Silicon Valley

Posts related to living in Silicon Valley

Ask Not For Whom the Bell Tweets…

/ Adam Nash

If you have not been following the latest Twitter drama, I thought I’d share it briefly here on this blog.

Meet Connor Riley.  She seems like a nice enough young woman, going to UC Berkeley.  Her personal student website is online, including her resume.  After all, she is looking for a job.

Good news!  She landed an offer from Cisco, one of the few big tech companies hiring these days.

Unfortunately, she sent out a message on Twitter that was captured as follows:

theconnor_ciscofatty2

She apparently didn’t understand that by default, everyone can see  you tweet.  Needless to say, someone at Cisco saw this tweet (likely from a saved search or TweetDeck stream for “Cisco”) and responded as follows:

fatty_answer

Thus, the “Cisco Fatty” incident was born.   Amazing stream – you can see the Twitter search here.  The drama.  The intrigue.  A couple articles:

Connor actually posted a response on her personal site.  It’s not much, but here’s a sample:

Sometimes in the course of applying for a job, it becomes apparent that it’s a job you don’t want to do. I declined one such job early on Tuesday, and then, because I live at some distance from many of my close friends, I decided to use Twitter to tell them about what I had been thinking.

Let me tell you about how I use Twitter: I have 45 friends. I know all of them. They know me. 95% of them have lived in a dorm or a house with me. I practically can’t offend them, although sometimes I try.

So one checkbox stood between my using Twitter correctly to suit my needs and my using Twitter in a way that would make @timmylevad start baying for my head.

It’s not really that compelling.  Still, I found myself thinking about a few things:

  1. Get used to “Bad Tweet” stories. We’ve heard a few “bad tweet” stories before (remember the “Memphis” incident?).  And we’ll hear more.  It’s the new, hot social medium, and these stories will take on a life of their own.
  2. The bad economy affects reactions. This would be one of those classic schadenfreude stories, except that with the economy where it is, people are particularly indignant at anyone who would flaunt and dismiss a great job at a great company like Cisco.  It’s overstating the case, but in some ways, this taps into anger the same way the AIG bonuses do.  This is just a Gen Y, techie version.
  3. Twitter seems private, but is public. There are least two very clever aspects to Twitter that have helped its member and usage growth.  The first is being designed, from the ground up, to separate “following” from “follower”.  Who you see is kept separate from who sees you.  The second, however, is a play on privacy.  Twitter feels private, and the interface leads you to believe that only the people following you can see your tweets.  However, in reality, everyone can see everyone else’s tweets by default.  The advent of realtime search streams has only made this more obvious.

People use Twitter like its a new, better form of group chat… but it isn’t.  These messages don’t just go to friends and family on your buddy list.  These updates don’t go only to your connections. And until the interface changes to suggest to people that their tweets are public, we’re going to see more and more people make the same mistake that Connor did.

Would You Pay $12.99 for 5 Hours of Facebook?

/ Adam Nash / 2 Comments

This is a note I meant to post over a week ago, but didn’t get around to it.

The question is, would you pay $12.99 for 5 hours of Facebook?

The reason I ask is, until a couple weeks ago, I would have assumed the answer was no.  Facebook has become the latest in a line of great, free internet products.  It flows open and free, like a trillion pages of Niagara Falls, unimpeded by usage charges.

Then I flew Virgin America to JFK & back.  5-6 hours each way.  And the flights had Wi-Fi.

(The Wi-Fi was fantastic, by the way.  I got a phenomenal amount of work done on the plane, and having live access to email and the web was incredibly useful.  Having realtime access to Twitter wasn’t as useful, but certainly was fun.  I also saw some funny behavior patterns – like people watching live sports on the laptops while their seat-back television was on CNN or CNBC.  Anyway, I digress…)

For $12.99 you got wi-fi… for about 5 hours.  Worth the cost most likely to make the flight productive for work, especially compared to an $8 snack pack.

Next to me on the plan was a woman, likely 20-25 years of age.  As soon as we were allowed to use our laptops, she flipped hers open, paid the $12.99… and went to Facebook.

I was sitting one seat away from her, so I could see what she was doing.  She spent about 3 hours on Facebook, with a small amount of miscellaneous web surfing mixed in.  But it was almost all Facebook.

It was interesting to me, because the economics of Facebook have been fodder for discussion in the Valley for a couple of years now.  And here I was, watching someone pay $12.99 for Facebook.

It then occurred to me how much money the “dumb pipes” of the internet are really making.  How many people upgrade their internet service to broadband because they want to make YouTube faster?  How many people are effectively paying the service providers to access content created by others?  How many people pay charges for internet service at hotels, airports, coffee shops?  To wireless providers, cable providers, satellite providers, phone providers?

