pretty much everything that wasn’t about selling something on the Internet. After writing a super-crappy business plan and code-named it “Sampa” (the nickname of the city I was born) I went selling the idea. Of all places, I started selling this idea inside Microsoft in mid-2003. I spent a 1 year trying to convince a half-dozen VPs/GMs to take on this project without success. The feedback was always the same “it’s a great idea and building websites is broken today, but we only do [package-software/client-software/biz-service/consumer-web] on my division”. At the same time I was working on MSN Search (now Bing). In early 2002 I had become the Dev Manager of MSN Search on a very small team. Once Gates, Ballmer & Co. realized search was worth something they invested heavily in bringing more people.

 

What attracted me to MSN Search when I went there in 2000 was the small team, lots of technologies and cutting edge on web work. But by mid-2004 MSN Search was already 150 people (about 45 developers) and growing non-stop. My scope of influence was diminishing (like everyone that works on a big team) and the pleasure of coming to work every day had turned into depression. So we took some vacation time and went to Europe. I remember sitting at the Musée d’Orsay in Paris and in an instant it hit me. I should just quit Microsoft and do “Sampa”. So I told my wife “When we get back to the US I want to talk about a new career” without telling her anything else.

 

We went back at the end of September and I started documenting my thoughts on this business. Then on a Saturday afternoon I told my wife I wanted to quit Microsoft and invest everything we had into this new startup. Silence. I don’t think that’s why she thought about a change of career. After the silence came her panic. My wife is not the risk-taker. She saw it as just too risky. It took her a week to absorb the idea and accept it. For two weeks I spent time talking with people who had done startups before asking them tips and ideas on how to do this, and at October 15, 2004, a Friday morning I went to my manager at Microsoft, Randy Kern, and told him I was quitting to do a startup. I’ve got a few offers to stay, but my soul was already sold on the idea. There would be no position and no amount of money that would’ve made me stay.

 

We were about to ship the first beta of the new MSN Search using our own indexing technology, so I stayed for 6 weeks while wrapping up my part of the project (mostly Instrumentation, AB testing and data logging). After Microsoft we took a month vacation in Brazil (in December/04), followed by a one month “warm-up project” (A social network for college students in Brazil… wait, isn’t that Facebook?). And in February/2005 I started working on Sampa.

 

I coded non-stop between February and August of 2005 at the office on my house. It was very hot during the summer to the point of being unbearable on that room, so I would code between 10am and 3pm, then take a break because of the sun and go back to code between 6 and midnight. In September of 2005 I decided to get an office space so I leased one and at the same month I released the “Alpha 1” of Sampa to about 75 friends. It was the first time most of my friends learned what I was doing. At the end of October I released “Alpha 2” and on December 31st, 2005 I released “Alpha 3”. I remember sitting on my new office with my pregnant wife while she was finding dozens of bugs per hour and I was fixing them. We left for a New Year’s Eve party around 6pm after I just released “Alpha 3” of Sampa.

 

On May 19, 2006 I dropped the “Alpha” phase and released the public beta of Sampa. About 20 friends signed up the first day. I’ve got some press coverage (but not much). The problem at the time was that I was running out of money. So, I started poking around the investment community and end up going to a meeting at Ignition Partners. I presented there twice and they passed on the deal. But I end up meeting Paul Gross, which was doing an EIR (Entrepreneur-in-Residence) stint. Paul, joined Sampa as a Chairman and investor in October of 2006 and in March of 2007 he became a full-time CEO.

 

If you go back to the original concept of Sampa, which was an easy-to-use Website builder for everyone and anyone you can see that it lacked market focus. That all changed when Paul came onboard. The first thing we did was to drop all the business-oriented features. Sampa was then converted into a consumer product. In 2007 we started to focus more and more the product into a “family website builder”, then to a “family web presence”, then to a “family personal private place on the web” (ah, the magic of marketing).

 

But there was a big issue: Our UI was built by an engineer (me) and had complexity written all over it. The business was not getting traction and every investor/advisor we talked to said we need to fix our website, our sign up process, our product. That’s a lot. Between 2007 and 2008 we released about 5 different homepages, each one making the message clearer, cleaner and the sign up process smoother. So we were getting a good number of sign ups (thousands per month), but we were retaining just a tiny amount of users. The product was very broken from a consumer point-of-view. From the get go, the product was not task oriented (as in “I want to upload some pictures and do a blog post about it”), but very object-oriented (an hierarchy of folders, with security, objects, drag-and-drop, etc.).

 

Starting in late 2008, with the help of our marketing firm, a great designer and a UX expert, we worked until March 17, 2008 when we released Sampa V2. A new, beautiful, streamlined experience to build a personal private place on the web. In a week we knew we had a product hit. Our engagement and viral numbers doubled, our churn was half and everything was great. At the same month we closed our Series A round of financing, raising $1MM from angel investors. Having worked on MSN Search for so long, particularly with lots of expertise in data-collection and analysis, I knew it would take months for us to really understand the changes the new V2 would have on the long term viability of the business.

 

In mid-2008 we hired an engineer to help us revamp our “stats database”. We need better ways to analyze all the data we were collecting to make better decisions. After some time, the data started showing its ugly face. Sure, the viral coefficient had double after V2, but the decay and churn were not under control. In other words, the business was organically growing at a turtle pace.

