Projects

Amitt's Space

GDC 2011 Presentation Slides (+Extras)

by Amitt on Mar 27, 2011
Thanks to everyone that came out to my talk at GDC 2011. As requested, presentation slides are now available at the GDC Vault and are also embedded below. In addition, I also have included slides from my GDC 2010 and Web2Expo presentations.





Porting Competitive Advantages

by Amitt on Oct 18, 2010
Porting Competitive Advantages After attending startup school this year, and seeing the variety of different startups presenting, I was trying to break-down the difference between some of the successful but smaller startups and the 10B+ market-cap companies.

One difference that I noticed was the portability of competitive advantages to new problem sets. The large companies solved a problem early on and then kept finding new ways to leverage their solution to grow and grow.

One way to visualize this is to imagine a wrecking ball of momentum that can break down barriers of entry into new markets (or as MML co-founder Zao puts it, a "wedge"). This momentum can be so large, in-fact, that it can completely disrupt existing markets beyond just overcoming the barriers of entry.


Google


Google is a good example of this. After building their initial search product they developed a skill in being able to take large datasets and analyze them in a way that provided useful information to the user. Take that approach in search and apply it to email, news, maps, website analytics, etc. and you have new types of businesses where you have a pre-existing competitive advantage.

Let's specifically look at Gmail. There were several large webmail providers before Gmail came out. While the massive free space proposition was one of the appealing factors to users, I'd argue that the rethinking of the webmail user experience is why most people use Gmail today.

Google leveraged their core strength of being able to analyze large amounts of text data to improve the core user experience of reading email. For instance, they provided better spam filters (cross-checked against their index of sites on the web) and email subject aggregation into threads using their ability to compare content.

Facebook


Facebook has a large set of personal data and social relationships that can then be applied to different problems.

There is a hypothesis that everything is better with your friends. So many activities are performed with friends that Facebook has a near limitless number of ways to use their data to improve existing products:


  • Photos are more interesting when you are able to see exactly who is in the photo.

  • Events are easier to plan when all of your friends names and contact info are already readily available.

  • This even works offsite on sites like Yelp, where it is useful to know where your friends eat-out and what they recommend.


The success of the Facebook platform is a testament to the fact that there are more opportunities than Facebook can even go after themselves. So by opening up the data, they are able to crowd-source the development of wrecking balls and then cherry pick the best ones to go after themselves.

The rapid growth of the valuation of Facebook reflects the fact that they have a huge upside if they keep executing on their vision of making everything social.

Quora


A more recent example. Quora has yet to use their potential leverage to enter a new market but it is possible. Quora is building a community of experts and a Wikipedia-esque DB of Q&A. If they manage to get scale and grow to expand to every topic, they can then use this data to provide user & expert generated answers to a variety of topics.

Similar to Facebook, they can now enter new markets using this data. Imagine, for example, a travel site where the information is structured using common questions people may have regarding a given destination: What is the best place to stay in Paris?, What should I see in Paris?, What are some good but cheap Parisian restaurants? and so on. They could then monetize such a site using AdWords style advertising since anyone using the site to find answers to these types of questions has some purchasing intent already. They have a leg-up on existing travel sites, as well, since they have a system of aggregating information from several sources into a definitive "answer summary".


Other Examples


What other examples can you think of where companies have used a solution developed originally for one problem and spun it into a new disruptive business? Share your thoughts in the comments.

Hiring "Smart"

by Amitt on Oct 06, 2010
At my last few jobs we used a mish-mash of semi-unique languages: UnrealScript at Epic, PHP/Flash combo at Zynga, and so on. It's rare that someone possesses experience in every technology the job requires of them so we needed to find a different way to hire.

In addition, if you're at a hot startup experiencing rapid growth, more likely than not, your required set of skills is going to change. In order to sustain that growth, you need a team that can grow with the organization and adapt to the new challenges the company faces within its lifetime.

So now when I evaluate candidates, two of the things I look for are:


  1. 1) Can they learn?

    Note that this is different from the traditional "Are they smart?" question popularized by Joel Spolsky1. There are a lot of people out there that are incredibly smart and talented but unable to pick up new skills or refuse to challenge their conventional way of thinking.

    This applies to efficiency as well, not just general knowledge, since a person that can learn can learn the "tricks" to become a more effective engineer.