It’s an interesting counter-balance to the argument that the service providers give for bandwidth throttling and other pricing power maneuvers.  They would still argue they aren’t getting enough of the pie.

Still, I’m pretty sure that Facebook got 0% of that $12.99.

Makes you realize why AOL actually worked back in the day.  You know, in simpler times.

PDMA 2008: Building a World Class Web 2.0 Product Organization

/ Adam Nash / 1 Comment

Last year, I had the opportunity to speak at the PDMA International 2008 conference in Orlando, FL.  I gave a talk entitled:

“Building a World Class Web 2.0 Product Organization”

While I posted this presentation to Slideshare and on my LinkedIn profile, it turns out I never actually posted it here on this blog.

Christina Wodtke, author of Elegant Hack and a Principal at LinkedIn, gave a talk this week on Product Management and borrowed a few of my slides.  As a result of that talk, I saw this blog post, about the definition of a product manager, come through my Google Alerts today.

pdma_adam_nash_product_manager

For those of you who’ve worked with me, it’s a classic “Adam Nash slide“.  The tell-tale sign is the use of simple geometric shapes, typically in pastel colors.  (I’m not proud of my limited PowerPoint skills.  In fact, you could say I’m proud that I don’t have advanced PowerPoint skills.)

Anyway, I’m glad to see that the content was useful/interesting for both Christina and her audience.  It was also a great reminder to post the deck here too for anyone who is interested.

BTW The second edition of Christina’s book on information architecture is now available on Amazon.  You might want to check it out.

Startups, Technology Companies & Giambattista Vico

/ Adam Nash / 4 Comments

I had one of those “delightful” newspaper moments today.  I was going through my Sunday morning ritual, page-by-page through the Sunday New York Times, when I happened upon an interesting editorial in the Week in Review.

The article itself was interesting, but likely one I would have ignored in the online version.  (It’s still one of the virtues of print that I put myself in the hands of the editor, and read the Week in Review from beginning to end.)  What was delightful about it was its philosophical reference to Giambattista Vico.

You see, until today, I had no idea who Giambattista Vico was.  However, it turns out that this 18th century Italian philosopher published a theory of societies that happens to match, almost exactly, my recent theory about start-up technology companies and their development into large, successful enterprises.   Here is a summary from the Stanford Philosophy website:

Nations need not develop at the same pace-less developed ones can and do coexist with those in a more advanced phase-but they all pass through the same distinct stages (cursi): the ages of gods, heroes, and men. Nations “develop in conformity to this division,” Vico says, “by a constant and uninterrupted order of causes and effects present in every nation” (“The Course the Nations Run,” §915, p.335). Each stage, and thus the history of any nation, is characterized by the manifestation of natural law peculiar to it, and the distinct languages (signs, metaphors, and words), governments (divine, aristocratic commonwealths, and popular commonwealths and monarchies), as well as systems of jurisprudence (mystic theology, heroic jurisprudence, and the natural equity of free commonwealths) that define them.

In other words, Vico outlines three distinct phases for societies:

  • An age of gods, when man and immortal walk amongst each other
  • An age of heroes, when the gods have departed, but their children or disciples perform wonders with their power
  • An age of men, when their is equality and democracy among men, and a lack of the supernatural

(Yes, I’m grotesquely paraphrasing.  Bear with me on this one for a moment).

When I left eBay in 2007 to join LinkedIn, many people asked me why I was interested in joining a startup at that time.  Being an avid fan of Greek Mythology, I told friends that there were three phases to the tech company lifecycle in Silicon Valley:

  • The golden age, when gods (aka founders and first employees) walk the floors.  This is a time of incredible vision, passion, and risk.   The events and people of this era become myth and legend rapidly.  The company typically at this time has a product/concept, but no proven business model or engagement with customers.  The company is usually measured in tens of employees.
  • The bronze age, when the gods give way to the heroes, the first wave of executives who help grow and scale the company and fulfill its destiny.  Usually this is a time when the business model has proven out, and the larger risk to the company is its ability to manage growth and scale the organization in both talent and execution.  This is still a time of passionate debate and eccentricity, but now at a larger scale as the organization and business broadens.  This is when the company goes from tens of employees to thousands.
  • The iron age, when the gods and heroes have fled, and the company is managed as a large, public technology company.  At this point, the company is typically measured in tens of thousands.

Amazing similarity… no doubt both Vico & I were both fans of the classics.

When I joined eBay in 2003, it turns out that I joined the company well into its bronze age.  Many of the early employees (and a founder) had left, but most of the original heroes who worked under them and with them remained.   There was no separate corporate entity, and the PayPal acquisition had just happened.  In 2003, a product manager would still present a product strategy directly to Meg at times.  But by the time 2007 rolled around, as many of the heroes  departed, it was clear that eBay had entered its iron age.

