Starting Up by Rahul Pathak

Tom Evslin has a great blog and I just came across his post about the opportunity in the local internet. While I’ve written before about the challenges in building a local business that scales, Tom makes some great points about why the potential in local online is so great.

Tipping points happen all at once and they’re only visible with hindsight.  We’re at a tipping point now: three-quarters of US Internet users or half of all US households have broadband connections. Not all of these connections are equal but most can download and watch videos, surf complex websites, and search the Web itself at a comfortable pace.  The importance of these statistics is that there is now a critical mass of people who can and will take advantage of and interact with rich LOCAL content.

The opportunity, if you’re an entrepreneurial sort, is that there isn’t much rich local content to interact with for the very good reason that, until now, the local communities were below critical mass of enabled households.  It used to be that the big bucks were in developing and enabling services with large audiences drawn from a worldwide or at least national prospect pool.  Now the opportunity is to bring the benefits of the interactive Web to every local community.  It’s huge!

We’ve learned  a lot about the local segment at Judy’s Book and I think Tom is right about the value of rich local content to consumers. The challenges of aggregation and fragmentation still apply, but it can definitely be done. I think winning approaches will be those that figure out how to crack the local aggregation nut - both from an audience and a content perspective.

I just came across this great post on ProBargainHunter about Modoshi and Dealspl.us. Yan talks about the time needed for community building and the commission vs. the flat fee approach to getting users to submit and scrub deals on the site. There are some great ideas in this post that are especially topical given that we are striving to combine savings and community here at Judy’s Book. Pure community approaches are challenging and the key to incentive systems is to be aware of unintended consequences. People are quick to optimize against whatever model you put in place and if you’re not careful about measuring and adapting your system, you’ll quickly find yourself paying for contributions that don’t create any value. I realize this is sort of a ‘duh’ statement, but you’d be amazed how hard it is to implement.

Fred Wilson gave a talk entitled “Does Information want to be Free?” in which he talks about the value shift from information to attention. He makes some great points and I would urge anyone in and around the content business to take a look at his presentation which is available on the web.

Creating ways for content to be pushed off websites and consumed in whatever manner users want is definitely the way forward. Putting metrics and monetization around these content bites also makes a ton of sense. Ultimately, getting into user flow is critical. It’s far better to make information available to people wherever they want it rather than to force them to come to your site and do jumping jacks to get it.

I think there are a couple of dimensions to data portability. One concerns consumers of the data and that’s been discussed. The other concerns creators. They should have portability as well. Why should they be locked in to a particular site or publishing platform. Dave Naffziger had some good things to say on this subject a little while back and it was around my move from Blogger to Wordpress.

In an ideal world, people would create data wherever they wanted and it would be available for consumption wherever needed. Fred’s assertion that RSS is the glue is right on the money. Having said that, I think RSS needs to come up with simpler language to get beyond the blogosphere and those of us in the industry. (See Andy’s post about RSS & “Ask a Skeptic’s Mom”)

For fellow flight nuts out there, I came across a cool site (via doggdot.us) which provides an interactive view of a B777 flight deck. Enjoy.

Kathy Sierra’s post on putting some randomness in your product is a great one. I think she hits on some great ideas about putting something unexpected in the product to delight users while still striving to create a product that is largely predictable and therefore allows the user to focus on their task and not on the software.

A couple of her suggestions really struck home and got me thinking. They would be great to implement at Judy’s Book or at any other user-generated content site for that matter. They are:

• Staff Picks - Business of the week, photo of the week etc.
• Unexpected Uses of [x] - Truth is stranger than fiction. I bet this would generate really cool and entertaining content
• Something unrelated - Occasionally introduce things that aren’t related to the primary purpose of the product

The focus is on providing consumers with something unexpected while they interact with your product and it makes a ton of sense. Kathy’s blog is full of gems like this post. Highly recommended for anyone interested in designing cool products.

Dave has a series of great posts on his blog about things that are “Interesting, Cool and Useful” The posts themselves are informative and entertaining - always a killer combination. Highly recommended.

I was told about two great posts today. The first is by Peter Rip at Crosslink Capital about “failing fast and failing often.” You have to get into the market, see what works and adapt.

The classic venture model has been to fund to milestones 12-18 months out. In consumer web services, there are only two meaningful milestones — (1) are you getting a lot of users and (2) have you figured out how to make money? We use other metrics in other sectors (like management, product, etc.) as proxies for real economic progress. We also use them because (we believe) they would have residual value in an asset sale or merger.

None of this is true in consumer web services. You’re either hot or not. Second place generally sucks.

The problem is that it is hard for entrepreneurs and VCs to know a priori if something is going to be a hit. The only way to know is to try, and trying takes time and money. So here’s the real rationale for what it makes sense for these companies to raise “a lot of money” and not blow it. They have to run lots of experiments.

One thing I would add to this - when you’re in a mode of getting into the market quickly and iterating, you have to do enough work to make sure that what you present to the market is good enough that their reaction to it is meaningful. If you release a buggy piece of crap quickly, you might conclude that no one cares. The flip side to this is that truly great ideas find a way to win, but in general, not shooting yourself in the foot is a good idea.

The second great post is on Mark Cuban’s blog and it’s about his relationship with Bobby Knight.

You said, and Im paraphrasing: “Everyone has got the will to win, its only those with the will to prepare that do win”

Clearly, a lot of factors go into success and even the definition of that word is a personal one. There’s no question in my mind though that the will to put yourself in a position to win and the ability to evaluate your situation and change what’s not working are critical.

