2 days at Mesh, a roundup

Over the past 2 days I was at the Mesh, Canada's Web Conference, in Toronto. Overall it was a great time and a lot better than I was expecting. To be perfectly honest I didn't pay too much attention to the sessions, other than the keynote talk by craigslist CEO Jim Buckmaster. It was great and you could tell that they've got an awesome thing going on there. Some tech specs since I'm a nerd: they're serving 7 billion page views a month on 200 servers. And over the past 12 years they've never had a programmer leave. That is just crazy.

The best part of the conference was meeting various people in the community and putting faces to the sites that I'm familiar with.

Robleh Jama (Sneaker Play) - cool and smart guy that just graduated Uni similar to myself. They're getting great traction and attention so it was cool to talk to him about their accomplishments.

Alex de Bold (ChickAdvisor) - I joked with him a bit accusing him of stealing the "love it" idea :P

Scott Brooks (ConceptShare) - I've known Scott for a while now, and he's a great guy and Concept Share is doing fantastic, which is amazing for a web company out of Sudbury. If they keep it up maybe Sudbury really will be the next valley HAH!

Sean Wise (Wise Words) - one of the people behind Dragon's Den which I friggin LOVE so it was very cool meeting him. Sean also just released a new book Wise Words, and is a columnist for the Globe and Mail.

Saroop Bharwani (Ogrant) - another person I've met with before, but we talked for a while at Mesh. These guys haven't launched yet, but they're doing something pretty cool and I'm rooting for them. I'm not 100% convinced of the idea, but Saroop and his partner impressed me with their intelligence and business savvy so I'm sure they'll figure it all out.

Anthony Carbone (MadWhips) - another guy I know well so it was good to see him again and chat it up.

Simon Pulsifer (QuillSoft) - the King of Wikipedia, this guy has 96,000 edits to Wikipedia which makes him the most active contributor in the world, so it was very cool to meet him. We talked for a while and he gave some of this thoughts on ProductWiki which were quite helpful.

Sunir Shah (FreshBooks) - one of the first people I met that is heavily into the wiki world it was great seeing him again. Sunir now works for FreshBooks and they were interviewing some of the startups to profile them on their blog. He interviewed me, so hopefully we show up!

Now, after 2 days of web and tech and startup talk I'm officially burnt out. I don't want to hear people talk about VCs, or bootstrapping or anything like that for a little while. Time to get back to products!

Toronto DemoCamp13

After attending my first DemoCamp (number 12) and seeing the other very technologically knowledgeable attendees, I knew that this would be a great venue to talk about ProductWiki. Just a little bit of background, DemoCamp is an informal get together of tech enthusiasts, designers, etc where certain people get to present something that brings value to the audience.

I contacted David Crow, one of the organizers, to see if I could present ProductWiki at the next meetup (numero 13), and we were selected to present, woo! Now that presenters are given a earnest 5 minutes of face time, we're all required to write a blog post answering the 6 main questions that the community wants to know. Well here they are!

Have you attended a previous DemoCamp?
Yes, DemoCamp12

Who are you? Previous experience, what makes you qualified for us to listen to, etc.
I'm cofounder and CTO of ProductWiki. We've been working on this for 2 years now and received attention from a variety of media outlets including popular bloggers such as Mark Evans. And the fact that I'm good looking helps.

What does your product do?
ProductWiki is attempting to create a comprehensive resource for high quality product information. Using the power of the community, and volunteer contributions we want ProductWiki to house FREE product information: FREE from bias, FREE from baseless marketing speak, FREE to use by anybody for anything, FREE from the noise of clutter and spam, FREE from editorial red tape so that it always improves and stays relevant.

What hard problem, interesting insight, or cool feature will you be demonstrating?
The project of creating a high quality, comprehensive resource for product information is an incredibly daunting task. Information-wise: what is a product? What is high quality information? How do you have the benefits of many bits of information without being overwhelmed by the noise?

An example of an innovation that came out of answering this question is the collaborative review system, where many people can come together and review a product individually while creating one authoritative review. This is an example of what I like to call Accordion Information.

What are you hoping to get out of presenting?
Feedback and advice on how to communicate our message and techniques on how to identify the right people that can become a core part of the project. Also (and I know this is sales pitchy), if ONE person is able to see our vision and how it'll effect the world, and becomes a contributor, then it'll be worth it.

What does the community gain by hearing you present?
What is Accordion Information and what are its benefits relative to other kinds of information systems (especially the user generated kind).

Learn from our experiences in growing a community and that it's all about the core people.

Get people to appreciate the problem of low quality product information, and the possibilities that come about when high quality information is FREELY available.