Obviously, there are later phases for technology companies that can be interesting.  (Believe me, as someone who joined Apple in the mid-1990s.)  And there are always outliers (Google has stayed in its bronze age longer than most.)  But these phases do a fair job of describing the cultural dynamics of those first few phases of a technology company.

For companies, there are no clear delineations between the ages.  The transitions tend to be gradual, and as often as not tend to reflect the four-year pattern for stock option vesting schedules.  In the last few years, however, I’ve found this framework fairly effective in describing how company cultures evolve, and how that influences the enjoyment and job satisfaction of employees who prefer one phase over another.

Maybe the reason this analogy has been useful for me personally is because, as Vico supposed, it reflects a more general description of how groups of people evolve socially when they dedicate themselves to a single social contract.  For Vico, that was a nation.  For Silicon Valley, it’s a start-up.  It’s interesting to consider that the venture capital financing model and stock option vesting model tends to encourage this type of phasing almost naturally over the growth of a new technology venture.

Something to think about, of course.

Refinancing? Try Pentagon Federal Credit Union (PFCU)

/ Adam Nash / 4 Comments

With rates plummeting these days, many people are choosing to refinance.  My family falls into that bucket, as we refinanced our home almost five years ago (2004) at the low, low rate of 4.5% for a 5/1 mortgage.  (In case you are curious, we ended up getting a 5/1 because while 30-year rates were also low, we felt it unlikely that we’d be staying in our current house more than 7 years.)  At the time, I remember looking at historical rates and saying:

When will we ever be able to refinance at 40-year lows again?

Silly me, the answer turned out to be about five years later, in late 2008.  Since our rates were about to float, and not trusting the bank to keep the rate reasonable, we decided to lock in rates for at least another five years.  In the process, we evaluated almost every web-based pricing agent, several internet deals, and one professional mortgage broker.

The winner: Pentagon Federal Credit Union

To get a mortgage, you must join the credit union.   You can do this for free if you or a family member served in the armed forces and has proof.  Otherwise, you can sign up to join the National Military Family Association for $20.  Yes, that’s right.  $20.

I’ve been extremely happy with the service.  In fact, when we originally engaged with them, the rates on a jumbo 5/5 mortgage (a unique mortgage they offer that resets every 5 years based on US Treasury rates) were at 5.375%.  They have dropped twice since then, and they allowed us to reset at their current price of 4.625%.

Yes, I know there is probably a better deal out there.  But I also know that most are worse.

In any case, they are worth checking out.  Their rates are updated daily.  Low closing costs.  They depend on Fannie Mae for securitization, so it’s really a good deal if you have a good credit score and fit within guidelines.  No points for loans under 70% LTV.  0.50 points for loans between 0.70% and 0.80% LTV.

If you find a better deal out there… don’t tell me.  I’m happy enough as is.  I may ask for you help again in 2014.

Or, at the rate we’re going, if rates plummet to 100 year lows in 2009, we might refinance again.

3% mortgages?  That’s the magic of deflation…

LinkedIn and Reid Hoffman: Recession Ready

/ Adam Nash

I don’t usually post every article here about LinkedIn, but this BusinessWeek piece is worth reading:

LinkedIn and Reid Hoffman: Recession Ready

The online version is an extended version from the print magazine.  I think the only reason to read the print version is the unique and slightly intimidating picture of Reid across the center spread.  🙂

Here is my favorite passage:

It was at the bleakest stage of the dot-com bust, in 2002, that Hoffman began to build his empire. He had been a key partner at PayPal, the online payment company ) bought that June for $1.5 billion. Flush with his share, he looked for next-generation investments—and found himself nearly alone. “The common wisdom was that the consumer Net was dead,” he recalls, and “that it was controlled by Yahoo, eBay, and Google. I thought it was just beginning.”

So he devised a strategy. He would start a company to run the business side of the social Net—LinkedIn—and he would participate in the consumer side as an investor. “The huge majority of things I have invested in are massively successful,” he says. Many of the investments, of course, are still locked up in companies, including Facebook and LinkedIn, which haven’t yet gone public. The return on Hoffman’s holdings hinges largely on how they navigate the coming downturn. Still, he continues trumpet the economics of Internet businesses. “This is the only time in human history when, for somewhere between $5 million and $30 million of capital investment, you can create a sustainable ecosystem for 10 million-plus people,” he says.

When discussing his career, Hoffman can sound positively utopian. He says, for example, that he left academia for business because he wanted more “scaled impact” in his quest to make the world “a much nobler place.” He regards LinkedIn as a system where the good are rewarded by the community for their deeds, while liars and cheaters are exposed.

I feel exceptionally lucky to work for Reid and to benefit from both his experience and his unique strategic insight.  When you remember the issues the PayPal team had to navigate, and the timeframe in which they did it, it’s an even more impressive story.  How many companies managed to IPO successfully in 2002?

