Identity, Personalization, and Social

The rise of Facebook has cast a large spotlight on social functionality in the last few years.  Social graph and Feed are becoming relatively common parlance, at least in the Valley, and increasingly there's a feeling that everything has to be social or that everything is inherently social.  I share the excitement but not the conclusion.  My cofounders and I started aardvark, despite great jobs at Google and elsewhere, because new APIs to social networks suddenly allowed us to create functionality we'd always dreamed of.  I think Social is a sea change.  But, I see Social as only one of three complimentary axes of value along which you can satisfy users.  The other two axes are Identity and Personalization.  Social may be an evolution of identity and personalization.  But, your product or service might be well served by devoting resources along either of the other axes at the cost of underemphasizing Social.

 
I'm a geek, so my mind generally goes to the data that is an input or output from my experiences on the web.  In that context, a service delivers value around Identity when I can see data that pertains to me (think MyWare).  When Identity comes into play, I have a different experience because I see something different than another user.  Many services don't have to deliver any value along Identity; when I read a paperback, I enjoy the exact same experience as anyone else doing the same.  Other services don't make any sense outside the context of Identity; my bank primarily delivers value to me by letting me have an account that has "my money" not "your money" in it; I would get no use out of receiving an anonymous or typical bank statement every month.  
 
Personalization happens when a service sees my implicit or explicit data and provides me with a different experience in result.  Mint.com is an example of a site that delivers significant value along the axis of Identity, showing me my data.  But Mint also makes a powerful play along Personalization, showing me specific offers that could help lower my spending in various categories (like mobile services or banking).  If Mint doesn't see evidence of spending in certain categories, it won't give me offers in those verticals (or at least Mint will deprecate such offers relative to those that seem personally relevant).  Generally some elements around Identity are a prerequisite for Personalization but that's not necessarily so.  A good salesperson offers essentially no value along the dimension of Identity but she will size you up and deliver a pitch that is finely tailored to you while refining her guidance around your reactions.  Hunch is a site that provides very personalized recommendations without needing to know who you are through a decision tree model.
 
Social happens when others see my data.  The power of Social is that my experience is materially affected by others and improved by the immediate or persistent network around me.  Most of what I actively do on Facebook (i.e., pressing a "like" button), I do because other people see the effect and, through their reaction, motivate me to do more.  As Facebook inevitably personalizes things based on my "likes", they will be layering Personalization on top of something that was initially Social.  [As my co-founder Damon convinced me] Social is its own equivalent pillar to Personalization and not just a refinement of it.  As for the relation of Social to Identity, it's reasonable to suppose that typical social functionality will require log-in but that isn't necessarily the case.  Good parties are all about being social without the use of nametags and without the host shepherding your experience.  Sites like Iminlikewithyou and Flikr can provide a sense of community and be social without requiring log-in.  
 
Being able to offer Social (especially around a person's existing real-world network) on the Web is an incredibly new and fantastic possibility these days.  You'd be a fool to ignore the potential effect of Social on your product.  But Social is not the be-all-end-all.  Lots of services could use much more functionality around Identity and Personalization.   I would even be willing to sacrifice some of the features I have around sharing and connecting with friends on Facebook to be able to get more Identity and Personalization value; The number one most useful feature of Facebook for me is photo storage.  My photos had to get uploaded and tagged, through killer social functionality.  But now that those pictures are there, I would pay a meaningful amount of money to have those photos for my own reference even while I would hesitate to pay to be able to see new photos tagged with me.  [I recognize that this is largely out of irrational overreaction to loss versus gain.]  As for Personalization -- it drives me nuts that, when I x out an ad and say that it's repetitive, I inevitably get shown the exact same ad.  
 
Ultimately, I subscribe to the mantra that you should focus your efforts fanatically if you want to succeed at something hard.  There's almost certainly more danger in trying to pursue Identity, Personalization, and Social at once than in devoting your attention to a single line of development.  In fact, many companies would be well served by culling the efforts they often inadvertently staff along whatever axes are non-core to their user value proposition.  In conceiving a social product for example, challenge yourself to provide a major social value without requiring log-in or without providing an experience to one user that is any different based on their activity.   Still, caveats aside, I'm hugely enticed by building products that are best of breed along all three dimensions of utility at once.  

