Recognizing Deven

Product Graph

Matt writes a new update to his article about Like.com after talking to ThisNext.  He makes a few interesting points I'd like to comment on.

a) He agrees that Like.com is ahead in key metrics like revenue and visitors.  So the core of my argument that we are wining on these metrics is not refuted by him.

b) He states that ThisNext being venture funded doesn't have to worry about revenue right now so instead let's just assess their progress on the quality of their vision.  This is just crap.  Like.com is completely venture funded also and hence has the same standard.  Visual search is no small vision either.  The only difference is that our team has decided that  you need to get off of living on venture capital sooner rather than later.  In fact, thisNext launched 1 year (I think) before Like.com so they should be farther along the maturity curve.  If we hit our target of breaking even by the end of 2008 it will be two years since we launched Like.com to break even.  Eventually  Matt and others should all hold ThisNext to the standard of revenue and profitability.  If they don't eventually the market and investors will (especially with a downturn coming).

c) Gordon the CEO of ThisNext states that social shopping does make sense in the form of recommendations, reviews, and friend's opinions.  My data shows that this is less important in many soft goods categories than in hard goods.  Are you really going to read a review of that tie before you buy it?  Are you really going to read a review of dress pants before you buy it.  It is just not as important as it is for a digital camera.  ThisNext 1.0 didn't work (I believe due to this).  We'll see about ThisNext 2.0

Probably the most interesting thing he said is that ThisNext is going to try this new product graph concept.  I don't know exactly their definition of it, but in a way visual search has already created an algorithmic product graph.  We have shown what products are linked (via visual similarity) to each other and which match to each other.  There may still be room for a social product graph, but to date a few apps on Facebook have tried (notably Glimpse) and that hasn't work.

In the end time will tell.  If 3 or 6 months from ThisNext is still selling vision only (and not metrics), I will not be impressed.

One irony here is that there are probably other folks this next would consider closer competitors than Like.com and vis a versa. 

Like.com Live with GA now

Like.com 1.0 GA

It is finally out.  You can see it here.   This new release is the culmination of 1 year of tweaking, hard work, and learning from our entire team.  I sure it won’t be perfect but am still so proud of it and our team.

The parallelism of my personal and professional life right now is uncanny.  As I write this post I am on Sofia (my new 7 day old daughter) duty tonight.  She is just the cutest thing – I love her so much that I don’t even mind that fact that I’ve had to change her diaper several times in the last hour and gotten poop on my shirt already once.  I guess for love you’ll suffer through anything.
This new site represents our key strategy for the year.  Focus, focus and more focus.  We launched Like.com in Nov 2006 and immediately lots of folks asked us if we were going to follow with generalized visual search, with visual search for dating (which we know doesn’t work), or with visual search for stock photography sites.

We decided instead to keep focused on shopping.  In fact we didn’t just focus on shopping we focused on just on soft goods (clothing, shoes, handbags, jewelry, etc – things that are largely aesthetic and hence hard to linguistically describe – perfect places for visual search as they are places were text search falls down).  Finally, we didn’t just focus on soft goods we picked 2 categories of products and just focused on creating the best user experience possible and the best business possible in those areas. 
I remember the first time I did a startup I was afraid to focus this narrowly because it might mean that we became a niche and my visions of grandeur would not emerge.  Now I realize that you first have to be really really good at one thing to ever be great at many.  2007 was all about focusing on one thing for Like.com

This release embodies much of that learning:

New UI: It contains a whole new UI designed by Pedro our UI designer, an outside consultant, and Jacquie.   Jacquie has been screaming for a year that our site looks like Walmart from a fashion sense when instead it needs to look like Macy’s or Bloomingdales.  The nerds amongst us were a little slow to understand this, but eventually we realized they were right.  They chose a design direction where the product (being sold) is king and where the rest of the UI fades into the background.  They chose purple – a risky color and dark tones because that is what most luxury sites do.  Jacquie calls it the Like.com makeover from Ugly Betty to … well… you decide.

