Yesterday I attended a panel which I've been thinking about constantly, ever since the panel ended. The panel was entitled "Why Marketers Need To Work With People Media" and it had a set of rock star panelists (in the SXSW interactive, not music, sense of the word!). The participants were:

• Tony Conrad of Sphere
• John Battelle of FM Publishing
• Toni Schneider of Automattic

While the panel referred to it as 'people media', it could also be described as 'social media,' 'read/write media', 'conversational media' (as John Battelle has been describing it lately) or even 'user generated content' [Ed: that last term is beginning to get unpopular]. Regardless of what you choose to call it - and I'll use the term 'People Media' in this post because of the panel title - these sites are a significant part of online traffic and so delivering high-value advertising is key. Note that John Battelle covered most of the material from the panel in a recent post. 

One of the interesting things about Federated Media is that it gives blog authors the ability to reject advertisers from running ads on their site. In the post mentioned above, he explains:

"The approach of having the authors approve the companies which advertise on their site seems obvious, but when you think about how traditional media works, it's downright revolutionary. The lunatics were running the asylum! But it turns out, when an author approves a company to advertise on his or her site, they are, in essence, inviting the company to join that sites' conversation. A permission has been given, a trust established. To this day, every single ad FM sells is approved by our authors before it appears on their site. And despite our initial worries that author approvals would be a hurdle to marketers - after all, they're used to getting their way - it has, in fact, turned into an overture, a conversation starter that has led to all sorts of examples of new approaches to marketing online."

John provided an interesting example of one of his bloggers rejecting an ad. Cory at BoingBoing has been a very vocal opponent of DRM. When a pro-DRM group approached Boing Boing (via FM Publishing) to run a $35,000 a month ad, they turned it down. At this point, this is the largest advertising opportunity that has been refused. However, since launching, John estimated that only 1% to 2% of all advertisers have been rejected. In addition to these small percentages, many specific ads have been refused until a better creative was developed - which leads to the next point.

Evolution of Advertising Creative

In John's post he wrote about the creatives of 3 campaigns which actually leveraged the conversation. They included an ad for Dice.com, which used JavaScript to ask "Why does your job suck?", and then allowed respondents to enter their answers into a text box - which ended up displaying in the ad. On the panel, John said that currently only about 15% of ads on FM Publishing leveraged unique creative like this. John said that these ads are performing very strongly and maintaining their performance. Typically, a display ad's performance (regardless of interaction metric) degrade over time as visitors to a site grow used to seeing them.

Summary

The web is only going to become more social. However, the dirty little secret is that those pages don't monetize nearly as well as many other pages - which are either contextually rich (in some areas) or the direct result of a user request (such as search). So the challenge is to figure out how to adequately compensate the individuals behind 'people media'. John made an interesting point about the gap in time between search engines emerging as the method of navigating the web, and those queries being monetized effectively. Today, search advertising is a multi-billion dollar business. To do this it took the correct metric (Cost Per Click) and correct ad unit (text ads). 

I'm confident that as the interactive advertising and media industries wrestle with the challenge, we'll come up with the correct metric and ad unit for advertising on people media. However, just like it took time in search - it will take time with people media.

This morning I attended a presentation by Matt Mullenweg, the Founder of WordPress, on 'Scaling Your Community.' Matt started by defining scaling your community as "being as useful to the last 100k people as you are to the first 100k." He talked about the four steps a community goes through as they scale.

The steps he laid out for leaders of a community were:

1. Build a Good Foundation
2. Bootstrap
3. Let Go
4. Personalize

Step 1: Principals for Building a Good Foundation

The principals that Matt emphasized to build a solid foundation all revolved around taking a simple idea and articulating it clearly. Also he said that once you, as a leader, commit to a set of simple principals, it is essential to make sure all of the product development is consistent with that vision. Matt made the point that while this is very straight forward, it is often difficult to execute on. For example, at Word Press they have committed to "all free features will always be free". Unfortunately, this has proven difficult at some points because users have leveraged the system in ways they never expected (such as unique ways to track their blog's activity). Because of their commitment, they have decided to continue offering those features for free.

