Archive for the ‘Entrepreneurship’ Category

Tips to improve the front-end performance and browsing experience

Monday, August 31st, 2009

The other day, I read in Gigaom a post about web experience, performance of web applications, and tracking tools to improve it. It was very appropriate because at Pixable we are currently focusing our efforts on improving not only the usability but also the front-end speed and eliminate the current slow browsing experience. The blogger was Yahid, MySpace’s client-side performance architect (so I guess he knows what he is talking about :-) ). Quoting the post:

In our era of immediate gratification, creating a quick and responsive web experience is often as important as the product or service you’re trying to sell. Poor performance can frustrate users to the point that they will leave your site. Recent studies show that on average, 20 percent of a page’s load time is spent on the back end as the server processes and generates HTML, while the remaining 80 percent is spent on the front end while the browser tries to load page assets and render the HTML. This rule applies to any size web site, whether it receives 100 or 1 million visitors a day. Keeping in mind this 20/80 rule, we at MySpace give as much attention to the front end behavior of our pages as our back-end systems.

The article talks about the typical bottlenecks and recommends some tools to track and improve the speed and bottlenecks like HttpWatch, Google’s page speed, firebug, AOL’s pagetest, MySpace’s MSFast, Fiddler, and IEInspector’s HTTPAnalyzer. Through the post I also found this great presentation about designing Fast Websites (here the slides), and this repository of information (see this best practices for example).

Some friends in the Consumer Internet space gave me their view about the tools. Mario from Asesorseguros.com recommended me Firebug and Jeff from Polleverywhere.com recommended mainly Firebug, Fiddle and Google Page Speed

Firebug

Pixable at the World Innovation Summit

Friday, July 31st, 2009
HITlogoAt the end of June, Pixable was present at  the World Innovation Summit (WIS). We attended the WIS event because Pixable was selected as one of the top 20 global startups with best growth potential. We went in representation of MIT.

For three day, gurus, inventors and entrepreneurs met and networked to learn the latest trends in innovation and its role as the way out of the global economic crisis, to share strategies to face the challenges ahead and drive the business projects with great potential. Among the plenary speakers I saw Rosabeth Moss Kanter, Ray Kurzweil (most famous inventor and futurist in the world), Gary Hamel (world most influential business thinker), Vijay Govindarajan (from Tuck) and Michael Eisner (former CEO of Walt Disney). As an inventor, Ray Kurzweil has been called the new Thomas Edison. I recommend listening to any of his talks or reading about his predictions of the future, including his famous book “The Singularity in near“. It was fascinating to hear his predictions about how medicine is transforming into an information technology, which by its nature advances at an exponential rate implying “radical life extension“.

The World Innovation Summit also featured the first edition of the Global Entrepreneurship Competition, an international early-stage tournament that called for start-ups from around the world with high future growth projections, in total 28 projects from 21 countries out of 10,000 initial submissions. This was a pioneering initiative bringing together entrepreneurs and investors from around the world at the World Innovation Congress HiT Barcelona.

Pixable was the startup with highest number of votes for the audience favorite award with over 350 votes. Thanks to everyone for voting! The best part of the congress was the networking events and the PR coverage (together with Ray’s talk). We were interviewed by reporters from the BBC, CNN, or El Pais. Now we have been nominated to the World Economic Forum Davos Technology Pioneers 2010. The PR through the event has helped us to start early conversations with very interesting partners.

HITbarcelona
HITbarcelona2

Why do we use cloud computing and Amazon Web Services?

Wednesday, July 29th, 2009

At Pixable, we use Amazon Web Services (AWS), an infrastructure web services platform in the cloud. Our Pixable service is accessed via the browser and the user does not have to download any software. Some startups are not aware of these type of services and we think that we can share some basic information to help them.

With AWS you can requisition compute power, storage, and bandwidth, and a suite of elastic IT infrastructure services as your business demands them, and pay as you use these (utility billing). This is the basic definition of cloud computing, with dynamically scalable and virtualized resources provided as a service over the Internet without having to babysit the servers.