2 months in the trenches

If you haven't been to ProductWiki in the past little while then go there www.productwiki.com. Now! Check it out! It's great! We just released a new update on Monday that introduces a bunch of new features. For the full details check out the notes on the site. In this post I'm going to run through some of the major changes and talk about why we did them.

Before getting into the details, here's a sample of some of the buzz that's been generating.

Download Squad - a very positive assessment of the site that appreciates the unbiased and comprehensive nature of the information

Maple Leaf 2.0 - an indepth interview with Erik, discussing the history of our site and the new features.

Mashable

Emily Chang's eHub

The Collaborative Review system is something we're super excited about, and it looks like it's already a success. Community members of the site probably already know that there was no formal review system with stars and all of that. There's the tapping system and you can always write reviews in the comments, discussions or as an article, but nothing like Epinions or Amazon's reviews. This is actually very deliberate on our part, we did have one of those systems working before we launched in 2005. But even back then we knew there were problems and so didn't include it as part of the site.

Though we didn't have an advanced review system we never stopped thinking about product reviews and how to do them properly. In fact I even wrote a pretty lengthy article examining the problem with video game reviews. Looking at expert reviews I saw that a lot of them contained the same information, a big hunk of text devoted towards describing the product, with some superfluous information thrown in, and the useful opinions at the end. User reviews, the good ones, followed the same model. The short user reviews are useful but you have to read a lot of them to build up a comprehensive picture of the product. Lastly, the most important parts of the reviews are the ending pros and cons where people say what benefits the product brings to their life, and the cons which are the aspects of the product that should be improved. What ends up happening is you read a ton of reviews and start building a list inside of your head.

Eureka!

Let's just do this FOR people explicitly. This helps the person writing the review, and the person reading the review. When you're writing the review you don't have to reinvent the wheel and come up with points that already exist, if someone has already added that pro or con, then just agree with them! Or disagree if you experienced something different. And for researchers now you have the convenience of reading one review while getting the confidence and coverage that many people's perspectives can bring. And if you still want to see how an individual feels we need to have that as well since it's still useful for a lot of people. Combine the best of both worlds.

It all sounded nice in theory, but now that we've launched and seen people use the system, it looks like the reality is following what we thought would happen. Which is great! Because as anyone in software development knows, the way you THINK people will use a service is often different than how they ACTUALLY do.

The second major new feature we added is OpenID support. When we first learned about OpenID it was kind of confusing, I mean what the heck does a "decentralized, distributed identity system" mean anyway? But after wading through the technical mumbo jumbo it turns out to be a pretty simple concept. All it is, is that you have a login account with one website which gives you a name, and then you use that name on other websites that support OpenID and that's it.

Simple right?

Heck, after dealing with some poor technical issues and documentation, supporting OpenID on our end wasn't that big of a deal either. The biggest annoyance was the lack of a proper testing server that we could experiment against, instead it was 'try this'... nothing happens 'try something else'... nothing happens. Which can have a dramatic impact on development times! Overall we're extremely happy with OpenID. I was skeptical going into it, but after getting it working and using it internally I started to appreciate how easy it all worked. So far it's been a success. Quite a few of the new sign-ups are OpenID people. And considering the OpenID folk are usually the ones that like to try out new technologies, I say welcome!

We've also incorporated some other features into the site that should improve the overall experience. A couple included some functionality that the community has requested including being able to filter your user list by product category, and deleting your own comments.

It's been a long and busy road so we'll chill out on the hardcore coding for a while and focus on community building. We hope you enjoy everything we've worked on, and as always if there are any questions, suggestions or concerns let us know!

An important note: we're looking for some moderators, so if you're interested shoot me, Erik or Amanie an e-mail and we'll start that process.

The Internet went down yesterday, and nobody noticed

I was surprised to find that no one could access ProductWiki for about an hour yesterday at 1pm (EST/DST). I did some research and found that GoDaddy (our registrar) was having "technical difficulties". It turns out that a whole whack of websites were inaccessible for a short period yesterday. It seems likely that this was connected to the new DST time shift that had happened the night before. GoDaddy said they were ready, but it seems not.

A few things that I thought were noteworthy about the event:

• a huge portion of the Internet was down yesterday because of a single point of failure
• a ton of people submitted the story to digg, and, surprisingly, it DIDN'T make it to the front page
• GoDaddy covered it up very well; they only once, on one page, mentioned they were "having technical difficulties" without any information as to why, or when they would be coming back up
• No mention of it on Bob Parson's blog
• This event made DST 2007 more significant than Y2K (which is kinda pathetic)

How does this happen? How can, in the "new media", a company like GoDaddy get away with a blunder like this without any viral bashing? No YouTube videos. Very few blog posts. No caustic commentary on digg. Nothing.