Repost of Interview with Rob Go (www.robgo.org)

Product Leadership Series: User Driven Design at Aardvark

This is the first post in a series of interviews on Product Leadership.  I’m very pleased to kick this series off with an interview of Max Ventilla at Google.  Max was a co-founder of Aardvark, a social search company backed by Harrison Metal and August Capital that was acquired several months ago by Google. 

What struck be as unique about Aardvark early on was the way the initial product came together, and the fact that the company is a 30 person organization with no product managers.  Instead, they are heavy on user focused designers and product focused technologists. 

I’m not a journalist, but I’m going to keep this in Q&A format, but won’t jot words verbatum (although the content was approved by Max).

RG: What problem were you trying to solve with Aardvark and how did you go about it?

MV: We were solving the problem of searching for subjective questions.  Different people can search for the same concept but are looking for very different answers because they are different and the context of their question is different.

Startups are most likely to succeed when founders really understand a space and they are taking advantage of timely trends (making the wind at their backs) 

Our team really understood search and the trend of increasing openness of personal information was at our backs.  The opening of the Facebook platform was a catalyst for us to really go after this problem. 

RG: You talked about being user driven in your product design approach.  Tell me about what this meant for Aardvark. 

MV: My last startup had a great product but limited user traction.  We decided to approach this one differently.  The very best product managers are right 20% of the time.  Most are right less than 10%.  There’s too much “hunch” in product decisions, so we set out to build products that could be tested.

It is tremendously beneficial to choose a product that you CAN be user driven about.  It’s hard for some enterprise products, or a search system that requires that you process responses in milliseconds.  We tried to test products that we thought could be appealing even if they were in a very raw form.  

RG: So, phase 1, you have a problem and no idea how to solve it.  What did you do?

MV: We self funded the company and released very cheap prototypes to test.  What became Aardvark was the 6th prototype.  Each prototype was a 2-4 week effort.  We used humans to replicate the back end as much as possible. We invited 100-200 friends to try to prototypes and measured how many of them came back.  The results were unambiguously negative until Aardvark. 

Once we chose Aardvark, we continued to run with humans replicating pieces of the backend for 9 months.  We had 8 people managing queries, classifying conversations, etc. We actually raised our seed and series A rounds before the system was automated - the assumption was that the lines between humans and AI would cross, and we at least proved that we were building stuff people would respond to. 

As we refined the product, we would bring in six-twelve people weekly to react to mockups, prototypes, or simulations that we were working on. It was a mix of existing users and people who never saw the product before.  We had our engineers join for many of these sessions, both so that they could made modifications in real time, but also so we could all experience the pain of a user not knowing what to do. 

This is a step beyond the “launch early and often” approach. It’s very very qualitative in addition to quantitative.  Launch early and often and being extremely data driven works, but when you have thousands of users.  It takes a long time before startups get that many active users. 

Stealing from Max’s blog here: “I’d estimate that we moved about half as quickly as if we’d just gone with our gut consistently.  In return, we dramatically reduced the chance that we would make wildly wrong bets…. Ultimately, investors gave money as much for our process as for our team and concept.”

RG: There has been some discussion about the limitations of “lean startup principles” and the Henry Ford adage that “If you asked users what they wanted, you would have gotten faster horses.”  What do you think of that?

MV: In our first phase, the concepts came from the team.  We tested the viability with people, but the vision of the products came from us.  The concept of having a “human” that you would talk to via IM was based on our own intuition, which we tested.  When we were developing the product, we also don’t ask users what they want us to build. We see what works, what doesn’t work, and the underlying problems behind our users struggle. We have a blog post on our methodology on our site here. 

RG: Thanks Max.  Best of luck at Google and we look forward to the continued growth of Aardvark and your new endeavors. 

Stay tuned everyone for our next two interviews with Adam Medros at Tripadvisor and Ben Foster at Opower. 