The 2nd change in this release is the focus on detailed features that really matter in shopping for soft goods.  We now let you choose the material type for many items in the left drill downs (try it in shoes).   Very few shopping comparison engines let you do this.  They let you choose megapixles for digital cameras but have not done this level for work for soft goods.  Tanvir and our new expanded feed team made this possible.

Third Baris, Navneet, & Diem used our visual and linguistic artificial intelligence technology to create a new set of style clusters.  We are the first shopping engine to have these.  You will see these on the top left.  They analyze your search results and visually group items into styles and let you select the style of shirt you are looking for.  This is a great feature that shows off the Like.com “beauty and the geek strategy”.  Jacquie developed the style ontology and Baris, Navneet, and Diem worked to analyze text and images to determine which style each item fell into.

Fourth, we are adding lots of new niche merchants. Not all of these are live right now but over the next few months we will bring many of them online.  One of the people who has just blown me out of the water over the last year is Mehul Nariyawala.  He has turned out to be one of best hires we made in 2007.  In fact, if you remember he found us and said he would work for free, we didn’t hire him.  Now he our Dir of Merchant Operations and is just driving things forward at a rate we have not seen before.  Together with Wil, Calvin, and Tanvir, Mehul has become someone I trust implicitly and someone whose leadership skills I wish I could replicate.  I love his sense of ambition and risk taking.  He is truly a rockstar.  Combined with Jillian our new rockin merchant sales person (she literally has called hundreds of new merchants in just 6 weeks) and several great new partnerships with feed aggregators, we have gotten a whole host of new merchants on the site.  Again we focused on building out the categories even deeper instead of broader.

Fifth, we have the first color search that really works.  Color search reinvented if you will.  Yes everyone else lets you choose the color of the item from a color wheel only to realize it doesn’t exist in that nuanced shade of purple.  Like.com is the first to only show you the color swatches of the colors available.  This is the brain child of Jacquie and Baris and implemented by Robert, Diem, and Eric Lee.  Try it and let us know.

Sixth, we now cross-merchandise matching shoes for a number of clothing items.  I think it is fair to say this area is in alpha (the matching is not that fashionable yet).  However, we are the first site to automatically show you the “visually” matching shoes.  It is a simple feature we plan to expand more.  A recent addition to our team is Marissa Goodman.  Marissa was a clothing designer for Old Navy, Gymboree, and Espirt for years.  We are literally creating a whole cross-merchandising algorithm based upon her expertise.

Seventh, we have updated out core relevancy and visual algorithm and will make even more changes later.  Ahmet, Baris, Navneet, Xioafan, Diem, Jacquie, and now Marissa have been working to make more and more fashionable items come to the top and hence help folks find what they are looking for, and to make the visual search better.  Our visual signatures are now 10x larger (hold more information about the item) that when we first launched a year ago.   I won’t go into specifics here but we are very excited about the possibilities.

Eighth, we have redone our visual search results in formatting based upon user feedback and more importantly just in-lined the results.  I don’t remember who came up with this idea but we realized that why should people need to click the visual search button.   Why not just show them a few right here in the results.  This seems like a simple feature but the back end server requirements were significant.  We are now firing off hundreds of extra similarity searches per second.  Dan, Diem, and many others worked days on the right technical solution.  Like.com was already the fastest visual search engine in live product, now it is even faster to allow us to inline the results.

Nineth, our business is all about testing and we buy traffic from Google and others so we need to make sure it converts.  With this release we will have a brand new multi-variate A/B testing system.  We have learned so much by just rolling it out to 50% of the audience and seeing if out performed.  Our world class SEM team of Ph.Ds have figured out how to grow our business, but I won’t go into that either, except to say that Vikas, Danny, Achal, and Sam are so scary smart – sometimes I wonder if I am the dumbest guy in the room.

