The Gong Show

There’s plenty of people talking about the TechCrunch acquisition, so I hesitate to pile on too.  I’ll say this one piece and then I’ll shut up:

My favorite asset in Techcrunch is not the blog network or the conference series.  It’s CrunchBase, their wiki-based, structured data asset of companies and corporate transactions.  

I doubt CrunchBase contributes to TechCrunch’s revenue in any meaningful way, so when I say “my favorite” i don’t mean it in financial terms.  It’s my favorite because it’s:

• Incredibly Useful: I use CrunchBase everyday, and it rarely fails to deliver value to me with every usage.
• Uniquely Defined: When CrunchBase launched, no one else was doing a crowdsourced approach to aggregating transaction data around startups.  Databases like VentureSource existed at the time, but all the content was populated top down and the system was (and still is) crazy expensive. 
• Solid Network Effects: CrunchBase has two network effects going for them. 1) Their data gets more valuable as the network of users edit the data in a typical web-like form of reciprocal altruism and 2) Their PageRank is gaining steam and will only continue to grow as they become the default link for startup transaction data.

Who at AOL will be in charge of CrunchBase? I’d love to talk to him/her.

Y’know how nearly every company in America is a Delaware company, regardless of where they actually do business? There are a couple incorporation-enabling service providers they give you little more than a postal address in Delaware, so a company can take advantage of the well establish corporate legal precedence that has been established there.

Well, the equivalent in indie rock is Brooklyn. If I ever start a band, regardless of where my band lived, I would claim we are a Brooklyn band. If doing so required getting a local Brooklyn PO Box, that would be fine.

The indie cred that comes with being a Brooklyn band feels overwhelming. It’s like a prerequisite just to be allowed to email Pitchfork a copy of your first EP.

Same goes for being an author BTW… Unless you really use your non-Brooklyn geography to your advantage, like Annie Proulx.

The latest chapter in the Wikileaks saga is playing out: Daniel Schmitt does an exit interview as he leaves the project.

Based on this interview, it looks like a combination of technical development stress, national politics, and infighting amongst team members are wearing Wikileaks down. I love knowing that Wikileaks exists, and it’s painful to read about the turmoil the site continues to face.

Does anyone remember the data haven built in Neal Stephenson’s Cryptonomicon? Quick plot summary from Wikipedia from those readers who don’t know what I’m referring to:

Lawrence’s grandson and his cohorts in Epiphyte(2) corporation work to create a “data haven” on the island Kinakuta in southeast Asia, a place where it would be possible to share information freely.

That book feels so prescient now as I watch the Wikileaks drama continue to unfold.

A brief prediction: I think we’ll see a successful, for-profit data haven flourish in the next 10 years. Usage will be gray (much like Wikileaks), and the site will be sustainable, trustworthy, (relatively) secure, and free from political pressure. I doubt it will be hosted in the United States, but it will be hosted sustainably and all nations not subject to government firewalls will have free access. I bet it will be very lucrative, and possibly even publicly traded. The effects of this freely available data haven will be at first profound, and eventually become commonplace and taken for granted.

Emily Green of the Yankee Group said something insightful at emTech last week that’s worth repeating.

To paraphrase Emily: “The most important innovation that the Kindle introduced was not related to reading. Instead, it was the way in which consumers pay for the data plan for Whispernet. Amazon created a great business development relationship with Sprint to bake the price of wireless communication into the cost of the device up front (the price is also likely baked into the cost of the books purchased on the network).”

I agree with Emily, the ease with which one can get a Kindle online was pretty revolutionary considering the state of wireless on-boarding at the time it was launched. Getting a wireless device active on the network is typically a painful experience in which you have to A) pick a network, B) pick a plan, C) setup billing D) register the device on the network. And incompatibilities across networks due to competing 3G/CDMA standards only exacerbate the problem.

Compare that lengthy set of steps to the process of getting a device on the power grid.  You just plug it in and it works, right?  So simple.  I think wireless network pricing will move in this direction eventually.  Some day, getting wireless connectivity will be as easy as plugging a TV Electric Razor into a wall. [Ed. Note: According to one of my readers, it’s impossible to just simply plug a TV into the wall these days… so I’m changing this to an electric razor]. The carriers are incentivized to move to such simplicity because simplicity will lead to increased usage, hence increased revenue.