Check out the full article if you haven’t already.  More to come.

J-Curve & The Hype Cycle: Potential Exits

/ Adam Nash / 1 Comment

Will Hsu had a very interesting post on his blog, Hitchiker’s Guide to 650.  (Yes, it’s a pretty cool blog title)

Will overlayed the now infamous Hype Cycle and a hypothetical startup valuation J-Curveover each other, like this:

2989415251_27295b1663

(Minor nits – the J-Curve here likely shouldn’t start at zero, but at some higher amount.  The founding team and the concept itself has some value, and typically, while the startup is nascent, the value hinges on that alone.  In fact, it probably rises initially as risk is taken off the table with a few key hires/revisions.  It doesn’t change the insight from the overlay, however.)

He then postulated a few different exit points, with reasonable valuations and time frames, and then highlighted the different ROI values for each.

  • Exit #1: 2~4x, 50~150% IRR (assuming 1.5~2yr hold, 1~2 rounds)
  • Exit #2: 2~4x, 30~70% IRR (assuming 3~5yr hold, 2~3 rounds)
  • Exit #3: 10~100x, 30~70% IRR

(You can read the full details here)

I must have seen versions of the  J-Curve and The Hype Cycle curves a hundred times, but for some reason, seeing them overlayed in the context provides some unique insight into the highs (and lows) of a venture backed startup.  It also highlights the incredible cost to being caught flat-footed (ie, needing cash) at the wrong points in the curve.

I also like the clear, numerical validation of a simple truth of venture investing (and entrepreneurship):  you achieve the highest internal rate of return by cashing out quickly.  But to achieve truly game-changing cash returns for investors (ie, return the fund), the big win is required.

The numbers really aren’t as material as the visualization of the two curves together.

Marc Andreessen Joins eBay Board of Directors

/ Adam Nash

This is literally yesterday’s news, but was worth a mention here.  From the eBay Ink blog:

Marc Andreessen has joined eBay’s board of directors, effective immediately.

Andreessen is most noted for co-founding Opsware and Netscape, and served as AOL’s CTO immediately following its acquisition of Netscape. His current venture is Ning, a new consumer Internet company founded in 2004 that is focused on building a next-gen platform for social networking. Rather than having its users join one all-encompassing social network, Ning encourages and allows users to create their own social networks for anything they’re passionate about. In four years, more than 480,000 social networks have been created by users on Ning.

I had a chance to meet Marc briefly as part of the OpenSocial launch @ Google last year.  There is no question in my mind that eBay will benefit from having his perspective on the board given their current challenges.

Some interesting facts & links:

In particular, I’m going to flag a post I wrote over a year ago about how eBay missed its opportunity to buy Ning cheaply, and why that acquisition would have made sense.  I caught some flack for that last summer… feeling at least partially vindicated here.

The Latest Large Prime Discovered: 2^43,112,609 – 1

/ Adam Nash

From Science News:

Here’s a number to savor: 243,112,609-1.

Its size is mind-boggling. With nearly 13 million digits, it makes the number of atoms in the known universe seem negligible, a mere 80 digits.

And its form is tidy and lovely: 2n-1.

But its true beauty is far grander: It is a prime number. Indeed, it is the largest prime number ever found.

The Great Internet Mersenne Prime Search, or GIMPS, a computing project that uses volunteers’ computers to hunt for primes, found the prime and just confirmed the discovery. It can now claim a $100,000 prize from the Electronic Frontier Foundation for being the first to find a prime number that has more than 10 million digits.

Don’t worry prime hunters, there are prizes still to be claimed:

The Electronic Frontier Foundation became interested in prime hunting because it makes an excellent challenge problem for cooperative, distributed computing. “The award is an incentive to stretch the computational ability of the Internet,” says Landon Noll of Cisco Systems Inc., one of the judges for the Electronic Frontier Foundation prize and a discoverer of a former biggest known prime. More prizes remain to be claimed: a $150,000 award for a prime with 100 million digits, and a $250,000 award for one with a billion digits.

In case you are wondering why I’m posting this here on my blog, I do have some personal historical trivia that makes the issue of large primes sentimental for me.

The first job I ever had writing software was an unpaid high school internship at NASA Ames Research Center, here in Mountain View.  My project was to build a simulation model to evaluate error rates for different fluid dynamics algorithms.  In order to do the project, which was executed on a Cray X-MP supercomputer, I had to learn Fortran.

The sample project I chose to do to learn the language was a simple program to take as input a Mersenne Prime, and then generate the actual digits for the number in a large output file.

As a side note, this was the first time I also ever became familiar with the operating costs of these type of high end systems… I remember being fairly shocked when the scientist I was working with explained to me that my program had taken several hours of Cray time, which was billed at about $2,000 per hour.