Preaching User-Driven Design

My cofounders and I started a company in July 2007 very deliberately before we’d decided what product to build.  We had a problem that we wanted to solve -- that it was too hard to make and carry out every day decisions using the web.  But we were not committed to a particular solution. In fact, our CTO had compelled us to put off choosing an idea until 2008.  Having the right set of cofounders, with whom I had a deep history and highly complimentary skills, was the most important factor in our success.  But that commitment to not pick an idea for five months, no matter how alluring those early products were, was ultimately the second most important factor and represents a criticial principle for other startups to follow.

If you are going to be doing a startup for an average of four years if you are unsuccessful and eight years if you are successful, it makes sense to devote six months to deciding what to do in the first place.  If you go with whatever comes to you first, unless you are very lucky, you will likely waste far more than half a year before you can pivot to success.  More often, you’ll run out of money and morale and end up in the dead pool. 

When you don’t have any way of knowing whether one idea is way better or way worse than the ideas that will come later, it helps to burn some early concepts to establish a baseline.  But it’s incredibly frustrating to be an entrepreneur and come up with one idea after the next only to have people not use them.  It’s also very discomforting to not even have a name for your startup.  So, when October 2007 came around we adopted the catchy but temporary name “Launching September”, committing ourselves to coming out with something by a year later (ultimately the month an end-to-end Aardvark went live). 

But it’s better to spend a month on an idea than a year.  It’s also essential that you restrict yourself to the kind of products that you could reasonably prototype (i.e., no critical low-latency component and not highly visual and not for enterprise).  At Launching September (later to become Aardvark), we would fake functionality with people on the backend and see if our early users -- many who didn’t know they were using a faked prototype -- kept coming back.   When users don’t click through on an invite, they wouldn’t want your value proposition even if you’d built it. 

Aardvark was the sixth idea that we tried out.  But, we didn’t just stop at being user-driven in selecting an idea.  After we committed to the Vark, we spent nine months doing wizard of Oz testing.  In the last months, before we switched over to an automated system, we had eight people pretending to be the system (though real users answered questions).  In every interaction, a person would be classifying queries, routing questions, and managing conversations.  By the time we turned Aardvark on, we’d had tens of thousands of sessions from which to learn how the system should be built.  That end result, informed by actual usage, was materially different from where we would have ended up had we gone off into a garage with the concept and emerged with an implementation three quarters of a year later. 

Even once we had an end-to-end system up and running and didn’t have to involve humans on the back-end, we continued to test the vast majority of features manually before they were engineered.  We built a tool so that a team member could put users through alternate flows and experiences.  After each test, our designers would email anyone who had experienced the experiment.  Over the lifetime of the company, we consistently enjoyed more than a 50% response rate on those emails.  Moreover, the quality and quantity of feedback improved as users became less likely to be friends or family of people working at the company.

At the end of the day, being user driven is a tax.  I’d estimate that we moved about half as quickly as if we’d just gone with our gut consistently.  In return, we dramatically reduced the chance that we would make wildly wrong bets and have to double back, abandoning large periods of work.  Ultimately, investors gave money as much for our process as for our team and concept.  Being truly user-driven also created a much better environment for our engineers, who got to work on features that had been vetted prior to beginning development.  Finally, our process let us avoid the tantalizing product concept ratholes that could have used up years of our collective energy and time to no avail.

Ban Gadgeting while Driving

A few days ago, I took a bus north to Tahoe from SF and was astonished to see how many people are using a mobile device of one kind or another while they drive.  Having succumbed to the temptation before of using those idle minutes in stop and go traffic to check email, I can attest to how distracting it is.  There are lots of studies, of varying merit, that conclude that texting while driving is as dangerous as driving drunk.  So it stands to reason that we should deter people, through penalties and by shaping public perception, from playing with gadgets while driving.