Wireless pricing is due for further innovation.  Which carrier is going to be first to pull the trigger?  More competition amongst carriers would accelerate this process because simplicity is also a competitive advantage.

[Emily used the power grid analogy too in her comment… due credit to her]

I could use some advice. I hit a wall in Excel badly today. It’s not a huge data set, only 42K rows, but it’s big enough that when mixed with a bunch of lookups across tables, the document no longer reliably saves properly and is very unstable. Any new formula that causes the doc to recalculate will freeze up the doc for at least a minute.

I need to export my data into something else that can stand up better to my usage and help me get back to work. I’m not looking for the perfect solution, I want to optimize for getting productive again as quickly as possible.

I’ve dabbled in R and Matlab in the past, but I’m not certain if either is the right answer, and I’m not excited about having to rewrite all my queries in those languages.

Any recommendations for how to graduate from Excel less painfully and where to go next?

I’ve been thinking about the dynamics behind social gaming and whether or not they are sustainable. I don’t have a clear cut answer, but blogging my way through my thought process will hopefully produce some insights.

Lets assume we’re friends (and if you’re reading this blog, we likely are friends IRL), and I ask you to water my plants because I’m going out of town and don’t want them to whither. You will very likely say “yes.” I’m asking a favor of you, and even if it’s slightly inconvenient, you’ll probably find a way to make it work… because we’re friends and that’s what friends do for each other.

This behavior is exactly the social dynamics that games like FarmVille use to create viral growth. It’s a type of social altruism that we feel compelled to fulfill in order to maintain friendships. For the digerati audience we all know that when we see a notification from a friend asking for help fertilizing their crops in FarmVille, it’s just viral marketing, but for a broad audience of users, that’s simply not true… these marketing “calls for help” are interpreted literally by FB users. Ted Rheingold said exactly the same thing a few weeks back using anecdotal evidence from his family:

People like my mother and sister get steamrolled by these types of [viral functionality]. I know because they ask me ‘do I have to do this’ and ‘why does your brother want me to do this, I don’t want to hurt his feelings by not doing it’. We realize it’s just a viral conversion strategy and can make an independent decision. Most users at the moment think they are hurting a friend’s feelings if they say no. - Ted Rheingold

If I’m right that people *actually* think every call for help to fertilize their friends’ crops is legitamite, then this is not a sustainable channel of viral growth over a long time frame because people are not dumb. Eventually, people will learn that some of these “calls for help” are really just viral marketing triggers, and just like the banner ads of the 90s they will learn to ignore the non-legitimate ones.

I’m not preaching doom and gloom here for social games (on the contrary, I’m quite bullish on the sector). If FarmVille users genuinely care about the quality of their farms (and I think many of them do) and those users are genuinely intending to ask their friend to help them fertilize their crops, then these viral tactics will continue to be a strong channel for Zynga because the social altruism Zynga is playing on will be real. But, any company that is falsely triggering social altruism cues to the friend networks of their audience can only do so in a short time frame until those friends learn that the cues are not genuine.

I’ll be curious to watch what tools users will develop to help them tell which “calls for help” are genuine and which ones are just spam. One tool will be stereotyping (read: all “calls for help” from XYZ game are fake, so don’t trust them). Another tool will be IRL confirmation (read: Sally asking Johnny IRL if he REALLY meant to ask her to fertilize his crops). Another tool will be usage confirmation (read: if Johnny’s FarmVille farm is massive and well-manicured, then his “call for help” is probably genuine because he cares about this game). But, we’re still in the Wild West of social game development, and some tools will emerge that I can’t even fathom yet because they will be responses to viral features yet to be invented. Which tools am I missing that are more obvious today? And, is there a business opportunity to help people understand when social altruism tools are genuine? (I think FB would answer: “yes”… probably Zynga too).

This post is sparked by the following quote from one of Jon Steinberg’s recent posts.