Amazon Web Services was the perfect fit for our application. Our business is extremely seasonal, with high peaks on dates like Mother’s Day, Father’s Day, Thanksgiving or Christmas and we cannot afford acquiring high capacity servers in advance to absorb these peaks of demand, and then use less than 5% of their capacity the rest of the time. Moreover, for us it was difficult to anticipate the exact demand. Finally, in our business, we create value every time computing occurs (because it means people are creating new content). Therefore, we love variable costs (or OPEX) vs. fixed costs (or CAPEX) because those happen when me have a transaction. Starting a company with only variable costs is what entrepreneurs want.

The general benefits of AWS are

  • Pay per use model with no contracts or time commitments
  • Instant scalability (elastic infrastructure)
  • Reliable/Redundant/Secure
  • Simple to deploy and manage
  • Most services accessed via simple APIs
  • Amazon experience and commitment

Overall, AWS provides us peace of mind in terms of IT infrastructure  (security, scalability, etc.) avoiding CAPEX in expensive hardware so we can be more focus on the business side, the web product (programming new features, usability, etc.) and the physical product at Pixable.

In terms of the specific configuration and programming languages, Pixable is entirely programmed in LAMP (Linux, Apache, MySQL and PHP). We also use Flex (for the album creator) and AJAX. The servers are configured in a 3 tier architecture to be able to do horizontal scaling. In particular, our web application has a web server, a database, an interface with Facebook and a powerful album creator (providing image handling and modification, algorithms to mix a photo book automatically, etc) which does not require any software download (fully programmed in Flex), therefore, consuming a lot of resources. Finally, we use a lot of computation for rendering the images shown in the browser so they are ready to be printed. To host this platform, we use EC2 for the computation (web server, MySQL, rendering, album creator, PHP, and link to Facebook content) and S3 for the data. For EC2, you will need various instances as you grow, so you can separate the tasks in various instances (the Apache server, the rendering, the load balancer, etc.).

I want to clarify that we use AWS because it is the cloud service that has been in the market for a while, but other companies like Microsoft  (Azure), Slicehost/Rackspace, or Google have started offering similar services and it is not clear which one is better (it is like mobile phone plans, which are all almost the same).

Oher cool companies/startups with similar IT infrastructure requirements that also use AWS are Smugmug, Pixily, Animoto, Netflix or Justin.tv.

Given the seasonality of our service (and rapid growth), when we observe that we have peaks of demand, we start new instances, only while they are required (with AWS you can scale up and down in an hourly basis). It is utility billing to the extreme, pure computing on demand with ulimited resources!

power_by_amazon

Pixable has new offices in Barcelona

Tuesday, July 21st, 2009
Pixable has new offices in Barcelona at the 22@ high tech district. We are located in the Imagina Building, right next to the Agbar tower, in the same building as Yahoo Research Europe, Bazar.tv, and the TV studios of Mediapro (Gol Television, La Sexta news, etc.) and also where “Vicky Cristina Barcelona” was made.
http://geographyfieldwork.com/Yahoo.htmThis is a transitory space for us, which we have for 1 year while we decide whether we continue having presence in Barcelona, where it is easy to find technical talent and digital designers.
This adds to the locations we have in New York City, Boston, Caracas, and Delhi! right now, Alberto, our CTO, is spending one month in Delhi.

Pixable has new offices in Barcelona at the  Technology District 22@. We are located in the Imagina Building (Av. Diagonal, 177), right next to the Agbar tower, in the same building as Yahoo Research Europe, Balzac.tv, and the TV studios of Mediapro (Gol Television, La Sexta news, etc.) and also where “Vicky Cristina Barcelona” was made.

This is a transitory space for us. We have it for one year while we decide whether to grow in Barcelona, where it is easy to find technical talent and digital designers. We have to thank the people at 22@ who have been very helpful organizing meetings with key people in Barcelona.

Now Pixable has headquarters in New York, and this new location adds to the locations (i.e., at least there is one team member or office space) we have in Boston (MIT), Caracas, Duke University, and New Delhi! can we be more global? Alberto, our CTO, is currently spending one month in New Delhi with our software development team located there.  Alberto, I hope you can cope with the weather!

edificimagina

Pixable at Global Entrepreneur Competition, Hit Barcelona 2009 (Spanish)

Tuesday, June 16th, 2009

Breakfast with Chris Hughes, founder of Facebook and New Media Strategist of Obama’s campaign

Saturday, April 4th, 2009

Gcplogo Today, together with a small group of MIT and Harvard students, I had breakfast with Chris Hughes, cofounder of Facebook. Chris joined General Catalyst Partners (GCP) as an entrepreneur in residence 2 weeks ago. The breakfast was held in a small room at the offices of GCP. Joel Cutler and Hemant Taneja, Managing Directors of GCP, were also at the breakfast. It was a very interesting 90 minute discussion in which I also had the chance to talk about Pixable. The founder of The Ladders (from Harvard) was also in the breakfast.

