The “Talent War” Revisited

What Silicon Valley companies are doing the best job of attracting talent? This turns out to be a complicated question, not to mention great fodder for bored designers to try to answer visually.

Recently an infographic started making the rounds, showing the ratios of employees moving between six top tech companies. This was published on the blog of a social job referral service called Top Prospect, generated using two years of their data. The story they told was one of small, up-and-coming companies poaching talent from more established companies.

A designer (Gene Lu) took issue with the fact that all the flow arrows were given equal weight, masking some important relationships. He did a nice redesign adapting the original to paint a clearer picture. The flow lines scaled to volume do a much better job of visually showing who the winners and losers in the “talent war” are.

However, there’s a more fundamental problem with the data underlying both these interpretations. Here’s a hypothetical to highlight the issue: Let’s say we have two equally-awesome companies looking to hire as fast as possible. They have the same low turnover rates of 1% per year. However, one company is much larger than the other. Company A has 10,000 people and Company B has 1,000. After a year, 10 people will have moved from B to A, but 100 people will have moved from A to B. So, even with everything being equal, the natural movement of people in the workforce automatically gives the smaller company a 10x ratio of hires.

This makes some intuitive sense. Even a large company that’s doing well will have a huge number of employees leaving in any given year. Having some of those people end up at a smaller company isn’t all that surprising.

So, I decided to run the numbers again, but this time scaling all the ratios according to the number of employees at each company. You can see the calculations on this spreadsheet. The resulting graphic is below.

This is a much different picture. Microsoft, with its huge employee base (almost 90k), is actually retaining its people quite well. It may seem to a small company like LinkedIn that tens or hundreds of ex-Microsofters showing up is a big trend, but to the Microsoft leviathan that’s a drop in the bucket. Scaling the ratios by company size shows in fact that there are a disproportionate number of LinkedIn employees actually leaving for Microsoft.

Otherwise, the scaled ratios are all relatively small. Google, Facebook, and Apple are at close to parity. The only other big story here is a sad one, and that’s the hemorrhaging of talent that Yahoo is undergoing.

Now, there are a lot of potential issues with this visualization as well. Is a linear scaling the correct way to adjust for company size? What about the base data itself? This data comes from Top Prospect’s small, proprietary sample. Fast Company has noted that the data might be skewed since it’s a Facebook-seeded referral service.

For these reasons and many others that I haven’t anticipated, I actually generated the infographic you see above with a tool I cobbled together using standards-compliant HTML, CSS, and JS. It’s up now. You can it to try out a different model or assumption, and generate new infographics.

For example, it was easy to pull Microsoft out and see the results right away. Enjoy!

No Comments »


postedby on July19th,2011 tagged art, coding, personal

I had in my head the idea that animated moire patterns would be beautiful, and I hadn’t messed with the Canvas element, yet, so I made this thing on a couple flights, last weekend. You’ll need a modern browser to see it, but I’ve included a screenshot below in case you’re not using one.

Click through, and refresh to see new, random permutations.

No Comments »


Two weeks ago, I decided to spend fifteen minutes with nothing but a text editor, letting out a string of thoughts that had been knotted in my mind for months. I was amazed by how much crap came out onto the screen. As I wrote — no edits — the page got ever messier, but ideas began to crystallize in my head. At fifteen minutes, I stopped. Nice: I’d gained a bit of insight.

And then the text editor crashed. But you know what? All the juicy insights remained. It was the process of writing (not the writing itself) that had cleared my mind.

A couple nights later, I joked with friends that someone should create a text editor that crashed every so often as a feature. After a bit of refinement of the idea and some JQuery that night, the idea became “ephemawrite.” A little video of it is below, or you can try it yourself.

Note: The quote in the video is from James Gleick in an interview about his book, “he Information: A History, A Theory, A Flood.”

Also: You can also browse the source (all 50 lines of it) on GitHub. Thanks to Kasima for some suggestions on the interaction design.

No Comments »

All in the Family

postedby on January2nd,2011 tagged personal

For about five years now, a favorite pastime for me and my cousins is to play Mafia:

Mafia ([...] also known as Werewolf, Assassin or Witch Hunt) is a party game created in USSR modeling a battle between an informed minority and an uninformed majority. Players are secretly assigned roles: either “mafia”, who know each other; or “townspeople”, who know only the number of mafia amongst them. In the game’s “night” phase the mafia covertly ‘murder’ a townsperson. During the day phase, all of the surviving players debate the identities of the mafia and vote to eliminate a suspect. Play continues until all of the mafia have been eliminated, or until the mafia outnumber the townspeople. A typical game starts with seven townspeople and two mafioso.