There is definitely an “appification” going on in the mobile space.  We only live in the “now,” and for right now, given mobile broadband speed, users prefer apps.  This may change down the road, but not for right now.   -Jon Steinberg

Jon’s right, we love apps on our iPhone.  But, why? I think it all comes down to user experience design.  There are a few UI Innovations at play here:

1) SAT Analogy Analogy Time.  URL Bar : Command Line :: App Store : GUI Desktop

The command line is fast and unobtrusive when you already know what you want, but it can be very frustrating when you’re in a more exploratory mode.  The command line is nothing more than a blinking cursor visually.  It tells you nothing about what’s behind it unless it is properly commanded to do so.  Your options at any juncture are rarely self-evident by visual inspection alone.

Out of this UX frustration with the command line, the GUI Desktop arose.  When you’re sitting at a computer with a GUI Desktop, you always have all your options visually laid out in front of you.  You’re not wondering what’s sitting behind a blinking cursor, you can *see* your options at all times, and good product design in a GUI environment should always properly cue you with affordances as to what the next step in your path can or should be.

So, I’d argue we’re seeing a similar UX shift in the App Store as the shift from the command line to the GUI desktop.  When you’re in a browser, you don’t know the world of applications that are sitting behind any URL or Google search.  If you want a great web service to manage ToDo lists, you can try googling “ToDo List App”, but the results will be 10 blue links from content pages, not necessarily specific applications.  So, sitting in the URL Bar or the Google search box is like being back on the Command Line, not knowing what options are available to you.

By contrast, the App Store is more akin to the GUI desktop. It’s much easier to explore and learn about new apps.  Granted, there are a myriad of startups out there trying to make the App Store discoverability be even easier, but the App Store was a huge leap forward for getting things do on the phone, a jump that parallels the historic jump into GUI Desktop computing.

2) Drop dead simple purchase and install.

Think about how we used to pay for software (whether or not it was on the mobile phone). If I want to buy and install software on my laptop I need to go to a retailer (online or offline), buy boxed software, open and insert the DVD, install it, enter the license key, etc… It’s silly how difficult the process is in comparison to buying an app in the App Store in two simple taps.

I’ve argued before that the Penny Gap is largely just a usability problem, and the App Store is a significant step forward in solving the Penny Gap. I won’t rehash that argument here, but I’ll simply say it has never before been so each to purchase and install a web app, and ease of this process has created a fundamental change in the way we consume applications.

A slight digression, but along the same train of thought… I can’t believe Apple hasn’t opened an App Store for OSX yet. How is it that I can pay for and install an app on my iPhone in 2 taps and yet for apps on OSX I still need to jump through the same box-software hoops that were created 10-20 years ago? The only plausible reason why they have not done this yet is because they’re phasing OSX out in favor of iOS over the next year or two, with an App Store baked into the new laptop version of iOS.  The iPad will be seen as a stepping stone in that OS evolution if my hypothetical reasoning is true.

3) Rich media for gaming

Games built with HTML5 to be played in the browser on the iPhone look pretty sad when compared to games developed as native apps.  I’m sure browser-based graphics libraries will get better over time, but for now, if you want to consume iPhone games, the experience of the App Store is infinitely better than Googling around for iPhone-compatible HTML5 games.  The biggest reason this is true is that most web-based games are written in Flash, and the iPhone doesn’t support Flash.

4) Leaderboards are baked into the distribution

How do you know if a particular web app is any good? Perhaps you get a personal recommendation from a friend, or you read a good blog post about a particular app.  But, in general, you don’t have much to evaluate the quality of web applications short of signing up and using the service.  In the app store the ratings system and the leaderboards for each vertical are baked into the distribution channel, so you can easily see what is the New Hotness at any give time.  Additionally, leaderboards give developers a place to compete and iterate against their competition, and the winner of that cycle is always the end-consumer that gets better products.

=====

I’m sure I’m missing some reasons for why we love the app store.  Please fill my gaps in the comments.

My experience of the web is now excellent at serendipity, thanks largely to Twitter. But getting info at the *right* time is still hard, particularly for information that is originally designed to be consumed in real-time feeds.