Matt encouraged leaders to do three things during the 'bootstrapping' phase of community development:

• Be your most passionate user;
• Talk to people (like an earlier panel, he pointed to the Flickr example);
• Finally, he admitted that while somewhat controversial, he encouraged leaders to 'pre-moderate'. He explained that "there is a point in every community where the signal to noise ratio degrades ... you need to pre-moderate before comments are publicly available to protect the community from this degradation."

Step 3: Let Go

Once you have an active community, he explained that one of the most difficult things to do is give the community over to its members. He talked about how he has recently done this with WordPress Ideas, a website that allows people to enter their ideas for new Word Press Platform features and allow the community to vote on them (think Digg for features). So far the site has received over 487 ideas. He encouraged other open source project leaders to let go, because Open Source "is more than just providing software and source code" - explaining that it was also about opening up processes and decisions.

Step 4: Personalization

Finally, communities need to encourage their users to express themselves as individuals. Websites need to to treat "every tag and click [as] sacred." He discussed how interacting with a personalized site can build stronger affinity. He pointed to a recent article from The Onion about the effectiveness of Amazon's Recommendation Engine. Matt explained that systems like this really help encourage your community to spend more time interacting with your site. As an example, he relayed a story where he spent 8 hours on the phone with Amazon trying to merge all his accounts together.

Conclusion

Obviously, this is something Matt knows a lot about - having built a very active open source development community for the Word Press Platform. He also has helped create a destination site that hosts 3/4 of a million free blogs (wordpress.com) which have more than 35 million unique visitors a month.

Each of these four steps are critical as your community evolves. Hopefully, you're able to diagnose where your community is and have identified a few techniques to help you get to the next level.

On Saturday we covered an SXSW panel called Web App Autopsy, which examined four live web applications (RegOnline, FeedBurner, Wufoo, and Blinksale) for things like conversion rates and revenue per customer. Today there was a similar seminar entitled "Barenaked App: The Figures Behind the Top Web Apps", which looked at 5 web applications and what it took to build and release those products. This time the focus was very much around the financial costs to build and deploy these web applications (as opposed to elements like lines of code or revenue). It also touched on what it costs in monthly maintenance. Here is an overview of the data they shared:

• DropSend: Build $48,012 / Monthly $3,625
• Freshbooks Build $20,000 / Monthly $46,000
• Maya's Mom: Build $70,000 / Monthly $30,000
• Mobissimo: Build $60,000 / Monthly $150,000
• Wesabe: Build $200,000 / Monthly: $3,000

One other interesting thing shared was that FreshBooks cost $430k total to build and maintain, until they reached break even ($140k of those expenses went to marketing).

Photo: Ryan Carsen

Today's keynote was a conversation between Limor Fried of Adafruit Industries and Phillip Torrone, the senior editor of MAKE magazine. In the conversation, they discussed a new movement called 'Open Source Hardware'.

Definition of Open Source Hardware

Open Source Hardware involves releasing all of the information necessary, to allow individuals to acquire the individual components of a device and understand how to assemble them together into a functioning device. Also, many individuals end up then extending a device to incorporate entirely new and novel uses. Limor talked about multiple types of open source hardware - including releasing the following:

• mechanics in an open interface markup file under creative commons;
• circuit level design in an open format released under creative commons;
• firmware source code (more like traditional open source licenses);
• data sheets and parts lists including where to find the parts (it is increasingly becoming hard to find each of the individual parts);
• finally there is the firmware and APIs, again under a more traditional open source license.

Limor and Fried reviewed a number of examples, including:

• The LinkSys WRT54G Line of Routers - provided the ability to upgrade the $75 router's firmware to make it significantly more valuable;
• Roomba - Provided access to the Hardware API, so that you can integrate additional hardware into the robotics platform. This was so successful, that iRobot just released an Educational Bot which is the Roomba without the vacuum.