Of course, I’m fairly certain that my new 8-core Mac Pro is significantly faster than those old Cray supercomputers… 🙂

An Obama Article During the Republican Convention

/ Adam Nash

I generally don’t write about politics here on my blog, largely because I tend to be more issue-oriented than party-oriented, and that seems to bring out fire from both sides of the aisle.

Right now, I’m hopelessly behind on keeping up with the conventions – I’ve downloaded all the speeches from the DNC, but haven’t watched them yet.  Similarly, I haven’t yet watched a single minute from the RNC.

Strangely enough, however, I was forwarded a link to an article Marc Andreessen wrote about meeting Barack Obama in March 2008 that is worth reading:

A Hour and a Half with Barack Obama

Marc’s blog, by the way, is the blog that most closely resembles what I wish my blog could be.  Or should be.  Most of the articles are deep, interesting, sharp, and reflect frank advice and perspective that you don’t typically find in either professional news or popular blogs.  Worth subscribing to if you don’t already.

Ganzbot: I Could Not Be More Proud

/ Adam Nash

I am fortunate to manage a really cool team at LinkedIn.  How do I know this?  Because when Steve Ganz, one of our more senior web developers came back from vacation, he was greeted by… GANZBOT!

Ganzbot is a lot of fun… he is hooked up to Twitter, and to a custom queue application, so you can make him say almost anything.   I sit right across from him, so I’m blessed with hearing a lot of the best Ganzbot interactions.

Now featured, of course, on:

Now, if only we could teach Ganzbot to write modern Javascript…

A Eulogy for eBay Express

/ Adam Nash / 47 Comments

If you follow eBay closely, you may have heard the news already. If not, I’m sure you’ll be reading more about the big eBay announcements over the next few days.

AuctionBytes has coverage, as does Business Week, but I actually think Randy Smythe has the best summary I’ve seen to date.

There are a huge number of changes, and I’m not going to cover them all. Instead, this post is dedicated to one of the smaller bullets in the announcement:

Closing eBay Express: The best features are now on eBay. We’re continuing to bring the best features of eBay Express into eBay.com including more selection in Fixed Price merchandise, improved buyer protection from PayPal, and easier, more intuitive ways for buyers to find your relevant listings. So we’re closing eBay Express and focusing our resources on improving and bringing buyers to eBay.

Since my name was so closely associated with this effort at eBay during my last two years at the company, I figured it was appropriate to post a few thoughts here for those who are either personally or professionally curious.

First off, there is no way to avoid the fact that I feel sad to see eBay Express close. When you build a team and put literally thousands of hours into something, you want to see it continue to live, grow, and flourish after you’re gone. But I’m not going to spend a lot of time on what might have been now.

Instead, I’d like to reflect on just a few key topics: why eBay launched eBay Express, what we got right, what we got wrong, and why eBay Express likely doesn’t fit with eBay’s current strategy.

Why eBay launched eBay Express. This is one is pretty simple, and was publicly discussed in several forums, but I rarely see it accurately reflected in regular press/analyst coverage. It all started in Q4 2004, which was a real wake-up call for eBay. It was the first quarter where the metrics made it clear that there were significant issues with the way buyer demand was scaling on eBay.com.

eBay Express was the culmination of three years of various forms of market and customer research that effectively argued a simple truth: as e-commerce continued to become more and more mainstream, an increasing number of buyers were looking for a different shopping experience. At the time, we called them “convenience-oriented buyers”. While buyers loved the value and selection of eBay, convenience-oriented buyers were looking for more convenience and trust in their shopping experience. They wanted good prices on fixed-price items from reputable sellers, with first-class convenience in checkout and customer service.

When we looked at the needs of both buyers and sellers to make this type of market successful, we found that they were radically different than the auction model eBay.com was based on. eBay Express was the culmination of one possible solution to that problem – a site that leveraged the tens of millions of high quality fixed price listings that eBay already had, while providing a brand-new shopping experience for buyers.

The key to this bet was that with literally zero additional work for sellers, we could boot-strap a brand new marketplace with millions of sellers and tens of millions of items from day one. Once the marketplace had traction with buyers, we would then be able to roll out new seller features and services more appropriate to a high-volume, fixed-price venue.

What we got right. Without getting into the weeds here, there were quite a few things eBay got right with eBay Express. Not all of them may be appreciated by those outside the company.

First and foremost, eBay Express represented a radical break with the way eBay designed and built products. We had volumes of research from over the years, and we literally went across every page, every flow, and asked the tough questions on why this couldn’t be simpler, easier, better for the buyer. The team had two fundamental principles:

  • Keep the site “seller agnostic”, ie, 100% backwards compatible with existing seller process. Selling on eBay Express should be so compatible, sellers shouldn’t even necessarily know that their items were selling on eBay Express.
  • Always ask, relentlessly, “What’s best for the buyer?”