Google tries to accomplish the task of getting me the right info at the right time. When I’m about to go to the movies, I’ll search Google for movie showtimes and movie reviews.  Google tries its best to get me all the relevant information to make an informed decision at that moment, but it’s not a solved problem.  

Google misses relevant real-time data in my movie example. What if when I searched “movies 02118” (adding my zip code to get local showtimes), Google could also source recent movie-related tweets so I could know what my friends liked in the past few weeks?

I know that my friends tweet about movies in general, but I can never remember who said what about which movie.  I could do a Twitter search, but that would tell me what everyone on Twitter thinks about movies, not just my friends, which is far less useful due to the low signal-to-noise ratio. I want to off-board all that memory to a service, and that service should do one thing really well: recall the right information at the right time. 

I had the pleasure of hearing Randy Pausch (famous for The Last Lecture) give the keynote at CHI 2005. He talked a lot about his time working in Imagineering at Disney and also about the HCII at Carnegie Melon, all of which was interesting, but he made one comment that really stuck with me.

Randy really drove home the value of specific, unfiltered, personal feedback. He mentioned a class he taught in game design in which all students have to work closely in teams for the whole semester.  At the end of the class, you write detailed feedback about each class member you worked with, and then all the feedback for each student is aggregated together on a single sheet per student, which is given to the student being reviewed.  Randy described this sheet of feedback as “the single most valuable piece of paper you will ever receive.” Nothing about his delivery of this line made me think he was overstating the importance.

I have received relatively little high-quality peer feedback in my life.  I never took any classes in school that generated a feedback product as valuable as Randy described.  And, I’ve never worked in a job that did real 360 reviews, or really any peer reviews at all.  I’ve certainly received feedback from managers/bosses I’ve had over the years, just like a received grades and comments from professors in school, but that’s fundamentally different from peer feedback.  You don’t work *for* your peers, you work *with* your peers, and the difference matters because your manager/professor likely values aspects of your behavior differently from your peers (think: obedience vs effectiveness). 

In an effort to solicit more peer feedback in my life, I’m going to link to my Unvarnished profile from the sidebar of my blog. You can leave anonymous feedback for me there. Ideally you would do so after substancial in-person interaction, but if you want to review me based on the merits of online interaction only, that’s cool too. I don’t know if it will generate much activity, but it can’t hurt to try.

I’m not convinced that Unvarnished is the perfect platform for getting peer feedback. It’s the best platform I’ve found so far, but its issues include:

1. All reviews on Unvarnished are anonymous, and anonymity online is a double-edged sword (see: 4chan).
2. Facebook Connect is required to enter (for some good reasons tied to the negative side of anonymity online), but mandatory FB Connect might turn off some users unwilling to connect their account.
3. Sometimes feedback can be more effective when it’s not said in a public forum. I have no problem with negative comments on my unvarnished profile, that’s not what I mean… but sometimes feedback has to be very specific about behavior in a private meeting, which by necessity can’t be public.

But, like I said earlier, I think Unvarnished is the best place to aggregate both positive and negative reviews for people out there today, so I’ll run with it. I’m open to other suggestions.

If you’d rather not send me feedback publicly via Unvarnished, you can always either A) email me — which will be private and I will know your email address or B) you could use the Tumblr Ask feature in anonymous mode — which will be private and anonymous.

Don’t interpret this post as one giant call to action to leave feedback for me right now. Please do write a review now if you feel inclined, but my ideal scenario would be: if after people met me they felt like they had a good forum to give me feedback, perhaps they would. The purpose of this post is to say: there is now a forum available for leaving feedback for me.  The purpose is not to get a deluge of feedback tomorrow.

Remnant Ad Networks pay out nearly nothing to publishers.  They’re also generally quite ugly. The success and prevalence of the “slap the monkey” typical remnant campaigns is case-and-point for me.

And, when you run remnant ad networks on your site, the low-CPM yield from the network undercuts the efforts of your own direct sales teams and premium inventory.  It says to advertisers, “Hey, I don’t think my inventory is worth much more than what AdSense will offer me.” And it also gives advertisers other channels to come on to your site for cheap.