 

In September 2008, before the collapse of the banks in the US, Paul and I had a meeting where he laid out the sad reality: Sampa was not viable as is. That was probably the worst moment for me in Sampa’s history. He just told me we didn’t have a business (more on that on a separate blog post). We failed to achieve the traction necessary to have enough revenue to continue growing or to raise a second round of financing. It was hard on me. Sampa as a business was a failure because no matter how much we tried, it could not generate a profit at its current format.

 

So, we spent about 6 months pursuing partnerships with larger business that could bring either a large number of customers or a new business model, and preferably both. We found several candidates and several were very interested in what we had. One by one these potential partners started to fall off our whiteboard, because they decided to built in-house, or they acquired a similar solution to Sampa, or because they weren’t ready to do the deal.

 

On Friday, June 1st, 2009, our last chance was gone. The partner with the highest value to Sampa decided they are not ready to do anything on this space yet and this was the final blow to Sampa’s existence. We’ll be shutting down our servers for good in August (which will give our customers many weeks to export their content) and liquidating the corporation. That’s the end of the Sampa story.

 

It’s pretty hard to distinguish a founder from his startup. Oh, wait. Let me take that back. It’s not hard at all. It’s hard for the founders to detach themselves from their startup. It’s the “baby syndrome”. The way they see it, their startup defines them. It *is* them. I’m sure I don’t need to convince you of that’s pretty unhealthy in most cases.

 

I think for the first 12 months of Sampa that’s how I felt. Sampa was Marcelo. Marcelo was Sampa. Any attempt at telling me there was a problem with Sampa was a direct affront telling there was a problem with me. But them you start to listen, take in the feedback and soon you’ll be the first to tell people about the flaws of your own product and not feel they are your own flaws.

 

Sampa has been a rollercoaster on my life. Sometimes lifting me to new heights, sometimes dragging me on some dark tunnel for months.

 

When I started Sampa in 2005 I was completely unprepared to have a startup. Before I even left Microsoft I’ve read about 6 books on startups. I thought I knew exactly what customer acquisition, customer-focus, focus, user experience, conversion were. Heck, I’ve been building software for 15+ years, and learning how to do a startup hands-on couldn’t be that hard, could it?

 

The biggest impact Sampa has had on my life was to give me clarity on how much I just *love* the way startup works. Corporate life is pretty sweet, pays well and can be career-wise very rewarding, but it’s not me. Startup-life just works better for me.

 

It sucks really, really bad to fail. It sucks to have spent investors hard-earned money to find out there was no business at the end, a very expensive experiment.

Yes, the lessons I’m taking from Sampa are priceless. Brad Feld (a renowned investor) said on a blog post:

“…what makes a better CEO of a new startup – an experience entrepreneur whose last company was a failure or a big company executive with a stellar pedigree who has never worked in a startup? Give me the experienced entrepreneur whose last company was a failure 100% of the time… the dude that just came off a failure and is ready to go again is super-extraordinary-amazingly hungry for success”

That’s pretty much summarizes how I feel. I have many lessons from Sampa which I will share on others blog post.

 

And just to be clear, Sampa *is* a failure. It failed. Period. There is no other way to describe it. It can also state clearly that I failed to make Sampa a business. What I can’t say is that I failed in life. The way I see it Sampa was one of the many battles I had and will have.

 

What’s next? Startup, of course. I don’t know what shape or form my new lifestyle will be (yes, startup is a lifestyle, not work for me), but I can guarantee it will not be as a part of 300-developer team in a corporate office.

 

In case you missed it over the weekend, there was a fire at the Fisher Plaza building in Seattle between Thursday and Friday. That building is he house of a few Datacenters, like Internap and AdHost.

 

I can't say I'm an expert on Datacenters, redudancy, systems operations, etc., but I can say that I've been managing services and servers for about 8 years, at Microsoft, at Sampa and my own personal servers (yes, I actually have 3 personal servers).

 

I was reading Jeremy Irish post on the effect on Groundspeak and remembered one of the painful lessons I learned a while ago.

 

ALWAYS SET YOUR DNS TTL TO 15 MINUTES OR LESS.

 

If you don't know what it means, you can stop reading now.

 

The problem with DNS is that most DNS Servers come with the default TTL of 24 hours, which means that if you move your server to a different location or you move your site to a different server, some of your users won't be able to access your site for 24 hours. Now, this default of 24h has been around for about 20 years or so and it's just outdated.

 

All my domains have a TTL of 60 minutes or 15 minutes. I actually seen some hosting companies (like eNom) using a TTL of 3 minutes, which I think it's great.

 

The biggest justifications to have a longer TTL are usually two-fold: to reduce your DNS server load and to improve user experience by reducing latency and resilience in case of DNS failure.

 

DNS Servers can have millions of domains and server billions of requests per day, so the perf improvement on the server simply doesn't hold up. DNS is also the fastest protocol on the web (because it uses UDP) and there is almost no latency any user could ever notice. Not even if you are solving a dozen domains for the same page and finally, if your DNS is down *and* your backup DNS is down, it's most likely that your site is down as well (if both run on the same server).

 

I find it that smaller TTLs also help with all kinds of issues, like moving servers, adding servers, removing servers, etc. So, stop procrastinating and go check the TTL of the domains you own and change it to 15 minutes!

 

I also recommend an expiration of 24h and a retry of 1h.