With a strong, dedicated founding team, the effort drew many of the best and brightest from within eBay to assist with every area of the product and across technology, design, and product. At the time, most people at eBay worked on a large number of projects at once, with divided focus across many different features. With eBay Express, time was of the essence, so people had a chance to spend 100% of their time dedicated to the effort.

The end result was a huge leap forward in both technology, patents, user research, and design thinking for many product areas. A modern search classification engine. Relevance sorting. A full featured shopping cart. A completely rethought integration with PayPal. 24/7 Customer Service. No listing fees, with revenue coming purely from promotion and successful sales conversion. Even though the team did not win all of its feature fights to break with the old, the team asked the hard questions, and fought the hard fights.

Not as visible to end users, the groundwork was also laid for significant changes to the way eBay Express would integrate with other sites, both inside and outside of eBay. Half.com integration. Shopping.com integration. Dynamic CPC & CPA-based Featured Placement. API-based platforms to allow any e-commerce site to offer multi-vendor inventory to complete their offerings.

Most importantly to me, eBay Express was designed with extremely heavy involvement from our customers, both buyers and sellers, as well as development partners. In fact, it was reviewed so many times, that even at launch, I don’t think one “new” question came up that hadn’t been raised previously. That isn’t to say that every customer loved every decision made for the site, but it did mean that every concern, every suggestion was considered and incorporated into the design when possible.

What we got wrong. This could be a long section too. Like all 1.0 products, there were a lot of small things we missed. But there were a few big ones that seem so obvious in retrospect.

  1. Branding. It was a tough decision. If you don’t use the eBay brand, you lose any possibility of the positive affiliation and traffic that comes with a known consumer parent brand. But, if you use it, you are also stuck with the negative attributes. eBay means auctions to most people. We ended up going with eBay Express because in the end, it was eBay inventory and we expected traffic to flow from the eBay association. It didn’t, and it also didn’t generate any real unaided awareness for us.
  2. Traffic, traffic, traffic. One of the unanswered questions was how to drive sufficient traffic to the new site. We had initial stabs at this problem, but eBay was still in a phase where it believed in buying traffic. TV, Catalogs, Email, Paid Search. It doesn’t take an Internet genius to realize that buying traffic is horrendously expensive, and frankly, ineffective. Our biggest course correction post-launch was a crash course on how the rest of the e-commerce world looks at traffic generation. Figuring out how to drive traffic in volumes to the site, and build organic traffic in the long term became our 24×7 focus.
  3. Inventory and merchandising. It may be hard for most people to believe this, but eBay at the time was incredibly under-developed on many of the retail basics of merchandising, inventory selection, and promotion. Why? Well, because eBay.com isn’t actually a retailer of anything. We realized post-launch that we needed to develop that expertise, quickly, even to the point of understanding sourcing, distribution, and product selection. Having 10 million+ products is great, but it’s no good if you don’t have the right products at the right price.
  4. International. We designed and built the site, from the ground up, to meet the different needs of the US, UK, and Germany. In fact, I even spent time on concept versions for India, China, and a host of other countries. There were some fundamental disagreements about which model would be most effective, so we built a platform to handle them all. In retrospect, we should have done the US only, and only expanded internationally once we nailed the basics. The distraction, debate, and expense was counter-productive, and in the end, a mistake.
  5. Expectations. There was so much enthusiasm internally around the various aspects of the project, and it was impossible to contain expectations rationally. The reality is that building a consumer brand and a billion dollars in sales doesn’t happen overnight, and it isn’t cheap. Look at how long Amazon has been stretching to build it’s third party sales efforts. We believed we could cut that time in half, but rationally, that was still a minimum 5+ year effort. In the best of times, that kind of effort requires a company with long term focus and commitment. And as we all know now, 2006+ were not the best of times for eBay.

Why eBay Express likely doesn’t fit with eBay’s current strategy. If you’ve actually made it this far through the article, you probably already know the answer to this question.

At a high level, economics speak loudly here. eBay needs to focus on its core marketplace business, and for the most part that means that investing people, technology and dollars towards building new businesses has to take a back seat. You’ve seen other announcements from eBay about closing other businesses, and that stems from this simple truth.

More importantly, eBay has decided against the premise of eBay Express. Our entire reason for building a separate site was because we believed that the changes needed for buyers and sellers in a massive fixed-price marketplace were not compatible with the experience of the traditional eBay auction site. As I used to tell buyers and sellers, we built eBay Express so that we would not have to change the auction experience that millions of buyers and sellers loved on eBay.com.

eBay has now decided that it needs to fold the convenience and trust we identified into the core platform itself. So there is no need for a separate site to preserve the original.

