Social Network Content and the Lincoln MKS

As I was working on the whitepaper I am writing about the four "value components" of social networking sites (SNS) last night, I came across a situation while perusing Facebook that I had to write about now rather than wait to discuss it in the paper. The topic is regarding the content that SNS play host to that serves as the "real estate" upon which marketers pay to place their advertisements on and around. As you will see from the two pictures below, one a screen shot of an ad for the Lincoln MKS on Facebook and the other a screen shot of a Google search results page for the Lincoln MKS, the "real estate" is quite different.

As you will see, I have edited out the lewd gesture in the screen shot from Facebook, however if you have any imagination at all you should be able to guess what the picture shows. While in most circumstances this certainly is not content that would have a place on our corporate blog, I decided to go ahead and post the picture for learning purposes to prove a point and because if you speak with anyone from my generation they will tell you this kind of content isn't far off from being par for the course on Facebook. In fact, this was one of several pictures along the lines of this kind of content that I came across in the same visit I considered using for this post.

Now just about anyone can see why the real estate on the Google search page might be more desirable than the real estate on the Facebook page for Lincoln as an advertiser here. For one thing, when advertising on Google you aren't trying to sell a four-door sedan with Weir leather trimmed seats and an EcoBoost V6 engine to someone who is trying to figure out why this girl hasn't untagged herself from this photo yet, let alone posted it in the first place. On the Google page however, the real estate is all high quality search results with matching information to the advertisements.

I like to think of content as the mechanism by which web properties channel and drive certain user behaviors. For example, Amazon wouldn't be great at selling books if the content on their site did not contain information about books, such as reviews, pricing, vendor information, reading recommendations, etc. So after seeing the Lincoln MKS ad, I decided to go through Facebook and mark down all the page types with ads that I could find. Below is my preliminary list, please add pages I may have forgotten in the comments.

• Profiles (Wall, Info, Photos, Boxes, and any other application tabs)
• Groups
• Pages
• Applications (Including Facebook's core apps such as links and notes)
• Friend/People lists
• News Feed
• Search results pages

What I would be interested in seeing is the data on which page types have the highest click through rates on advertisements. My guess is the third party apps with specific information types driving a particular user behavior have the highest click throughs. Evidence of this could be the fact that Zynga, the second biggest advertiser on Facebook behind AT&T, spends approximately $2.5 million a month on Facebook advertising. Since users actively play Zynga games or other gaming apps while on Facebook, an ad for a new application would align with the content on those pages more directly than most other ads and pages found on Facebook. Again, I cannot say this for sure as this is simply a SWAG (scientific wild ass guess).

I believe moving forward allowing advertisers to have more control over what types of content their ads are paired with will be a good thing for social networking sites, specifically Facebook, in terms of developing better targeting for reaching users. What are your thoughts? What pages would you think are the best for reaching users through ads? Or do you think I am entirely off on this? Would love to hear your thoughts.

Indexing All for you

Well this week's post will be short and to the point. We have recently finished a major milestone in general content indexing across the United States top news sites by circulation. The current growth focus is targeted at diving deep into specific industry verticals (ie. venture captial, advertising, healthcare...) and guess who gets to do most of the work...YOU. Well not really, but we definitely need the voice of our users telling us what they read everyday. I will rely heavily on our elite beta users and even the new guys to drive what content Pinyadda will index. This is your chance to personally receive exactly what you are looking for on the internet. Perhaps you just have a suggestion for a new tag or even a genre of content, we'll make it happen. Please post a comment or send me an email (greg at pinyadda dot com) and I will respond and get what you want indexed live right away.

New sites this week (all from users):
zerohedge.com, chrisdixon.org, paidcontent.org, moconews.org, contentsutra.org, paidcontent:uk.org, npr.org, designsponge.com, smittenkitchen.com, csmonitor.com, weekly standard.com, and commentarymagazine.com

Until next week,

sligs

ps. pinyadda is currently averaging 30,000 trusted content items a day

New Look and a Look at What's New

We unveiled a couple of new features last week and though I mentioned them briefly in my last post I thought I'd take a minute to dig a little deeper. What follows is description of each feature and a little insight into the though process that went into them.

1) Universal Search. While search is a no-brainer for almost any website, it was a little different for us because we have so much content in our system and so many different content types. Page specific search functions will remain in some places, but we placed the new search bar in the header so that it's always available. You can now search all content types at once or specify a content type using the dropdown. Small feature, big usability improvement.