Nick Denton and Gawker knows this well. That’s why they have the Gawker Artists program, and they do not run any ad networks.. They run pictures of art in their remnant unsold inventory in the IAB standard ad placeholder blocks.  The benefits of doing so are numerous:

1) Gawker’s sales team is the only place to buy the Gawker audience.

2) Less inventory supply pushes CPMs higher through classic supply-and-demand mechanics.

3) The Gawker audience has a more pleasing user experience and is less likely to avert their eyes from the ad units.

4) Emerging artists get exposure to a larger audience, which is a nice community service, and good karma to Gawker.

Jason Calacanis makes the argument against remnant ad networks well. He focuses on all ad networks in his post, which I don’t necessarily agree with. But, all of his arguments have stronger force when applied specifically to remnant ad networks.

I don’t know of any publisher besides Gawker that does this.  I think there’s a business opportunity (perhaps for-profit or non-profit) for someone to aggregate either a bunch of charities or artists (or both) and create an entirely free remnant ad network.  Publishers would donate traffic to this network and reap the four benefits outlined above.  Charities or artists would contribute artwork/branding for free distribution.  It would be a win/win, and it could be a very interesting network to run.

Much credit to Gawker for innovating this; this post is largely about taking their great idea and distributing it web-wide.  If anyone working on Gawker Artists wants to chat about this, I’d love to meet up.  Or if anyone wants to take this idea and run with it, please do so.

Review sites of all kinds (music, movies, books, etc…) should provide two reviews for every subject (by “subject,” I mean a specific piece of content: a new book or a new album). One review should be by a serious fan of the artist and one should be a more neutral critic that covers the medium of the subject in general but isn’t a fanboy.

If a neutral reviewer just doesn’t “get” the subject (and often times outright admit they don’t “get” it), the review is useless to fans that want critical analysis beyond superficial understanding. And conversely, if a reviewers is a serious fan of the subject, the opinions of the viewer rarely reflect how people who are uninitiated with the subject will feel about the work of art. Serious fans bring a set of assumptions and expectations that first-time consumers rarely have, and so their experience is often starkly different. So, each side of this dual-review structure is useful to a different audience.

Sure, creating twice the number of reviews would be twice the work, which would make this idea difficult to implement for a review publisher. But as review aggregation sites like Metacritic drive individual reviews towards commodity, this could be an interesting point of differentiation that would not be easily aggregated up.

I recently changed my bio on Twitter profile, just on a whim. When I did it, I realized that probably no one would even notice the change because A) who reads Twitter bios and B) even if someone did read it, they wouldn’t know it used to be something different a short while ago.

The activity around a users account is currently implicit data (and metadata) that could be made more explicit to increase information, connections, and utility in the network.

What does that previous buzzword-compliant sentence mean?  Well, lets use a corollary example: Back in 2006 Facebook used to be completely profile-centric. You’d go visit the walls of your friends to see who was saying what to whom, who posted new pics, etc… there was a time when there was no news feed, no way to know what was *new* on the service.  You simply had to navigate profile pages (much like MySpace still is) in order to get any utility from the service.

Then, to the sounds of significant initial user backlash, Facebook created a news feed.  The news feed surfaced both data (such as my new relationship status) and metadata (the fact that my relationship status changed…  which is a subtle but *important* difference from just the relationship status alone). The surfacing of previously implicit metadata was a gray area with respect to users’ privacy, but in the long run it was a great feature change that increased both usage and utility of the service.

Twitter has similar, hidden implicit metadata, and they have the same opportunity to make that data more explicit and useful.  For example, when I follow someone new on Twitter, that could be a piece of data that could surface in my friends’ streams (or perhaps in a separate, metadata-only news stream).  Or when I change my bio… or update my avatar… all of these are events that could be surfaced. 

A third-party could use Twitter’s API to build this feature I’m describing, but it would require significant caching and constant polling to determine when anything changes in a user’s network, which sounds like a real pain to implement. So, ultimately I think this is Twitter’s opportunity to own. 

Disclaimer: Though I work for Spark Capital, an investor in Twitter, I have no idea if Twitter is thinking about any of this stuff or not.  This is strictly my own opinions, ideas, opining and does not represent my employer in any way.