I know that it’s taken decades and enormous work form public agencies and law enforcement to deter drunk driving to the extent felt today.  People still often drive drunk and we still have a long way to improve.  But, as someone that enjoys occasional drinking (outside a vehicle) and has a car, I consciously avoid drinking and driving largely due to the societal pressure and the legal risks associated.  Still, it doesn’t seem feasible to affect a “gadgeting while driving” ban with the same tactics as have proven effective for curbing drunk driving.

I would propose that the police cruise the highways and look for people gadgeting while driving (potentially using inconspicuously marked cars, although the distracted texter would probably hardly notice a normal squad car).  If police saw you using any device while you were driving, they would pull you over and fine you.  Penalties would be high but not outrageous (around $100).  But, critically, they would take your device.  You could go through some onerous process to get your device back in the event there was something on there that was really important or sentimental.  Still, for most people, the violation of losing their iPod / Nexus One / Garmin… would be far more impactful then the fine. 

Perhaps most importantly, reports of device seizures, especially when something quirky was taken, would spread virally and dramatically raise public awareness. Some protocol would need to be established for what to do when the device was attached to the vehicle.  But, assuming the car manufacturer had installed the device, one could lock the gadget unless the car was stationary or someone was sitting in the passenger seat (detected through the same mechanism that causes the seat belt warning to beep). 

One of the most important enablers of enforcement seems to be the level of individual outrage for being prohibited from a particular activity.  Given that police officers are normal civilians when they aren’t in uniform, their attitudes about enforcement are strongly affected by their take as civilians.  A gadgeting while driving ban would seem to be very enforceable from this perspective.  People do not yet feel like they are entitled to play with electronics while driving given how recently it became possible to engage in such activities.  Also, people feel far less protective of their freedoms in a car than in their home.  After all, we’re familiar with being pulled over for speeding and used to prying eyes as we drive. 

The longer enforcement lags hands-free driving laws the harder it will ultimately be to reign in habits.  I think most people that use devices while driving recognize that it’s dangerous and want to stop even while they indulge.  Fines could be calibrated to pay the direct costs of enforcement and marketing to impact public opinion.  California, rich in technology and automobiles, would seem as good a place to start as any.  What do you think?

The .Com Boom Caused the Recent Financial Collapse

Throughout the '70s, '80s, and '90s there was a pretty clear path to prosperity for someone who had had the benefit of a great education.  If you started working at a big bank after college and put in a couple decades, you had a good chance of amassing enough money by your mid forties that your family would never have to worry about money again.  I don’t think that money is the most important thing in life and I don’t think that most people think money is the most important thing in life but it’s legitimately a priority for lots of people.  For such people, in my admittedly oversimplified depiction, banking was a high-probability route to amass somewhere between ten to one hundred million dollars (in today's dollars) before the age of fifty.

One of the most game changing things about the Internet is that its so easy to put up a website.  Moreover, in the early days of online, enterprising people could get prime real estate on the Web (i.e., one word domain names for less than the cost of dinner) and get eyeballs.  The fact that one didn’t need to raise financing or manage staff or even get out of one's pajamas to have a successful Internet venture in 1998 meant that kids could do it.  In fact, computers favored a younger generation.  Also, the pace of progress became incredibly accelerated by the late '90s.  It would be shocking to make an offline business that became frequented by millions of users in less than a decade.  Online, where marginal costs were negligible and everything was a few keystrokes away, websites routinely got to millions of users in a matter of months.

Had the dot com era never happened, a generation of ambitious and competitive individuals in banking would have continued contentedly on their pursuit of an eight figure bank account.  But human beings are terrifyingly subject to upgrading or downgrading their aspirations depending on how people are doing around them.  Fascinating studies show, for example, that people are happier when earning much less money if they are doing relatively better than their peers.  By the late nineties, when the population online had grown to over a hundred million, the financial markets became willing to bet on the biggest sites becoming big businesses.  Many web destinations went public for enormous valuations and their founders became incredibly rich.  So when relativistic financial professionals saw people in their twenties becoming billionaires en masse around 1999 they upped their ambitions by an order of magnitude.