Tenth, we realized that there are lots of features that matter so much in soft goods that we had to do them right.  We let you show only what is on sale.  We dynamically generate price ranges so you can only see items in the range you want.  We let you only see items with free shipping.  This is critical in buying these items as they don’t fit much of the time, we tell you which merchants have return free shipping for the same reason – almost no comparison shopping engines show you this.

Finally, we added lots of merchandising.  Search is great once you know what you are looking for but we realized lots of folks wanted more guidance.  You will see this in the menu items and in the category pages when you click on an item in the menu drop downs.

Thanks to the entire Like.com team …  for helping us to grow in 2007, for setting us up to do well over $10MM in revenue in 2008, and for taking this site out of Beta and into GA.  As always my special thanks to the two professional pillars of my life without whom there is no company here (just a dream) – Azhar Khan an Burak Gokturk.

I’m sure the next few days will be rocky as not release it perfect, but for things you love a little poop is worth it… ;-).

Go / No Go Conf Call

Yesterday’s conference call repeated itself today at 6pm again.   This time I am in the car driving with Vijay and Deven to Santana Row.  Vijay hasn’t gotten out of the house for seven days since the birth of Sofia and needs to stretch her legs.  Like Silicon Valley dads all over the place I am on a conf call while she and Deven are running around Broders.  She knew I had the call but wanted to go anyways and said she was just happy to get out even if I had to be on the call.

I joined promptly and already Jacquie was going through the open issues one by one.   Unlike yesterday her reading went something like this:

1.    Drill down 500 error issue – fixed
2.    Flakiness of interface – fixed (it had been due to backend initialization issues)
3.    Formatting issues for other browsers – almost all fixed
4.    Fixed, fixed, fixed…

It seems that Robert Long, Diem Vu, Dan Chiao, and Andrew Miller had managed to work wonders.  Even our intern Mei Than had chipped in at the last minute.  Robert in particular seems to have knocked a lot off a bunch of issues in just under a day.  I was duly impressed with the whole team.  They rock!
One issue still remained – speed.  We really had no way of knowing if the new site was fast enough.  Some preliminary data said it wasn’t. 

However, speed at Like.com is a carefully measured.  We deploy a script called JStimer which measures for each user what the total time they experienced in drawing the page in the browser.  We add this to other times that we measure for each user and generate our true mean and median times.  We watch this number each and every day.  Since it comes from a measurement of every user to the site we know it is an almost exact measure to what our users are experiencing.  Andy, Dan, and a few others had put this infrastructure in place.  It was a great. 

Unfortunately there is no way to really forecast the JStimer number.  The only way is to roll out the site gather data and roll back if it is horrible.

If the site was too slow our numbers would crash. We have seen that even 200ms can matter and impact the efficacy of the site.  As Sergey and Larry are supposed to drill into people at Google – slow sites never win.

This is one of the reasons I hate hard PR driven launches.  In a normal setting you just launch gather data and if bad roll back.  With PR out there it is much harder to do this.

We talk about it for a few minutes, but after months of development we all know that this is a risk we are going to have to take.  There is just too much we will only learn when this goes live.  We go around the virtual table of the conf call.  I say are you guys on the bus to roll this out tonight – Jacquie says yes,  Dan says yes, yes, yes, yes from the others.  Finally I wait for the last votes – Azhar’s and Burak’s– they both say … Yes!  We are off.  In a way like kids waiting to xmas, our patience is done and we just want this site out.  (see photo of Andy in the foreground and children at our Christmas party).

However, we hatch a contingency plan.  We will work on more speed issues tomorrow and after 24 hours we will either roll back the new site or roll out a patch that improves speed.  We will have both options ready.

Dan coordinates with Andy and Dhiraj Kumar (who I have written about here) who is our super star release engineer, coding engineer, and jack-of-all-trades QA guy.   As I drop off the call I hearGuys, let’s start the IM chat at 10pm for the release. 

We would never have built this company without Dan ... that is a fact.