I think this event, or non-event, might trigger a shift towards websites using other name registrars (I know I'm considering switching). It wasn't that significant in terms of downtime, but GoDaddy's response, or lack of response, is telling.

UPDATE: It wasn't related to DST after all. It looks like a good ole' DDoS was the culprit.

FLIRTing with the community

I found a great blog post by Sami Viitamäki recounting the outline of his Master's thesis on the topic of customer collaboration.

In an earlier post, I dismiss the value of crowdsourcing. What I really was dismissing was the form of crowdsourcing that doesn't take what Sami is talking about. A major tenent of the FLIRT model that he's outlined is the idea of concentric rings of ever increasing membership, but with less involvement. He defines them as:

• Creators: those users who create original content
• Critics: those who criticize, and evangelize the work of the creators
• Crowds: confirm the work of the creators and critics through simple interaction (voting, comments) making the information accessible to the community at large
• Community (not in the diagram): everyone else

In trying to foster a community and grappling with the challenges of a collaborative website, I found this model to ring true. As I look around the web space and see a ton of UGC websites trying to make a buck off by leveraging the work of the crowd, I realize that without considering the true nature of collaboration, most will fail....

You know you're a web developer when...

• you can tell the difference between Arial and Helvetica
  
• your right-click in Firefox brings up a menu longer than (and wider) than your dual monitor setup
  
• 1px of margin can make or break your day
  
• you see colour in hex (it's not light grey, it's #BBBBBB!)
  
• corollary: you can tell the difference between this and this
  
• you check Alexa (or alexaholic) more than your bank balance
  
• Red Bull and Full Throttle are two of your basic food groups
  
• it drives you crazy that Myspace, craigslist, and plenty of fish are so successful in spite of their ugliness
  
• IE6 is like your creepy uncle; you don't like him, you don't want him around, but you still have to deal with him
  
• Google is simultaneously the most loved and most hated entity in your Universe
  
• corollary: you say "Google" more often than any other word in the English language 
  
• you want to get bought by Yahoo! before Google edges you out of your market
  
• you know your users by IP address
  
• you no longer think TechCrunch is relevant, but would kill to have Arrington write about you
  
• ditto for digg
  
• you register for every Web 2.0 site and use them for an average of 8 minutes
  
• you judge people by their site's pagerank (or number of comments)
  
• if someone brings up the topic of PHP/.NET/Ruby they better get comfortable, because they're not going anywhere for awhile
  
• you run a hot, up-and-coming startup, but your parents have no idea what you're doing; they tell everyone you work "on computers"
  
• you create a blog post entitled "You know you're a web developer when..." instead of working on your hot, up-and-coming startup.

The Wisdom of Crowds? Nay!

It's the wisdom of a few exceptional people. I think I need to delve into the mathematics of user-generated content (UGC)...

There is a buzz about the "wisdom of crowds", everyone seems to cite Wikipedia and digg as examples of crowd of people collaborating to produce high quality sources of information. The fact of the matter is that most of the content on sites relying UGC is created by a small minority. Jimbo Wales has said on many occasions (including in this month's Business 2.0 magazine) that 50% of Wikipedia was written by a 1000 people. This is in direct opposition to the commonly held belief that Wikipedia is an encyclopedia written by millions of people, each adding a small snippet or factoid of information. digg also has similar statistics. Very similar, in fact. seomoz.org pointed out that 56% of digg's front page was contributed by the top 100 diggers.

Let's dig deeper into the numbers of these two prime examples of websites that depend on UGC. Based on some rough alexa analysis, and commonly accepted figures, Wikipedia gets in the range of 10 million unique visitors per day, while digg gets in the range of 1 million. Either way, the approximate ratio is that Wikipedia has 10X the traffic of digg. This is the exact same ratio as with the core contributors. Does your site's readership directly relate to the number of core contributors?

Website	readership	contributors	% of content	readers / contributor
wikipedia.org	10,000,000	1,000	50%	10,000
digg	1,000,000	100	50%	10,000

If instead of focusing on the size of your crowd, you focus on your core contributors, the content of these users can support 10,000 (mostly non-contributing) readers. How does this happen? I believe that it's the exceptional intelligence and wisdom of the passionate that is able support strong readerships. (Not the mediocre intelligence, and wisdom of crowds)

How does the trend of crowdsourcing fit in to this model? Well, crowdsourcing involves paying (often very small amounts of money) for the content produced by crowds of people. This model does not likely lead to exceptional content by passionate people, but rather acceptable content by people motivated largely by money. This type of content, I believe, does not support the readership ratios seen in Wikipedia and digg, and you can verify this by visiting some sites that are known to crowdsource (without mentioning any names).

I put it out to you, that instead of budgetting for crowdsourcing, your money would be better spent catering to your 1% of passionate users. How do you that? This blog is a good start....