It’s essentially a game of logic and deception, and as my little cousins have gotten older (and wickedly smart), the game has gotten more and more fun. I think we logged about eight hours of it over the last few days.

Relatedly, I took a 360-degree panorama of us playing in my grandmother’s living room. I used an app for my Android device called Photoaf. It does a serviceable job of creating panoramas on the go. To put this up online, though, I re-stitched the photos using Photoshop and output the panorama using Pano2VR:

No Comments »

Sparring Redux

Mark’s out of town, so he has a friend subbing in on instruction. We sparred for a few rounds at the end of today’s session. She says I’m way snappier than I was than a year ago when she last did a few rounds with me. My progress feels really slow and my gradual improvements are hard to notice. So it was really nice to hear.

I took a video of one round just to see, and I agree. You can click through to previous posts to see videos of me from early 2010. I was even slower and more off-balance, then.

No Comments »

CircleVoting.com: Is your privacy protected?

I updated my last post on CircleVoting.com as they added features I was asking about, including ballot information for San Francisco and new privacy controls. The privacy settings are some of the clearest I’ve come across on a socially-connected application. I’ve included a screenshot, below. One problem still is that a new account is defaulted to being Public — I think it should be Friends-Only. But if you take a second to change your settings, the short answer is yes, your privacy is protected.

Screenshot of CircleVoting.com's privacy controls

Aside from the immediate visibility of your voting preferences, though, there is a larger question around how this data is used in aggregate. Voting preferences could be packaged up and sold to the highest bidder. This isn’t necessarily a privacy issue, since the data can be aggregated in a way that couldn’t identify individual voters, but it would be ideal if CircleVoting.com were up-front about whether it intended to profit from this information in the future. I asked Jesse Sanford this follow-up question. It seems that CircleVoting is still just beginning to think about this issue.

Q: Privacy and profit are often at odds. When my friends and I considered building a similar site, a while back, it occurred to us that having aggregate voting preferences combined with user demographics might be really valuable. With Murray’s [Edelman: co-founder and investor] deep experience in polling and public opinion, I imagine CircleVoting.com has thought of this as well. I think this valuable data could be monetized ethically, but there are a lot of ways to do it wrong. Do you make any public pledges to restrict how CircleVoting.com will use its users’ data, and how will you make your policies clear to your users?

Jesse: [T]he big case that isn’t handled right now is the case of a user who becomes more private and needs to remove listeners as part of that. Expect an interface that covers that case as soon as Dennis and I get to it.

To address the question about data retention, sharing and privacy, we posted the privacy policy we’re drafting with our attorney at http://circlevoting.com/static/privacy — parts of it are somewhat boilerplate at the moment, however, and it may change. The gist is, we will never share your personal information except to do business with you (maybe send you an email, etc.) or as required by law.

You raise an interesting question about sharing aggregated information, like statistics; all I can say on that is that we have no plans to do so at this time and I really haven’t given thought to it. My intention is that whatever we do will serve the purpose of cultivating more functional, distributed democracy.

Now, it’s true that in the long run the money in electoral politics always has a side, and most sites like ours will probably face pressure eventually to make money by hooking people up with campaigns and causes they care about (and for which they might eventually donate or volunteer). If we take steps in that direction, our intention is to do so equitably for the various sides of a race/issue, in a way that continues our goal of reducing the influence of money in politics. We’ll also be sure to ask users’ permission and take community views into account if/when we develop that offering.

So, no clear answer here and, as with any service of this sort, they can just change the terms of use. I would encourage the founders to think deeply about how to make their business model transparent to users. At the moment, CircleVoting.com is an interesting service with an appealing social mission. I’m definitely excited to be using it. But to be trusted at scale, it will need to have a more concrete ethical foundation.

1 Comment »

Q&A with CircleVoting.com

I’m really excited about CircleVoting.com, which was released as a public beta last week. Check it out, or read on for a quick introduction to the site, plus a Q+A I did with one of their founders.

Anywhere you live in the US, sorting through all the misinformation during campaign season can be a major chore. Most people just don’t do it. If you live in California, you know this problem is compounded by our extremely messy ballot measure system. Here in San Francisco, I’m facing research on 20+ measures and over a dozen races for public office.