This post on marketing innovation in the coffee market was sent to me and I had to share it:

Sex sells. Usually, it sells things like beer, various liquors and - of course - fatty fast food hamburger, but some Seattle area businesses are using it to sell something else entirely: coffee.

Regardless of your opinion on this particular form of marketing, I’d like to thank Nicole Weston for the term ‘Sexpresso’ and for providing evidence that you can de-commoditize anything.

There’s a really interesting article in the WSJ this morning about really walking a mile in your customer’s shoes. Or, in this case, viewing a scene through his glasses.

Mr. Sato is head of advertising and investor relations at Japanese eyeglasses retailer Meganesuper Co., which requires employees to wear glasses at work. “I change them every day,” says the 50-year-old, sporting a pair of black-and-white frames that clash with his navy-blue striped suit.

The everyone-wears-glasses policy at Meganesuper (literally, “Eyeglasses Supermarket”) is an extreme solution to a universal management problem: How to get a handle on customers’ needs.

Every employee, even those with perfect vision, is required to wear glasses. While this example seems a little extreme, using your own product regularly is the only way you’ll experience what your user does. That’s when you find ways to improve your product and processes.

Brad Burnham has a great post (which I found via Fred Wilson’s blog) about what’s coming next in the inevitable progression of the computer industry. He argues that we’re in a world where we no longer care about hardware or software but rather we care about data (Think Facebook, Craigslist). After data, he suggests that governance (how to get data that delivers the most utility for the least overhead) and values are coming next. He makes some very compelling points and I highly recommend the post.

One area where I might put things differently is to suggest that at any given time, we care about different parts of the stack in different sectors. For example, the iPhone, computers in cars, connectivity on airplanes. There are a number of sectors within the industry where the data we’re trying to get to is known and the value is in the boxes and pipes. It rarely stays there though - I completely agree with Brad in this regard - value shifts up the stack quickly. As soon as the hardware and software become transparent, the data and it’s quality become the only things that matter.

Anyway, that’s neither here nor there - this is a post about governance. At Judy’s Book, managing the quality of user-generated content is something we’ve been talking about lately and Brad’s post really hit home.

The genius of Craigslist is in its governance system. It is its lightweight governance system that allows 21 people to administer 300 sites in 35 countries. I believe that the basis of competition in web services will shift from the data to the system that manages the acquisition, and use of that data. The governance system that yields the most utility for the largest number of users with the least overhead will ultimately manage the largest communities with the most valuable data.

I also came across an article on Reddit a few days ago which talked about the early days of HotOrNot.com and some of the scaling issues they had to deal with. One section in the article focused on how they created a system for editors to quickly determine whether a picture was obscene or not.

Scaling the Human Element

We had another problem with some users submitting pornography and other inappropriate photos. Initially, we decided to solve this by adding a link under each photo that said, “Click here if this picture is inappropriate.” If a photo received enough clicks, based on a formula we had derived, the picture was removed.

This worked pretty well, but not well enough. I sent the chairman of a large advertising network a link to our site with a note proclaiming that: “The odds of getting an inappropriate picture are extremely low.” Ten minutes later, I received his reply: “Unfortunately, the first picture I saw was that of a topless woman.”

He informed us that if we wanted companies to advertise on our site, we’d have to filter each picture as it came in. Jim built an interface for us to do so. However, we soon realized that we couldn’t spend all day screening pictures. The system’s human component wasn’t scalable. That’s when we arrived at the moderator idea.

We decided to build a system in which moderators could vote on whether to approve or reject a picture before it passed on to the main site. If a picture got enough votes either way, it was approved or rejected. By making the decision collective, no single moderator could approve or reject a picture independently.

To help detect any rogue moderators, the system tracks each moderator’s accuracy. A vote is counted as wrong when the moderator’s vote goes against the final outcome of the picture. For instance, if one person votes to approve, but all others vote to reject, the one person is wrong. Moderators whose accuracy ratings drop below our threshold are kicked out.

We decided to take the moderator system one step further by adding security levels. The higher a moderator’s security level, the more his or her votes counted. We also gave higher-ranking moderators special privileges, like an expert mode in which they could judge pictures much faster. We gave the highest ranking moderators the ability to reject or accept moderator applications, and the ability to kick out rogue moderators. Today, these top-level moderators essentially run the moderator section of the site and decide on the specific guidelines for what makes a photo inappropriate. More than 1000 moderators are currently active, and they form our strongest community.

It seems clear that some combination of technology and human input is needed to create a system that is reasonably scalable without having to invest a crippling amount of money in a pure technology system. One of the problems with simplistic algorithms (’x’ votes, ‘y’ comments, user rating over ‘z’) is that it’s incredibly hard to tune.

Another question to consider here is the notion of democracy versus editorial and how to create a balance within the two. A pure democracy is hard to achieve online - given that there will always be bad actors in the system and that the good actors will have varying levels of engagement (think about the 80-19-1 rule) - you will always need some measure of editorial control. This could be via a group of users who are engaged and are given tools and special interfaces to control content. Even Digg, sometimes the poster child for democratic content control puts a fair amount of editorial power in the hands of power users. Pure editorial seems great but has one major problem - it’s hard to scale.

The truth of the matter is that different strategies will work for different businesses. One thing is clear though - if you’re in the user generated content business, finding a governance system that will let you maintain the quality and tone of data you want in a scalable fashion will be a competitive advantage. Presenting new users with information they find valuable earlier in their interaction cycle makes them that much more likely to stick around, participate and contribute their own $0.02.

I realize I’m not directly debating Brad’s point here, but it got me thinking, which in my mind is the hallmark of a great post. There’s always something coming next and that’s something I love about this industry.