Business Model

I found each of the above examples interesting, because they are 'for profit' companies leveraging 'Open Source Hardware' projects. Just like for profit businesses have leveraged Open Source Software.

Interestingly, Fried's own company Adafruit Industries actually provides kits with all of the components necessary to complete the electronics they develop and release, under creative commons with attribution license. According to Limor, this has actually become a profitable business; because people are interested in building these electronics, but they don't want to do the work to track down each component, and are willing to pay a premium to get all of the components from one source.

Conclusion

There was a lot of discussion today about the advent of "prosumers" in media and specifically web media. Today's presentation was very interesting to see how this same trend is being applied to another aspect of technology. I was actually tempted to call this post "Read/Write Electronics" :-)

This morning, I attended a panel titled "Using RSS for Marketing". The panel had a great set of participants including: Tom Markiewicz CEO, EvolvePoint (moderator); Emily Chang Co-founder, Ideacodes; Bill Flitter Chief Mktg Officer, Pheedo Inc; John Jantsch Owner, Duct Tape Marketing; Greg Reinacker CTO/Founder, NewsGator Technologies Inc.

Tom's style of facilitation (at least for this panel) guided the conversation to cover a broad range of topics extremely quickly. However, at a high level, the panel discussed:

• Where are we in terms of user adoption / understanding of RSS
• Reasons marketers should syndicate content
• What are marketers and publishers doing wrong?

The panel all agreed that user adoption of RSS is continuing to grow. They also all agreed that inclusion of an RSS reader in Internet Explorer 7 has helped increase adoption. Bill pointed to some research showing a 500% growth in the automotive vertical in 2006, as one example of this growth spreading outside of the technology and early-adopter crowd.

However, they also agreed that most people don't actually know they are consuming RSS content. Greg Reinacker did an excellent job summarizing the consensus of the group when he stated: "It is not about RSS at all, it is about subscribing to content". Yahoo and Ipsos did some interesting research (pdf) in Oct 2005 that showed there are a large percentage of unaware RSS users. Apparently, this group is continuing to grow in size.

Interestingly, Bill commented that even in pitches to advertisers, Pheedo has stopped talking about RSS and now just talks about 'content distribution' or 'syndication'.

Reasons Marketers Should Syndicate Content

The panel went through a number of reasons why marketers would consider syndicating their content via RSS. The reasons tended to fall into the following 'meta reasons':

• It's an easy and faster way to deliver information to their customers and other audience members;
• Marketers are becoming more like publishers (here's a good description of pubvertising - nb: free acct required); the panel said RSS is a great way for marketers to participate;
• It is a very easy way to optimize your content for search results, because if your feed is optimized then the content is crawled and archived efficiently for search results.

What are marketers and publishers doing wrong?

Bill talked about publishers "being too stingy with their content". He indicated that most want to restrict the feed to partial text, even though in the research they have done the difference in response rate (clicking back to the website) is not statistically significant between full and partial text. In addition, he talked about the importance of adding your company's name to posts and potentially even as part of the title (like the AP feeds).

John talked about making sure your content is easy to subscribe to. Going back to the state of adoption, he encouraged all bloggers, publishers and marketers to make it very easy to subscribe. He even encouraged marketers targeting lower-tech audiences to include a video or page description, explaining exactly how to subscribe to your content.

The panel also touched on the limited amount of tracking that publishers and marketers do on their feeds and the importance of including basic analytics. Note: I discovered very similar things in a yet to be published second phase report that I prepared for the Newspaper Association of America, on 'RSS Next and Best Practices'. Here is a link to a brief description of the full project, on my personal blog. The second phase involved surveying 70+ newspapers about their use of RSS.