    (As a side note, this can also apply to organizations and their ability to adapt to changing market conditions, more on that in future post)


  2. 2) Are they fun to be around?

    When training is such an important part of inducting someone into an organization, they need to be easy to get along with to make the experience rewarding for the TRAINER as well.

    If the experience is not enjoyable for the trainers, they'll be less effective in their training and not take ownership of actually making sure that new hires are getting the attention they need to succeed.

    Making sure that the folks we are bringing on are a cultural fit and fun to work with helps mitigate this issue.


Turns out this approach to hiring works surprisingly well (for us)

When we are scaling up teams at Zynga, we concentrated on finding folks that had a solid foundation in Computer Science and were willing to work with us to learn something new. Normally candidates would either know Flash or PHP but rarely both (and sometimes neither).

We would work with them and invest time to fill in the gaps in knowledge (through boot camps, code reviews, pair coding, etc.) so that they would become better developers, a cost we gladly paid to ensure we had a highly specialized team ready to take on the unique challenges we faced.

It's a win-win for everyone involved:


  • We get talented developers that are more effective at their jobs

  • The developers grow in their careers and acquire a broader set of experiences which will hopefully set them up for the next set of challenges the company would face



Treasure Isle Released!

by Amitt on Apr 06, 2010
Treasure Isle Released! Zynga just released our newest game, Treasure Isle! A lot of the lessons we learned from creating FarmVille went into this game and I'm super excited to see how people receive it.

The goal of Treasure Isle is to explore islands and collect treasures. By completing sets of treasures, players are rewarded with prizes that they can use to deck our their own island paradise.

Check it out here

Press and more info on Treasure Isle:

Treasure Isle: Everything you need to know
Zynga Gets Adventurous with New Facebook Game Treasure Isle

Rapidly Developing FarmVille Presentation

by Amitt on Mar 30, 2010
Slides from my GDC presentation are up on GDC's vault now if you
click here.

There's also a video presentation for those with access to the GDC Vault available here.

--

Update (embedded slides):

GDC 2010

by Amitt on Mar 10, 2010
GDC 2010 Had an awesome time presenting on building and scaling FarmVille yesterday at GDC, thanks to everyone that came and to the advisory board for allowing me to present!

For those asking for a recording of the presentation, that is up on GDC's Vault and will be available through them.

If you're at GDC this year, hit me up, will probably be there Thursday.

Building Big Social Games

by Amitt on Dec 18, 2009
I recently had the honor of presenting to a group of developers in Seattle at the Social Games Seattle event on how to build big social games. The talk focused on the design and technical decisions we made with FarmVille to support a massive audience.

I've included the slidedeck I used below. There were a few points which were talked about but not in the deck:


  • It is possible to release too much content at once because it overwhelms players


    We frequently throttle releases to prevent information overload. Furthermore, an increase the amount of content per release also increases the burden to test all of the new content. Usually, this results in decreased test coverage on any given feature, increasing the risk of a serious bug for a release.

  • Players look forward to new updates


    When you release updates to your game at a consistent interval, players come to expect those updates almost like they expect a new episode of a tv show each week. There's an impulse to keep coming back to see what new things have been added this week.

  • Invest the time in proper code architecture and development practices


    Building sustainable games is a marathon. If you hack together your base, it will make it much more difficult to adapt to evolving design requirements and release consistent updates later on in your product's lifecycle.

  • The iframe trick referred to in the last slide involves embedded an iframe in your canvas page footers. This iframe basically "pings" a php page that then performs FQL statements and stores them in memcache before the game begins loading

New look

by Amitt on Aug 24, 2009
Finally got around to making some much needed changes to this site.

I plan on using this space to showcase whatever happens to be on my mind when I feel like writing. In addition, it'll hopefully serve as an archive for the random projects I'm always working on. To start off, I put up all of my old projects (at least whatever I have available on my backup CDs).

I also took the task of having to redesign the look for this site as an opportunity to learn Python and mess with Google's App Engine. I'm somewhat impressed at how streamlined Google has gotten the process of creating and deploying an app. Though there's an un-shakeable feeling that there's a lot of magic going on behind the scenes and as soon as you deviate from their intended example usage pattern (a blog) things aren't as smooth.

About

Amitt

My name is Amitt Mahajan, I'm a entrepreneur and developer living in San Francisco. I love meeting new people and discussing technology, startups, and everything in-between. Feel free to contact me or subscribe below.

Projects