I like General Catalyst because it is one of the only firms in the East Coast with a very ambitious mentality even if it requires taking high risks. As Joel Cutler mentioned today, “different VCs do things differently. If you focus on breakeven too soon, it will limit the growth.”

Facebook-logo

Chris Hughes (born in 1983, only 25 years old) co-founded Facebook, with Harvard roommates Mark Zuckerberg and Dustin Moskovitz. He worked at Facebook for 3 years and most recently served as New Media Strategist for Obama’s presidential campaign and pioneered the use of online campaigning through the political social networking site My.BarackObama.com. Chris was featured as the cover story of April 2009’s Fast Company under the headline “The Kid Who Made Obama President; How Facebook Cofounder Chris Hughes Unleaded Barack’s Base- and Changed Politics and Marketing Forever“. He holds a bachelors degree in history and literature from Harvard, where he graduated magna cum laude.

The conversation was more about what we were doing and how we wanted to do it, than on his experience (although in the end, everyone was asking him about his experiences). Some good insights of the conversation:

  • About Chris’s experience at Facebook:
    • At Facebook, Chris focused on user experience and product design
    • Before Facebook, sharing personal info on the web was about dating sites. Now it is a different paradigm.
    • I asked about the challenges of managing older people with more experience and expertise. He said that you need to be surrounded by good advisors but you also have to trust your instinct. As an example, he pointed out that when they had the $1B offer to be acquired by Yahoo, older people just wanted to take it because they had lived the Internet bubble and burst; but the young ones wanted to make it bigger and thought that it was possible.
    • He left Facebook when there were 200 employees (now there are 900).
  • In the process of starting a company:
    • Identify inefficiencies in human interactions. E.g., recommendations of restaurants in the neighborhood, search of grad housing, etc.
    • Design a web product that can make the interaction easier
    • Think about how to make money out of it
  • Use your own experiences and pain points to identify an opportunity. And BE ARROGANT THINKING YOU CAN DO IT BETTER THAN OTHERS (he stressed this many times)
  • Look for opportunities in which the product demand already exists. Never create a demand that does not exist (e.g., never say “People may want more of this”). With Facebook, they saw that people were interested in “what are my friends doing?” but it was very inefficient to find out.
  • Again, focus on a product that solves your pain point. And make sure that there are a lot of people with the same problem.
  • Although you do not need barriers to entry at the beginning, think about them for the long term. For example, Kayak could have been copied in the beginning but they just focused on the product and now is when they have barriers to entry
  • About the board of advisors for your startup: look for genuine advising that complement your skills. They should be people that were helping even before becoming part of the board.
  • VCs make two types of investments: a) ambitious about the impact (and then they will figure out the final way to monetize it); and b) investments to make money.
  • There are a lot of opportunities to make money in other countries adapting successful ideas in the US. They gave the example of European Founders
  • He recommended us: Find a good idea and just do it; building a websites is easy and you can see what customers think. And later, write a business plan. You learn much more doing it this way than talking to VCs or experts (except when there is a big ecosystem or you are in a mature industry)
  • Go where change is happening in the world and learn from it
  • During the first year, focus on product product product. Everyone will tell you that what you are doing is wrong
  • As a founder/CEO, most time is spent hiring people (the most difficult part) and raising capital. Be always welcoming potential candidates.
  • Expertise is not important. You just need to suffer a pain point as a customer and have a vision and ambition.
  • There are 3 things that take a lot of time: hiring, product and monetization. As an early stage startup, you should focus on hiring and product development.
  • As an entrepreneur, you need to spend a lot of time with the startup but you should learn to prioritize how you spend your time.
  • About LISTENING TO CUSTOMERS: “PEOPLE ARE NOT GOOD AT EXPRESSING THEIR FRUSTRATION”. Listening to the customer is not about feedback emails or questions you ask your users. The best way to listen to the customer is THROUGH METRICS: heat maps, who comes back, who they are, etc.

Overall, a very interesting breakfast at GCP.

Chrishughes