The last panel I attended on the first day of SXSW was entitled "Web 2.0 and Semantic Web: The Impact on Scientific Publishing". The panel was moderated by John Wilbanks from Science Commons. John did an excellent job showing how a number of general internet trends are effecting scientific publishing. Specifically, I was impressed by three major projects the panel touched on:

• Open Access
• Connotea
• SemanticWiki on People at Ontoworld

Open Access

Melissa Hagemann, the Program Manager for Open Access at the Soros Foundation, was one of the panelist. If you aren't familiar with Open Access, it is an initiative trying to get all scholarly research available for free on the Internet. Melissa explained that many supporters are now requiring the research they support via grants, to be provided under Open Access. In fact, the US Congress is considering legislation to require all research supported by the United States Government to be distributed under Open Access. The reason is that the government and other funders are realizing they are paying for the research twice. Firstly, they are paying to have the information created and synthesized for publication. Then they are paying again to allow other researchers to get access to that knowledge.

Timo Hanay, Director of Web Publishing at Nature Publishing Group, discussed the beginning of an evolution in how the impact of scientific research is measured. Historically, the impact of a piece of academic research is measured by the journal that the research is published in. However, that is evolving - the impact increasingly based on how the information is distributed. One of the examples Timo showed was Connotea, which Timo described as a "del.icio.us for scientific publishing". I spent some time exploring the site tonight and it does seem to have all the typical social bookmarking site features, but focused around tags for scientific publications.

SemanticWiki on People at Ontoworld

As journals and other material are shared via Open Access, another key challenge is making it easy for individuals to discover the content they are most interested in. Interestingly, one of the most difficult challenges is understanding which individual wrote certain pieces of research. For example, a common name (John Smith) may be multiple people; or multiple spellings of a name (John Smith, John A. Smith) may be the same person. Proper attribution and understanding of the progress of a research project is very important, when trying to deliver meaningful search results across scientific publications. Many are hoping the 'semantic web' will make that easier. Interestingly, these projects are leveraging the SemanticWiki on People at Ontoworld to help with this.

Conclusion

While I was very familiar with all of the online media trends behind these projects, it was very interesting to hear how this is effecting the publication of scientific research and journals. It honestly was very encouraging! When I think about the productivity improvements that social media has brought to my life, I'm happy to hear that the individuals focused on finding a cure for cancer - and other significant projects - are using the same types of tools to improve their productivity.

After attending a panel on collaboration earlier this morning, I attended the panel "Under 18: Blogs, Wikis and Online Social Networks for Youth". The moderator was: Andrea Forte of the Georgia Institute of Technology. The panelists were: Danah Boyd PhD Candidate, USC Annenberg Center; Anastasia Goodstein Publisher, Ypulse; Kate Raynes-Goldie TakingITGlobal; Erin Reilly Exec Dir, Platform Shoes Forum; and Elisabeth Sylvan Researcher, MIT Media Laboratory. The presentation started with a slide which read "Young People Online are ... a constant mortal danger or fulfilling their inner potential?" This was a pretty good summary of the conversation.

Background Context from Danah Boyd

Each panelist opened with a brief presentation. Danah Boyd's opening provided great context for the rest of the discussion. She talked about how a hundred years ago 14 - 17 year olds participated in society and were mentored by adults. Then during the great depression the government instituted a policy of forcing 14 - 17 year olds to attend high school. She talked about how this began a process of creating a dynamic she called 'Age Segregation'. The concept behind this segregation being that society creates separate activities for teenagers. This however didn't change the fact that 14 - 17 year olds still yearn to participate and express themselves to society at large. The difference is that in the last few years they have begun doing it online. She pointed out 4 things that make this unique:

With this background, the panel focused on three questions: 

(1) What are young people getting out of their online lives? 
(2) What is reality when it comes to dangers for young people online? 
(3) What kinds of social, technological/design solutions are there once we identify experiences we want to facilitate and/or prevent?

What are young people getting out of their online lives?

The panelist all seemed to agree that while the medium has certainly changed and the four attributes of the medium mentioned by Danah certainly bring new implications to this, the things young people are getting out of an online experience aren't that new or different.