2) Email Notifications and '@' Replies. Instead of sending emails to users bugging them to come back to the site, we wanted to figure out how to give them something valuable that made them want to come back. You can now mention other Pinyadda users in your comments on any news item using the '@username' convention, and they'll receive an email with a link to the article and your comments. You can also choose to receive email updates when another user replies to a comment you've made. Our internal experimentation with this feature went great (we really liked it) and we've already noticed a big jump in user engagement since bringing it live.

The core idea for Pinyadda came in part from the need for a better way to manage group email lists where lots of great content in the form of both links and commentary were being shared. These features solve a big part of that problem that's something we're really happy about.

3) While it's not exactly a new feature, we've completely overhauled the Initial User Experience, including a brand-new Getting Started guide, a handful of prompts to help new users understand our core functionality, and host of interactive help icons all around the site. Making new users feel at home and letting them know that we're here to listen and help is huge priority for us and we'll continue to work hard to make each user's experience the best it can be. (If you're a user, feel free to email or call me at any time. I mean it!)

4) Again, not really a feature, per se, but we made a host of UI Changes and Language Adjustments to make things more clear around the site. We continue to abide by a "simpler is better" mantra when it comes to UI design and while we're not all the way there the latest changes are a big step in the right direction. We've included more visual cues, simplified navigation, and improved aesthetic consistency. If you noticed the changes, let us know what you think!

In the coming weeks you can expect to see even more improvements, including all new pages for adding People, Sites, and Tags, the ability to email items to friends outside of Pinyadda, and number of integrations with existing social networks. Suggestions and feature requests are always welcome. You can send them here.

If you're already a user, thank you. If you're not yet a user, head over to the site and sign up for an account. We can't wait to hear what you think.

Indexing Speed & Real Time Content

First of all, I should mention that I'm a few days late in my post. Halloween is our favorite holiday here at Pinyadda, and I was entirely too busy playing tricks rather than treating you to Friday morning with Sliggity. So here's my blog post for Oct 30.

As you know, the hype about the interweb nowadays is this thing dubbed "real time." In context, I guess it makes sense - real time, meaning right now or when something really happened. But what if that thing never really happened and someone simply made up a rumor or published bogus content. My biggest qualm with this term is services like Twitter and oneriot utilizing it so heavily. Yes, when someone tweets about something it shows up right away on Twitter and with a slight delay on oneriot. However, I am more concerned with actual 'real' time 'content', meaning that it has value and can be verified. We'll come back to this topic shortly.

Next item; indexing speed. I am running a test (as we speak) in order to measure Google's content indexing speed and the time lapse until I receive a Google Alert. I will clock the time from the moment a content item is posted from this blog until it appears in Google search results and I receive my 'Pinyadda" Google Alert. While I am doing this, you can entertain yourself with some side reading. Here is a post by Charles Heflin back in 2008 about clocking the speed of content indexing (I like how he touches upon page rank influence). To touch upon Alltop, an RSS-powered aggregator that (in their words) allows you to answer the question "what's happening?" Well what's happening when? Yesterday? 5 minutes ago? Right now? Well their answer is once every hour...boring.

Now take Pinyadda's backend, which I like to call a 'real' time 'content' indexing service. I like it because you have to focus on real time, we are indexing new content within minutes of when the item is published, and then focus on real content, which is trusted and valuable to you.

Long enough here are the answers you have been waiting for:

2 minutes for search*
still waiting on google alert. UPDATE 2:52pm, what is the reason for the delay here?
not a controlled test b/c google owns blogger, will try again next week.

New sites this week:
The Blade, Tulsa World, Dayton Daily News, Press-Register, Akron Beacon Journal, Syracuse.com, Kentucky.com, Delaware.com, The News Tribune, Arizona Daily Star, Oakland Tribune, The Morning Call, Philly.com, The Advocate, Sarasota Herald Tribune, Wisconsin State Journal, The State, The Post & Courier, The Journal News. Phew that wraps up the top 100 News Publications by circulation.

What's next? Tune in on Friday morning, Pinheads.

Cheers,
Sligs

A Response, New Features, and the Vision Arriving

Last week, Andrew Davis of Tipping Point Labs wrote a post called "Start Up Overexposed: Too Early, Too Far, Too Fast" that used Pinyadda as its primary example. I encourage everyone to go and read the original post, but in lieu of that commitment, here is Andrew's main point:

"...something I noticed early in my interaction on the new platform highlights one of the major concerns I have with early-phase new media channels: the integration of features that increase reach too fast, too early."

Before responding to this comment I'd like to thank Andrew for his honest critique of our product. One of the best features of new media and the social web movement is the ability to post honest reviews and the ability of small firms like ours to respond. Andrew's keen eye and expertise in the field is something that can only make our company better, and it's feedback like his that is the single most valuable metric of user engagement and satisfaction we have.