I hang up the call and walk with Vijay and Deven back to the car (Sofia is at home with Grandma).  Vijay is a little tired but the outing and fresh air was good for her.

The Revenue Alexa

As Michael, Matt, and Fred continue to discuss Fred's assult on Matt's journalism skills.  The topic of how reliable Compete.com, Alexa, and Comscore are comes up again (disclosure note I own a bunch of comscore stock via a  venture fund I am involved with).  All of these are fine services and on the whole do a good job with the resources at their disposal.  The issue is not which of these to use, but rather why as Web 2.0 matures we should use any of them.

I would ask every reporter to get a new metric for evaluating how startups are doing.  Having lived through 2000, I know this venture funding party will (has?) come to an end.  The falling stock market  / debt markets may have already started but process, but once it does revenue and profitability will be the only real metric for a startup.  Both in updating Matt for this article and updating Michael for our last article I showed the two of them our revenues.  I asked them not to write the specific number for competitive reasons, but I showed them the financials so that they could have the first hand information of how we were really doing by the main metric that matters for a business.  What if every startup did this?  What if there was an Alexa for revenue.  Obviously in some ways this is silly, but it isn't silly for reporters to ask to see this before writting a positive piece about a startup.

Years ago in 2000 (2001?) my last company Andale was covered by Nick Wingfield who was at WSJ at the time.  He was putting us in "Under the Radar" a column that show cased new startups.  He told me that the last few he had written about had gone under and he wanted to make sure we wouldn't.  He asked to see our financials.  I let him and he wrote the story.  Unlike hundreds of other startups at the time Andale didn't go out of business and grew into a profitable company.

Why can't Michael, Matt, Om, and others start asking for this info before writting positive pieces about companies.  They don't have to publish the numbers just use them to verify the veracity of the story.

On a side note - even revenue is not he right metric, we eventually need to get to profitability which we hope to achieve this year, but let's at least move in this direction.

Bloggers - show me the money

I woke up this morning to find that Matt has already posted his article ahead of the site going live.  Seems like we had a miscommunication.  Arrghh. 

His article is here

Interesting thing though is that a few other bloggers like Fred Wilson and Mathew Ingram beat Matt up about how he didn't run the Alexa, Compete, Comscore analysis on his quote from me that we are bigger than Thisnext.com.

Come one guys.  You all assumed that in talking about being bigger I was talking about traffic.  I was not and frankly don't care to be bigger in eyeballs (eventhough we are).  I care to be bigger in revenue and profits.  This kind of eyeball analysis is so Bubble 1.0-ish. 

When talking about ThisNext I was referring to revenue to revenue. Like.com will exceed $10MM in revenue this year with much of it already in the bag from a run rate perspective. While ThisNext is private and I don't know their exact revenues my understanding is that it is not as large (someone can correct me if I am wrong). In addition, traffic is not a good measure of revenue. For example if Like.com had the same click-through rate as it did a year ago (19%) at 3MM uniques it would have 4.5x less revenue than at 90%. Social shopping sites have had a challenge with this metric - Click through rate. So while they may be close in uniques they may still be far off in revenue. In the end this means that their revenue per visitor is lower.

As a 2nd time entrepreneurs, we at Like.com had decided to not focus on eyeballs or users, but rather revenue and eventually profits.  If you remember we first took this approach back in May of 2006 after we launched Riya (and while we had usage) and didn’t know how we were going to make money.  We retooled and launched Like.com, which we knew would make money.

However, building revenue takes time, but like preparing for the winter it is not something you start when the leaves are falling but something you start many months earlier.  With the economy in recession, I believe that winter will soon arrive to Web 2.0 and those who have revenue and profits will survive, not just those who have eyeballs.

Matt was shown our revenue numbers month by month and my quote refers to revenue.  Saying he didn't run the comscore graph is just plain silly and very bubble oriented.  It is time Web 2.0 companies start focusing on revenue and profits not just traffic.