How this new strategy will fair is good topic for debate, but for another time. With eBay’s new strategy, eBay Express will now live on as its feature design concepts and technology innovations become the basis for the new buyer experience on eBay. Of course, the team at eBay has made a large number of improvements and changes in the design concepts to adapt them for the needs of the core marketplace, both from a technical and user experience perspective. eBay Express also lives on as a relentless focus on building a great buyer experience, and a recognition that the needs and economics of high volume, fixed-price sellers are different.

In retrospect, I’m a little jealous of the progress Amazon has made with its FBA and API programs since then. These were all part of our long term thinking as well, so it’s nice to see the validation of their success, but it’s never as much fun to see someone else with that success. Maybe, just maybe, back in 2005 before Amazon had it’s run-up in stock price, eBay & Amazon could have merged, and the the eBay Express backend could have been used to power the Amazon marketplace. Easier said than done, of course.

For the 600+ people who had a hand in creating perhaps the greatest technology & product effort in eBay history, please do join the eBay Express Alumni group on LinkedIn. One of the great things about this industry is that we all get chances to take our lessons from each challenge, and then go and change the world again.

Go with peace, my friend.

Update (08/20/2008): Wow.  This post has been really popular.  Over 300 page views already.  Given the interest, I’m digging up some of my earlier posts on eBay Express:

Congratulations to Mike Schroepfer & Facebook

/ Adam Nash

If you missed the 39 articles currently on Google News on the topic, Mike Schroepfer announced today that he has accepted an engineering leadership position at Facebook.   Mike is the current VP of Engineering at Mozilla, and has been there for about three years.

Here are Mike’s comments, direct from his Mozilla blog.

In my opinion, of course, this is a huge win for Facebook, as they get a top-notch engineering leader to join their team.  It’s bittersweet, of course, because I’m also a huge fan of Mozilla and the team over there.  I think Mike summed it up best in his post by expressing confidence in the ability of the Mozilla team to continue to innovate and deliver on their mission and vision.

If you are curious, here is Mike’s LinkedIn profile.  You can be sure I’ll be on his case for him to keep it up-to-date. 🙂

Reminder: Why Apple Killed Clones in 1997

/ Adam Nash / 1 Comment

Some interesting press coverage over the past todays about Psystar’s announcement that they will be selling a $399 Mac clone:

Psystar Sells $399 Mac Clone: Return of the Mac Clones?

A quick snippet from the post:

Budget conscious Mac shoppers can save a bundle on a $399 mid-level Macintosh computer running OSX called an OpenMac sold by a Florida-based company called Psystar. That beats comparable offerings from Apple, whose cheapest similar computer, a Mac Pro, starts at $2000.

Now for the catch. The Psystar computer appears to violate Apple’s end user license agreement (EULA) for Macintosh OSX, which prohibits running the operating system on anything other Apple-branded computers.

The Leopard compatible Mac is built using standard computer parts with specs that include a 2.2GHz Intel Core 2 Duo, 2GB of DDR2 memory, Integrated Intel GMA 950 Graphics, 20x DVD+/-R Drive, four USB ports and a 250GB 7200RPM drive, according to the Website MacRumors.com. I would’ve pulled the specifications from the Psystar Website myself, but the site was not functioning and, the last time I checked, displayed the message: “Site is currently offline due to the massive influx of users in the last 24 hours.”

So, obviously, coverage like this is sensationalist.  This $399 machine is nowhere close to the specifications of the $1999 Mac Pro.  It’s much closer to the Mac Mini, which is $599 to start, although this offers more expandability.  However, everyone loves to talk about Mac clones, so you can forgive the urge to create a big story here.

Since I happen to be at Apple in 1997 at the time when Apple killed clones, I feel somehow irresponsible if I didn’t remind everyone why Apple launched clones in 1995, and why they killed clones in 1997.

  • First answer: market share.
  • Second answer: economics.

Here is the explanation in the article:

In 1997 Apple decided to halt its MacOS licensing program. Back in the Mac OS 8.0 days, Jobs–who was only a consultant for Apple at the time, though he soon became “acting CEO” –reportedly called Mac clones “leeches.”

Circa 1997 you could buy a Mac clone made by Power Computing, Motorola, or Umax that was faster and cheaper than anything Apple was selling.

At the time, Apple was losing OS market fast, so Mac clones were viewed as an important strategy for Apple to survive. PC World’s Charles Piller wrote: “Furthermore, no single company–no matter how creative and dynamic–can compete against an entire industry. The engine of innovation that will keep the Mac competitive has to include clone makers.”

Jobs didn’t agree with Piller’s analysis.

In one of his first major decisions as acting CEO for Apple, Jobs yanked the clone program. He saw Apple’s profits in selling computers, hardware, not licensing software. Microsoft, it was widely accepted, had already won the OS licensing race.