Corporations currently fill this gap, spending immense amounts of money to try to “inform” us on how we should vote. But we have a better source of information: our friends, colleagues, and trusted organizations. If we could summarize all this expertise and see it at a glance, we could make more informed decisions, quickly. Years ago, my friends and I brainstormed ways to do just this and un-break our ballot measure system. We came up with a bunch of ideas and even a prototype, but never finished the job. Happily, this year there’s CircleVoting.com, and they’re making the idea real:

Screenshot of CircleVoting.com's ballot interface

Above is a quick taste of the CircleVoting.com interface. Deciding how to vote on this ballot measure was easy. I can see that a couple of my friends are voting Yes on this measure, along with several organizations I trust. If I click, I can read comments they’ve made and official summaries of the measures. Awesome! Now I can focus my attention of the measures where a lot of people I know disagree. That’s where all the interesting discussion will be.

So, needless to say, I’m into CircleVoting.com. But I wanted to know more about the people behind the site and how it worked, so I got in touch with Jesse Sanford, CircleVoting.com’s CTO.

The team has a great set of complementary skills. How did you all come together to create CircleVoting.com?

Jesse: I met Murray about 2.5 years ago.  I’ve had a couple other projects that attempted to use new media for social change and that’s been a persistent interest of mine.  Meanwhile, I’d also been messing around in IT and web development, at the time as Director of Web Development at Beezwax.  I’d worked on several fairly big-budget sites of a comparable level of complexity to Circle Voting, so I had a good idea it would be possible to do what Murray described.

Is CircleVoting.com currently a labor of love? How do you hope to sustain the site for many election cycles to come?

Jesse: We are set up as a for-profit corporation right now, though we may cross over and reorganize as a 501(c)3 at some point.  Murray [Edelman] is our angel investor.  That said, we’re exploring about a dozen options for a revenue stream.  Verification and more personalized profile pages may be of value to campaigns and public figures.  I also hope to work more with streams and RSS, perhaps moving into the content aggregation space.  After the election I’ll do another round of grant applications and fundraising. Pretty much all the voter education organizations are behind where we are at with social networking, and it would be expensive and risky for them to develop software like ours, so I would think an existing non-profit may also have an interest in taking it on — a modest but constructive exit.

CircleVoting.com doesn’t highlight any privacy concerns around putting our political beliefs online. Shouldn’t we be worried about this very personal information being publicly available?

Jesse: This is a major concern of mine!  Would you please be willing to be a test user on our privacy stuff?  I have it mostly in place but I’m polishing the interface and doing some more testing.  I hope to go live with it this weekend.  In brief, I have developed a three-tier privacy system which will eventually default to “friends only”.  Because I’m testing it on the live site, though, and because we’re really working on expanding our content right now, I think users will have to switch themselves to “Friends Only” after they join.  I’m not sure — I’m likely to explore this with the community over the next few weeks.  Your concerns are essential to address here!

I’ll definitely be a guinea pig. I’ll let my network know how it goes :)

[UPDATE - 10/13: Nice-looking privacy controls are now available, making it possible to set all of your preferences to be Public, Friends-Only, or Private. They default to Public, however, so make sure to go to change your settings if that's not what you want. This is a flaw of the site, in my opinion. I have my profile set to Friends-Only.]

How are the social graphs that you glean from Facebook and Twitter being used? When I added people to my circle, I was presented with a lot of options, but only a few were friends from my networks. Was that just a coincidence?

Jesse: We’re not using Twitter right now — just haven’t had time — except for login.  On the other hand, we do ask Facebook for your friends and we automatically connect you with any friends who have signed up for Circle Voting.  That’s updated every time you log in.  We are thinking of writing a Facebook friend browser for sending invitations to Circle Voting and maybe some notifications (which you could turn off) when you do something public on Circle Voting.

How can a group become a featured option  on CircleVoting.com?

Jesse: Well, that’s pretty easy right now!  We’re going to show public opinions as a second option below people’s circles — i.e. from anyone in Circle Voting who is fully public in our privacy system; since there isn’t too much content in the site yet, it’s easy to have your stuff set up.  We also have a “default circle” set up to demo the site to anonymous users, which will include some of the best opinion-writers, hopefully in a politically balanced way.  We really want to feature people!  Just ask us for what you need here or let us know your ideas.

When will local ballots be available? I see CA’s on mine, but not San Francisco’s.

Jesse: The San Francisco stuff will be live tomorrow, I hope.

[UPDATE - 10/12: San Francisco measures are now up. They're moving fast!]

Awesome, Jesse. Thanks for taking the time to speak with me.