Like.com GA release ...

February 17, 2008

6pm last night (Saturday) we had a conference call with all of the engineers.  The new version of Like.com, which is supposed to go live yesterday, has issues. The new site is slow, every fifth click it hangs, and some parts are not working in all browsers.   This new site is a complete make over of Like.com.  It is the largest release we have done since launching almost a year ago.  We have been working on this for over 2.5 months.  Crap!   We gotta get this out.   

On the phone are: Dan Chiao our trusted lieutenant who has been there since day one – he has transitioned from the rapid prototyping hacker a few years ago to a methodical systems architect of a large part of the backend for Like.com; Andy Miller, who is one of the best Data center architects I've ever worked with and our guru on site performance; Eric Lee, a relatively new addition to our team, who has worked 5 weekends in a row for this release and finally took some much needed time off … on his own volition calls in from a ski resort and offers to help; Robert Long our front end engineer who I don’t know well yet, but who seems to be very very skilled and conscientious about his work;  Jacquie Phillips who in just the last few months has gone on to become one of the best product managers I have had the pleasure to work with. Jacquie is a no nonsense mid western girl who followed her passion to NYC and eventually put herself through school at NYU before becoming an early fashion blogger with shefinds.com.  In a company of computer science Ph.Ds she has provided the domain knowledge we needed to really make a better search engine.

She is the core of a new team strategy we have been developing all year (code named: beauty and the geek) that aims to create the smartest fashion search engine on the planet.  Visual search was just the beginning.  We recently hired a super smart and very talented fashion designer named Marissa Goodman (more on her later) to add to this strategy.

Azhar, Burak, and myself, and a few others are as also on the call.  While I am on the call, Sofia Shah, my new 7 day old daughter rests on my chest. Having her there is calming for me and all babies love sleeping on their parent's chest - something about the beating of the heart reminding them of being in the womb.   I am happy she is there with me. 

Can we get this out today but it is clear we are not ready.  The speed of the new site is 1.5 times slower than the current version.  The flakiness is unacceptable.  We decide to fix the issues and hold the release for another day at least.   

The urgency is that we already briefed Matt Marshall from Venturebeat a few days earlier.  I had walked him through the last year of Like.com.  What a year it has been.  We started out with zero revenue and now we were on track to do more than $10MM in revenue in 2008.  Our search engine used to now let people find what they were looking for.  Only 19% gross click through rate.  Now it was close to 90%.  We had not blogged or done any PR.  In fact I probably would not have briefed him except that he asked about a recent small financing we did and asked how we were doing.  We had just focused on building and building and tuning and tuning our search engine.  All the while our traffic continued to grow from zero to over 3MM unique visitors a month. 

Matt is supposed to hold the article until the new site is ready to launch so we end the call with the action items.  Dan, Diem, & Andy will look at the backend.  Robert and Eric will look at the front end performance.  We will reconvene tomorrow at 6pm to see where we are.

VP Finance / Controller

Friends - I need your help to find a VP Finance / Controller.  I tried looking for this position back in Feb but we were just not ready as a company for this hire.  Below is the description.  If you have any leads I would appreciate it.  Have them email me at "financejobs at like.com".  This email will come directly to me so anyone you send will not get lost in "the shuffle."

Like.com (formerly Riya) is a visual search engine.  Like.com is the first search engine that lets you shop by appearance.  We look inside the photo at the color, shape, and texture of an item and use that information to find similar items.  The company has raised over $19MM in financing and has filed over 7 patents. 

Over the last few 9 months our revenues have grown considerably and we need a world wide VP Finance/Controller based here in our US office.  The ideal candidate has/is

•    A certified public accountant
•    10+ years of experience as a controller
•    3+ years of experience as a corporate controller for a    
      technology startup
•    Experience with equity based accounting (FAS 123R)
•    B.S. in Accounting, Masters or MBA from a top 10 University is
      preferred
•    Successful completion of at least 3 years of audits by outside
      auditors
•    Experience preparing reports and presenting to the Board of
      Directors of a venture funded startup.
•    Experience with venture financing and technology mergers and
      acquisitions is preferred.
•    Micro-revenue experience and ecommerce revenue experience
      preferred.
•    Experience with CPC (Cost per click) or CPA (Cost per action)
      Internet revenue systems a major plus.