While a banker might accumulate ten million dollars of wealth with a reasonable success rate over a thirty-year slog in banking, it is incredibly unlikely that the same banker could make a billion dollars of wealth in ten to twenty years.  To even have a chance, that individual would have to ratchet up the risks they took.  But in chasing that much richer payout, the associated probability of massive losses (not just for the individual but for the firm) went way up.  And when the first big risky bet went sour, the only choices available were to shamefully bow out of the race or to double down or to attempt a fraudulent cover up.  Given the personalities involved it’s not surprising that many chose the latter option.  Given the example of the Internet, it's not surprising that people had an easier time believing that involving computers completely changed what was possible or sane.

Greed, suboptimal regulation, and lack of internal controls at big financial institutions certainly played big parts in the global financial collapse that started in late 2007.  But if the individuals involved in finance had had more reasonable ambitions they wouldn't have been driven to act so irrationally.  The risks they took in retrospect don't seem to make sense because the additional happiness one gets from getting more wealth than you can spend is pretty small for most people.  That additional pleasure pales in comparison to the pain of personal or vocational failure resulting when a big risky bet goes horribly wrong.  But, when one is chasing something hugely challenging, it's often difficult to asses the accurate probabilities of success and failure.  A healthy disregard for the impossible underpins most great human acts.  But more of a spotlight on the thousands of entrepreneurs that stumbled for every dot com billionaire that emerged in the late nineties would have gone far to prevent a generation from thinking that it was realistic to go so impossibly far in so little time.

Hello World

I’ve been meaning to start a blog for a while, but now that I’m going back to work at Google, I feel like the daily commuting time might actually keep me writing.  So, without further ado, welcome to my first post: Why Posterous?  I figure it’s appropriate to start with a topic as meta as this.  Also, on the heels of selling Aardvark to Google, Posterous provides a good opening to talk about one of the things that I think made Vark successful. 

The Internet is a place where network effects dominate.  Whether a business is algorithm based (e.g., Google), or content based (e.g., Youtube), or activity based (e.g., Facebook), or commerce based (e.g., eBay) – more visitors means a higher quality experience.  That means the site that has a traffic lead tends to build on that lead.  Once a powerful incumbent exists, a new entrant has to be much better along some critical dimension to change people’s habits. 

In the past, I tended to think that many opportunities existed to create new websites in categories where no incumbent existed if one picked carefully and executed well.  But when the time came to actually start a new Internet business in 2007, I concluded that the web itself had become saturated.  Excluding the most fanatical digerati, people do not visit more than a half dozen distinct websites on a regular basis.  In the last five years, I can only count one website that has entered my normal rotation (Facebook).  Youtube rotated in but rotated back out, as have many other sites.  Twitter is on the verge of earning itself a spot.   So it seemed foolhardy to think that I might build a new destination site.  Instead, my cofounders and I decided to build something that could be accessed from the places where people already spend their time online.  The idea was that we should add value to people’s existing habits not try to change them.

Aardvark was the sixth idea that we tried, following a string of failed prototypes (see my comment below).  But all our ideas were subject to the restriction that they could not be a destination site.  Any candidate idea had to be useful from within some other online application.  Aardvark is designed to be a contact that is accessible from anywhere that contacts go (email, phone, IM…).  It wasn’t until we were about eighteen months into the company that we finally built a full-fledged website.  That seemed pretty remarkable for a *web* company but I think it will increasingly be there norm. 

Companies that build their business on the back of SEO, or on iPhone, or as a Facebook application are all part of what I see as a new world order on the web.  If you can get people on masse to type the name of your site into their browser bar, you will probably be worth billions but you need to be a braver entrepreneur than I to chase that prize.  But you’ll start to see more and more companies like Zynga build enormous value in symbiosis rather than competition with the major online brands. 

So, in choosing a blogging platform, it just made sense to use Posterous.  The whole idea of the site is that I never have to go the website and can push content from any major web application to be accessed on all the major third party applications.  That means I can blog when I don’t have convenient access to the Internet (like from a shuttle bus) and I can support fellow entrepreneurs who are moving to build value on the shoulders of the big sites in this new era of open and semi-open platforms.