Sorry… this just isn’t accurate.

Gil Amelio launched the Mac clone market in a belated attempt to boost Apple marketshare.  The thinking was that clone makers would expand the Mac hardware base into niches that it didn’t currently occupy, growing the base of Mac users and Mac hardware for developers to target.  They assumed some small amount of cannibalization, but it was assumed that the overall pie would get bigger.

The problem was, the Mac wasn’t set up to clone easily, and Apple really didn’t have the infrastructure to support a large number of clone makers.

That, by itself, could have been just growing pains.  But after just a couple years, it was clear that Apple had to kill the clone market.

Why?  Economics.

The clone makers were not, in fact, expanding the Mac user base.  Market share for Mac OS machines was not improving.

However, that wasn’t the worst of it.  The real problem was profits.

Now, I know what you are thinking.  “Microsoft has huge profits!  Selling just the OS is far more profitable than selling hardware!  What are you talking about, profits?  Apple would mint money if they licensed the OS…”

It’s the difference between profit margin and total profit.

Let’s say Apple sells a $1500 Mac with a margin of 20%.  That’s $300 in profit.

Let’s say a clone maker sells an Apple clone for $1500.  Apple sells the clone maker a copy of Mac OS for $50, with a margin of 98%.  That’s $49.

Uh-oh.

That’s right, to replace the margin dollars of an Apple machine, you have to sell several clones.  That means the clone makers have to expand market share by 5+ machines for every one they cannibalize.

But it’s actually worse than that.  Manufacturing computers has a lot of fixed costs.  So, theoretically, if you cannibalize enough machines, the margins on product lines can decay.  You can hit a point where you aren’t even making money on the machines you are selling, without raising prices.

Now, Apple could have raised the OS price to the clone makers, but as you can see, not enough to make a difference.

The reason I tell this story now is that, fundamentally, Apple’s economics for Mac hardware haven’t really changed that much.  They still get 20% margins.  And Apple hardware is still, on average, about $1200-$1500.

Now, Apple is a much bigger company, and theoretically, it could eat the profit hit today, if it wanted to.  But make no mistake, it would be a real hit to profits.  And that means a hit to earnings, and that means a crashing stock price.  It’s not obvious how Apple can cross this chasm without multi-billion dollar dislocation in profits over a transition period.

As a final note, I really enjoyed this lesson when I learned it back in 1997.  In the 1990s, it was conventional techie & MBA wisdom that OS licensing was an obvious win for Apple, and that it was something that “had to happen” for Apple to survive.  Both, of course, were proven to be categorically false.

Why Everyone In My Family Has Blue Eyes, Except Me

/ Adam Nash / 2 Comments

Today, I discovered “The Spitoon“, the blog from 23andMe, the company dedicated to personal genomics.   Really interesting material.  I found this article particularly eye-catching:

SNPwatch: One SNP Makes Your Brown Eyes Blue

I’m curious about this, of course, because while I have green eyes, my wife Carolyn & my two sons have blue eyes.  It seems that this isn’t even due to a single gene – it’s literally a single nucleotide pair.  From the article:

Three recently published papers (here, here, and here) report that a single SNP determines whether a person’s eyes will be blue; every blue-eyed person in the world has the same version. The findings also suggest that the blue-eyed version of the SNP can be traced back to a single ancestor that lived about 6,000 to 10,000 years ago.

It’s been known for a while that eye colors like green and hazel (deviations from the brown color found in the majority of people) can be explained by SNPs in a gene called OCA2. The protein made by this gene is involved in the production of melanin, a pigment found in the cells of the iris. This is the same pigment that gives your hair and skin their color. Darker eyes have more melanin than lighter colored eyes.

But none of the known variations in OCA2 could explain blue eyes. The new research seems to have solved the mystery. A SNP near OCA2, but not in it, determines whether a person will have blue eyes.

The SNP, rs12913832, is actually in a gene called HERC2. Scientists think that instead of affecting HERC2, the SNP controls how much protein will be made from the nearby OCA2 gene. Low levels of OCA2 protein, caused by the G version of the SNP, lead to lower levels of melanin, which in turn leads to blue eyes. 23andMe customers can check their genotype at this SNP in the Genome Explorer or in the Gene Journal (Note: In the Gene Journal you’ll see other SNPs also associated with eye color. The combination of these SNPs with the blue-eyed version of rs12913832 can end up giving a person green eyes instead of blue).

What a great blog.  Sign me up for that feed.

As a side note, Michael Arrington has posted his account info from 23andMe on TechCrunch, so you can live vicariously through him in case you are short $1000.  I have to admit, seeing those results makes me jealous – I’d love that kind of genetic detail on myself & my family members.