This position reports to the CEO and comes with a generous compensation and equity package.  Like.com hires only the top 1% of candidates, compensates them generously, gives them lots of responsibility, and expects nothing but the best from them.

iPhone sucks buy Blackberry stock

I know I'm swimming against the tide on this one, but hey with the India reverse outsourcing I'm on a roll lately... ;-).  I am giving up my iPhone and going back to my Blackberry.  Those of you who know me know I love everything Apple.  I personally own 4 Macs at my house, one AppleTV, two apple LCDs, and 3 ipods.  So I am certified Steve Jobs fan. 

However, this iPhone just sucks.  Before you say I didn't give it a chance.  I did.  I used it for 35 days as my primary phone.  As a business user it just doesn't work for me.  Here is why:

a) It is not a better phone as Apple proclaims.

- You can't jump to an entry in the address book by typing a few words.  I have thousands of entries in my address book and the way you have to scroll to get from the beginning of the S's to the right Shah (you can't believe how many Shah's there are in my address book) is just ridiculous.

- It is very touchy.  I have accidentally called at least 10 people in the last week alone.  The deal killer for me was when a CEO of a competitor called me and I sent him to vmail because I was in a meeting and in actuality it answered the call and he heard parts of the meeting.

b) The is a horrible email device.

- After a month I can still type 10x faster on my blackberry than my iPhone.  Now before you say I have big hands... I don't ... I am only 5'6".  Before you say I didn't learn to trust the auto correct... I did try and trust it many times but frankly a few of the messages I sent were so badly corrected people wrote back saying "what did you mean by this email".

- It only has 200 emails - which is like 1/2 days email for me

- It has no search or sort

- The SMTP configuration between wifi and Edge mode doesn't work for my SMTP server at fastmail so I have to switch the server each time I switch networks.  I know there is a solution if you use SMTP.com or a similar service, but geesh... who has time for that.

I use my phone for email and calling (imagine that) and the while the other features of the iPhone are good (Deven loves to watch Cars on my phone when we goto a restaurant and he is bored), they are not enough to compensate for these two features.

So alas I'm switching back to my Blackberry and I'm guessing Blackberry has nothing to fear from Apple for now and maybe forever (I believe that iPhone can't be a good email device until tactile feel of real keys it somehow simulated).

Debbie Schlussel didn't do her homework

Debbie Schlussel wrote about our bringing some of folks from India here.  She didn't call us or even do her homework.  I left her a comment on her blog which I've pasted in below. 

Debbie,

I wish you actually took the time to investigate things before writting such an article. Our company is not bringing anyone over on an H1(b) visa. We are applying to an L1 - intercompany transfer. We have applied to the INS like all company's and they can approve or deny our applications.

On a more political level I think you are not seeing the forest through the trees here. On balance because of this action we are bringing more jobs back to the US by 3x (translate as hiring more US citizens) than the small 7 folks we have to transfer here from our India office. However, in order to create the least disruption our ability to bring these 7 folks here is key to our ability to bring 21 more jobs to the US.

In any event, next time call me before you write so a least you get the type of visa right.

Reverse offshoring is complicated and the key to the US welcoming back jobs is going to be some required flexibility in allowing the key employees to transfer back in order to provide the least disruption.

Episode 27 - Dhiraj Kumar

On my trip to India this week, I spent a day in Pune with my 84 year old Grandmother.  She retold me a story I’ve heard before about my father Girish Shah.  My grandfather Popatlal Hemani (our last name was changed to Shah by my Grandfather as was commonly done by refugees from the countryside then) was a tailor in Mumbai having fled Gujrat during WW2.  He was deaf and largely poor.  They lived in the then outskirts of Mumbai, in an area called Kandivali.  Their home was a single room with a kitchen where all six of them lived. 