Interestingly, Anastasia talked about how individuals are really doing the same thing we did growing up; but doing it online. Also, she pointed out that while young people are doing a number of activities online, research indicates that the thing they value most is the ability to do research online. This led to Andrea talked about how she is helping bring wikis to high schools. In doing this, she has realized how scared teachers are of wikis. This lead to a nice conversation about the dangers of the COPA (Children's Online Protection Act), which threatens to eliminate access to social sites at public schools and libraries.

What is the reality when it comes to dangers for young people online

This was probably the most interesting part of the panel. Danah talked about how she grew up with a good understanding of who were strangers and not to talk to them. However, strangers were completely contextual. For example, the other children she met on the first day of school were not strangers. She feels that most young people in her research have a similar understanding of who is and who is not a stranger online.

The challenge arises when the young person has problems in general. Now, because of the "invisible audiences", a young person who is having a rough time is much more vulnerable. Therefore, she feels we need to leverage this for good - for example, she asked 'what if we had online social workers monitoring MySpace?'

Danah also pointed out that young people can teach us a lot about the act of online collaboration. For example, why do young people break up in the comments of another person's MySpace? According to her research, the reason is that they know that any other electronic conversation (such as IM) can morph. In My Space, a set of 'digital bread crumbs' is left to keep the record straight. She said young people intuitively know this and can add a lot to the conversation - we just need to learn to listen and engage them in dialogue.

Technology and Other Solutions

This led to the most controversial part of the panel: should we try to develop some type of 'Cyber-Ethics'? The idea being we would cover topics like: Online Cheating, Online Plagiarism, and Cyberbullying. Some panelist were of the opinion that we needed to start developing these and educating people. Others felt it wasn't a good idea to break this out as a separate topic - but instead we just should listen and learn and try to teach in general. Also, one of the panelists made the excellent point that this isn't exclusively an issue for schools - parents and other mentors need to get involved too. This received a very warm reception.

Also, while the panel was fairly universal that technology doesn't solve all the problems, they were of the opinion that it is important to educate people about some technology solutions. For example, while there is a lot of concern around plagerism, there are tools now that can easily help teachers identify this.

Conclusion

It was nice to see so many at SXSW engaged in this conversation. I also was impressed at how many people raised their hand at the beginning to indicate they are educators (probably 10% of the audience) The challenges and opportunities are great, but this was a particularly influential group of web leaders with which to converse about these topics.

This afternoon I attended an amazing panel titled "Web App Autopsy". This was one of the panels I was most excited about when I was getting ready to come to Austin.

The panel examined four live web applications (RegOnline, FeedBurner, Wufoo, and Blinksale) at an amazing level of depth. If you are involved in any part of designing, building or marketing a web application, then you really need to go download a copy of the slides.

While the entire presentation was packed with useful information, the data I found most interesting was around the conversion of site visitors to customers - both free and paying customers. The slide below shows a summary of the data the panel presented.

The panel had an interesting conversation explaining the above results. For example, it is interesting that Wufoo and Feedburner have a lower conversion rate to the free services than BlinkSale - even though BlinkSale asks more questions during the registration process.

It is also interesting that RegOnline has a much lower percentage conversion rate to free accounts than the other services, but then has a higher percentage of accounts that end up becoming paid members. It was noted that RegOnline has the highest organic result for the search term 'Online Registration' in Google - so this may have something to do with the poor conversion rate to free accounts. As you'll see in the slides (you have downloaded them right?!), RegOnline also has the highest revenue per customer by an order of magnitude.

When I started my career as a consultant, we did a lot of benchmarking for best and next practices. At my firm, we had a very good repository of data, on many industries, to use for benchmarks. However, we never found a great resource for online web applications. I've been out of consulting for years, but this data would have been invaluable. Is anyone aware of other resources like this to benchmark the performance of your web application? If so, please leave a comment below.

UPDATE: Also see our coverage of the panel The Figures Behind The Top Web Apps. It looked at 5 web applications and what it took to build and release those products.