Jesse: Thanks for asking such great questions!


What is Google.org?

A lot of people ask me what Google.org is, given it’s hybrid for-profit/not-for-profit structure. My personal take on it is that while Google has adopted the motto “Don’t be evil,” then the mission of Google.org is to actively do good.

The reality, however, is much more complicated. I highly recommend this piece from the Stanford Social Innovation Review: “Do No Evil”. The article is a detailed look at the six-year trajectory of Google.org, highlighting its transition from a grant-making institution to a engineering-driven, distributed organization within Google. The article also highlights the incredible difficulty that faces the young organization as it attempts to tackle global problems that others have wrestled with for decades, with questionable results.

I’m proud to be a part of Google.org, in large part due to the scope of its ambitions. However, this piece is a healthy reminder for everyone attempting to do social good to approach their work with humility.

No Comments »

Samasource.org -> Google.org

Leaving Samasource.org

The Samasource.org team and I said our goodbyes Wednesday night, after a long evening of well-wishing and reminiscing (and strong drinks). We had a lot to celebrate about. I joined when Samasource was just Leila and two other insanely-devoted people working out of a conference room. Nine short months later, the organization has more potential than at any moment in its history. Every month, Samasource has brought in larger and larger contracts to bring ever-increasing numbers of people out of poverty in South Asia, East Africa, and Haiti. The young organization is proving its model.

The team has grown substantially, and fortunately I leave Samasource in very capable hands. David Yoon assumes technology leadership as the new VP of Engineering, after years of experience building sites like Donorschoose.org. Noah Bradach is the new VP of Sales, closing deals left and right. Chelsea Seale continues to wrangle order out of the chaos of their operations. That’s just a few of our amazing staff. The office is full of dedicated people like Caitlin, Luke, Rebecca, Joon-Mo, Kala, Marcia, Pamela, Tanya, and so many others before them who pour their heart and soul into the organization. And, of course, there’s Leila Janah, who continues to lead with immeasurable grit and intelligence.

The decision to leave Samasource was very difficult. The short of it is that I’ve found an opportunity to work on another problem of enormous scope. If you’re interested in this project, read on.

A prototype of Earth Engine shown almost a year ago at COP15Google Earth Engine

The world is slowly getting serious about putting a dollar value on greenhouse gas emissions reductions.* One of the mechanisms for mitigating climate change that gained substantial traction in Copenhagen last year was REDD (Reducing Emissions from Deforestation and Forest Degradation). REDD is based on a simple fact: the world’s forests represent a massive carbon sink, sequestering huge amounts of CO2 from the atmosphere. How much carbon? It’s estimated that the the amount of forest the world loses every year results in more greenhouse gas emissions than all the world’s cars, trucks, planes, trains, and ships combined. Clearly that the world’s forests represent an enormous value, even if we only consider the value of the carbon they sequester.

The major challenge to the REDD mechanism is verification: How can we know that our actions are actually preserving forest and reducing CO2 emissions? Monumental sums of money are at stake in answering this question properly, as the world must reduce CO2 levels at the lowest possible cost. Unfortunately, there’s a lot of uncertainty about the effectiveness of forest preservation efforts. For example, legislation that protects forests in one region may simply push logging activity somewhere else. What’s needed is a global forest monitoring solution. Until we have this, only token investments will be made in REDD. Today, we see nations pledging a few billion here or there for forest preservation, but the market should theoretically support orders of magnitude more than this.

My new employer, Google (and more specifically, Google.org), is aiming to provide exactly this with Google Earth Engine. The product will serve as a central clearinghouse for all available satellite data and earth surface monitoring algorithms. It will be a place for scientists and policy makers to answer big questions at global scale, and to prove that their methods are best. If our small team succeeds, we will enable these forest preservation markets at a scale we have all only hoped for until now.

I’ve had the opportunity to work with the Earth Engine team half-time for a few months, now, but next week I will transition over completely. I’ll be returning to my roots in frontend engineering. This means designing and building the web applications that people will need to access and understand the vast stores of data that Google Earth Engine will be making public.

Wish me luck!

More great articles about Google Earth Engine:

* How much are our coastal cities worth? Stability for hundreds of millions of climate refugees? The avoidance of massive water shortages and resource wars? Economists, scientiests, and policy makers are having trouble wrestling with a problem at the scale we are currently facing.

One algorithm for analyzing satellite data

One algorithm for analyzing satellite data


Left-Handed Monkey

I couldn’t resist…

Creative Commons License

1 Comment »