My grandmother proudly said that my father was the bright and ambitious one.  He scored at the top of his class and when it came time to go to college he got into IIT Mumbai (Powai).  Ironically, life isn’t always about skill.  My father was actually #2 in his high school.  At the time entrance into IIT was not by a national exam, but rather offered to the #1 student in each high school.  In his case the #1 student decided not to go.  To this day I don’t know who that person was, but he literally changed the course of my father’s life and mine.  Every time I see kids running in the streets of Mumbai, I realize that for one person’s decision that could have been me.

IIT was the first of two significant opportunities that changed my father’s life. My father studied hard and then got a second rare opportunity.  UC Berkeley offered him not just admission but a scholarship to come to the US.   The story of his immigration is iconic for his era. He took a steamship from India through the Suez Canal with one suitcase and $50 dollars. 

I don’t know who in the US thought of allowing foreigners these opportunities in the 1960s but my father and I have them to thank.   I also believe this created a bi-directional obligation.  One to give back to the US economy (which he did with his early work on building the Point-of-sale credit card payment systems and I am continuing through the jobs created in the US by the two companies I have co-founded) and one to eventually give others in India the right to come to the US and continue this obviously ROI positive process for our economy and for humanity.

As you have read here we recently decided to move our engineering back to the US for largely cost reasons (ironic isn’t it).  While this will create even more jobs in the US, a few of our key guys from India will be needed to functionally make this work.  So we are bringing them over. 

One of these guys is Dhiraj Kumar.

As a CEO, there are many times you have to make tough decisions which are not pleasant but occasionally you get to do something that feels great.  Being able to offer Dhiraj the opportunity to come the US was one of those times.

Dhiraj grew up in a small town in Bihar.  His father, a school teacher, died with he was 5.  His mother, also a teacher, brought him and his brother up by herself in their hometown of Rampur. Dhiraj was smart and ambitious.  He was the first person from his village to goto IIT. He spent much of his savings buying his brother a truck so he could remove sand from near the river and sell it as building material in the middle of town (vs carrying it by hand).

He joined Riya as a QA person, but he was not content to be only that.  He aspires to learn to be a developer.  Recently, we sent 5 people to Java training 6 months ago and he was the only one who learned it well enough to transition to programming.  Today, he doesn’t just find the bugs he now is able to fix them himself most of the time. 

Dhiraj impressed the heck out of both of us.  His dedication and attention to detail is amazing.  He once drove to our IT guy’s house in Bangalore (who was not being super responsive), brought his laptop and made him fix something on the spot.  Dhiraj is willing to call anyone in the middle of the night (including me and Azhar) to fix a bug if it is live on the site and critical.

As Azhar and I were making the list of people who should come to the US, we knew we needed Dhiraj.  We knew he was someone we wanted by our side.   He had earned it as much as anyone and we felt good to be able to give him the opportunity. 

I remember the look on his face when we told him about the opportunity to come to the US.  He looked like he was going to cry tears of happiness.  He later told us that if he comes, he will be the first one from his village to every go to the US.

I had a hard a week last week, but Dhiraj Kumar was the highlight of my week.  Dhiraj Kumar thank you for everything you’ve done for us and welcome to California.

27 - Leadership team, 26 - India grows up, 25 - Million mice, 24 - Land Invasion, 23 - Nuclear Power Plant,
22 - Fast iteration , 21 - Like.com prognosis , 20 -Launch Video and More , 19 -PR Coverage of Like , 18 - Like.com Launches , 17 -Pre launch , 16 , 15 , 14 , 13 , 12 , 11 , 10 , 9 , 8 , 7 , 6 , 5 , 4 , 3 , 2 , 1 .