The SXSW conference has multiple panels going on at the same time. I started my SXSW experience by attending the "World Domination Via Collaboration" panel. The presenters were: Jory Des Jardins Co-Founder, BlogHer LLC; Betsy Aoki Program Mgr, Microsoft; Jessica Hardwick Founder and CEO, SwapThing; Lisa Stone Co-Founder and Pres of Operations and Evangelism, BlogHer LLC; and Jenna Woodul Co-founder, LiveWorld

Overview

Jory Des Jardins opened the panel by stating that in 'web 1.0', community was a "nice to have". In 'web 2.0' the new reality is that community is "the business". The panel then jumped in, focusing back and forth between two topics:

(1) Tactics and policies to build a community organically, and 

(2) How to convince other individuals in your organization to actually spend time and resources developing a community.

The panel discussed two keys to build a community organically. First of all, you need to understand the members and ask them what they want. Second, you need to make sure you protect the community from spammers and trolls.

When discussing how you would understand a community, there were a number of tactics discussed. First of all, the panel pointed out the obvious - that you should greet your members and simply ask them. While obvious, it is amazing how many web services overlook doing this. As an example, the panel pointed to the greetings that early Flickr users received (interesting explanation here).

In addition to these tactics, Jenna strongly recommended that sites consider forming advisory boards. She pointed to the 'eBay voices' as a good example of this advisory board structure. However, she also emphasized the importance of rotating membership of this group. You don't want it to stay stuck with the 'old-timers'.

In terms of protecting a community from spammers and trolls, there was an interesting question posed by a member of the audience: "Is anonymous commenting the cryptonite of an online community?" While everyone agreed that at a minimum, you needed to allow people to create online personas that might appear annonymous to the community (i.e. not their real name), there was some disagreement around whether you should allow completely anonymous commenting. BlogHer does not allow people to comment with out first creating an account. This is so they can block individuals who repeatedly add comments that end up later needing to be moderated; or as Lisa said, they "are building BlogHer not BlogPorn".

Interestingly, Betsy talked about how she allows anonymous commenting, because if people want to take the time to write a comment on her blog (even if it is rude or errant), she wants to react to it. (She does block spam). Betsy also pointed to the Slashdot 'Anonymous Coward' as an interesting example. It shows that the community respects people who comment publicly more, but they are open to taking feedback even from anonymous individuals. She explained that while it has added some work monitoring flame wars in her role at MSFT, she did feel it added value to the Slashdot community. Unfortunately, as this topic was heating up the panel ended - but it was an interesting thing for community builders to consider regarding their own sites.

Convincing an Organization to Build a Community

Previous to starting BlogHer, Lisa was a consultant who helped a number of organizations embrace community. She consistently followed a 2 step process on these projects. First, she would monitor and circulate conversations occurring online, in real time, about the organization. Then she would transition to getting her internal champions (regardless of level) to start blogging in order to champion the concept.

Betsy pointed out that the employee blogging program at Microsoft skipped the first step, but was entirely 'grass roots'. A number of 'feisty people' felt it was important to blog and made it happen. She runs the internal email distribution list and said that they're even more feisty on that list! As senior managers became aware of the blogging, it was discovered that Microsoft had no formal policy around their employees blogging. Note: Microsoft has decided to continue operating without a blogging policy.

Jenna talked about her experience helping large corporations (like BMW and Dove) create community. She explained that at the end of the day her argument comes down to the ROI of customers who are engaged in community versus those who are not. She pointed to an eBay Case Study by Harvard Business Review (link here) that showed participants in an eBay community purchase 56% more and listed 4 times as many items.

Conclusions

This was an excellent first panel of the show. While I wasn't necessarily surprised by what was said, the examples like eBay, Flickr and stories from the front lines at Microsoft were fascinating. I'm sure the community at Read/Write Web has a number of other good examples of community building. Please consider leaving these in the comments below.