Andrew's main critique stemmed from our (now-obselete) set-up guide that asked him to import his contacts before becoming acquainted with the site. If I understand his points correctly, he argues that this sort of 'exposure' has the potential to taint the experience of the initial core user base by extending the product's reach ahead of its development - the internet equivalent of letting too many people taste the secret recipe before it's been perfected.

It's a great point and one that we've wrestled with for a while now as we decide when to push the product, what features are ready for implementation, and what the consequences of each of these actions may be. On one hand, it's very important not to poison the well, so to speak, by inviting people to an experience that's not fully formed. On the other hand, there is simply no substitute for having outside users tell you what works and what doesn't. In our business, particularly from a product standpoint, the work is never done. There will always be cleanups and feature enhancements and new functionality to experiment with, and there is at least as much art as science in the timing of these releases and updates. It is, as Andrew points out, a delicate balance "between inviting constant and sustainable new user adoption without risking too great an exposure to an audience that is not ready to adopt your new concept."

It's something we think about every day, and I'm happy to announce a couple of new features and other product improvements that we think greatly enhance the user experience on Pinyadda and will allow us to open our doors to more people. For the record, we've been in closed Alpha phase for several months now, and access to Pinyadda has been restricted to those we've manually approved on a user-by-user basis. While Andrew's point are well-taken, we have been pretty diligent to avoid over-exposure in this early phase of the product. Andrew was personally invited because we value his expertise in the field and knew that he would provide constructive feedback.

The set-up process is something I've written about before on this blog, and was where Andrew found his main rub with the user experience. I'm happy to say that we've completely overhauled the process, including a repositioning of the invite/import process he talks about. Coupled with a new landing page, the design focuses on providing simpler calls to action, giving users more information about the actions they're performing, and presenting the value of the product to the user, not the other way around.

We've also introduced a universal search function that allows users to find what they're looking for quickly and intuitively, whether they're looking for content, people, sites, or tags. It's the beginning of a series of changes designed to make navigating the system simpler and faster. Also new are several email integration features including the ability to send links and comments via email using a simple "@" operator, and the option to be notified when someone responds to a comment you've made. Both of these features have already begun to influence the way users are behaving by increasing engagement around high-quality content. Discussions are becoming more robust, comments more insightful, and the feedback has been overwhelmingly positive.

This kind of interaction is the real beauty of the Pinyadda product - the ability to engage in real discussion around content that makes people think. Seeing people begin to use the product to facilitate these discussions is an incredible experience, and one that makes the vision stronger and the will to execute that much more intense.

Once again, thank you to Andrew and all of our other testers who continue to help us make this product and this company a positive influence in the world and in their lives. Information is the currency of the knowledge economy, and we're thrilled to be playing a role in that economy that helps people stay informed, express their opinions, and enrich their world view.

If you've got any critiques or comments you'd like to share, or if you'd like to be a test user, please let us know in the comments! We've also enable Disqus commenting on this blog so that you can sign in with one of your existing accounts and track your comments from across the web.

Expanded Thoughts on Twitter vs. Facebook; Value and Trust

The other day as I was searching to find the end of the Internet (no luck yet but I am still convinced it is flat), I came across a post on Danah Boyd's (@zephoria for you Twitter-heads) blog titled Some Thoughts on Twitter vs. Facebook Status Updates. Clearly many other people saw this article as it has over 5,000 clicks tracked by bit.ly and from reading the comments it was apparent that the post resonated with many duel Twitter/Facebook users, as there was a litany of interesting comments about people's personal use cases. Most of the conversation was focused around the types of crowds people interact with on the different networks, with my favorite description coming from Ian Kennedy who quoted Marry Hodder, "While Facebook is like having a dinner conversation with friends, Twitter was like getting up on stage at a nightclub on open mike night." This is a great analogy and I think most people who use both networks would agree with the comparison.

What I find to be interesting about this Facebook vs. Twitter issue has less to do with who people are interacting with on the networks and what information they are sharing, but rather which types of relationship people find to be more valuable and more trustworthy. Now I know this is a bit like comparing apples and oranges since the two networks are used for sharing different information types (for the most part) with different networks of users, but let's forget about these two issues and simply analyze this based on the two different graph designs of the networks.

It is commonly known amongst SNA geeks and many people who study social sciences that the most famous SNA paper published to date is The Strength of Weak Ties written by Mark Granovetter in 1973. Granovetter found that weak ties, basically more distant friends in his study, were positioned to be sources of new information more so than close friends. This idea comes down to the fact that you generally know about the same things as the people you spend a lot of time interacting with, and that new information typically disseminates through your weak ties (technically bridges), people who interact mainly with people outside of your network, who would be sharing different information.

Now considering the two social networking site's (SNS) graph designs from a high level without getting into the different ways different people use the sites, let's agree for arguments sake that one typically uses Twitter to connect with weak ties and one uses Facebook to connect more with strong ties (even though we all have "friends" on Facebook that we aren't really friends with, but I will save the argument that a perfect SNS would have an infinite amount of relationship types for another day). So Twitter = weak ties, Facebook = strong ties. Immediately, the discovery of new and valuable information is more likely on Twitter, making it more valuable right? Well not so fast Ghostrider, some not so recent data (2008) for the real-time world that we live in found that far and away the most trusted source of information was "an email from someone you know", with 77% of people validating this referral type. On the other hand, only 43% of people actually trust the social network profiles of people they know, making me wonder how much they would say they trust the information they receive from people they don't technically know on a social networking sites (SNS) - i.e. a weak tie.

When Granovetter explored the topic in 1973, he considered only symmetric relationships as to not complicate his formal math experiments for his thesis (if you want to get into that go read the paper). Considering the expanded opportunity of developing new relationships on the Internet, it doesn't really make sense to define a weak tie on a SNS the way Granovetter defined them in '73 based on 1) amount of time 2) emotional intensity 3) intimacy (which he defined as mutual confiding) and 4) the reciprocal services which characterize the tie. Anyone who uses Twitter follows people with whom they are not intimate (based on Granovetter's mutually confiding restriction) and by nature and purpose the services aren't reciprocal, but the amount of time and emotional intensity for the follower could still be high, so how weak or strong really are the ties on these networks - sigh, the grey area expands.

This can all be boiled down to this: do you typically value a referral of some sort of information more from a symmetric "strong tie" on Facebook or from an asymmetric "weak tie" on Twitter? (Hold the information type constant in each situation). And secondly, do you trust a referral more from one tie more than the other? And without getting into semantics, yes the two are different (I did just kind of get into semantics huh?). Obviously there is no right answer considering it is somewhat of a subjective measure and a more complete argument would have to take into consideration different information types being shared, but I invite you all to share your own sentiments on the matter.

It is without a doubt in my mind that as we move forward and SNS's evolve we will begin to make sense of some pretty amazing structural, "macro-level" patterns that happen in our society because of the data we will be able to extract from the microscopic relationships within social networks. I have said it before and I will say it again, we are only at the tip of the iceberg on this stuff.

Value Components of a Social Network

Let me start by following Sir Sliggity's lead and introducing myself. My name is Chase Garbarino and I am a one of the Co-founders here at Pinyadda. I will be blogging frequently about a number of topics including news about Pinyadda's progress, our company culture, the social media industry, and a series of posts focusing on social network analysis (SNA).

As many friends of Pinyadda know - I am really into social network analysis. Some of you have been unfortunate enough to run into me at a bar and get locked into a conversation about measuring edge values of relationships on different social networks (social networking sites actually). Let me apologize right up front to those of you who have found yourself in this conversation at the Beer Garden on a weekend night - I often forget my passion for SNA is not the most exciting drinking fodder for others. As I have so astutely realized that most people who are interested in SNA with regards to online properties such as Facebook, Twitter and Pinyadda are not usually the people I am watching the Pats, Sox, B's and Celts with, I am going to channel these discussions through our blog here at Pinyadda.

I have recently been reading a lot of blog posts speculating over the valuation of Twitter, most notably Robert Scoble's post that pegs Twitter's valuation to be between $5-$10 billion. Scoble certainly generated a lot of buzz, with more of the comments and reactions seeming to suggest that most people think he is over shooting the valuation a bit at this time. While this certainly isn't the first time someone has taken a stab at valuing a large SNS and stirred strong debate (see TechCrunch's SNS valuation formula here, and an old post from Om Malik about Facebook here that will make you chuckle), this particular post pushed me to share some of my thoughts on the valuation of SNS's that I am writing a whitepaper about in the coming weeks. I personally believe that the more exploration, discussion and debate we can stir up around this topic, the better. While social media has reached critical mass with 83% of Internet users now using social media, we are merely at the tip of the iceberg on what we know about social media use and effective social media measurements will evolve an incredible amount in the coming years.

In an effort to help us develop a better understanding of SNS's and better ways to measure different forms of value on SNS's I would like to start a conversation about what I call the "Value components" of SNS's. Value components are simply the different components that every SNS has that combine to make up the overall value of of a SNS. After several months of on and off research, I have come up with four value components - they are:

1. User Behavior
2. Content
3. Microscopic network design - the design and types of relationships of a SNS
4. Macroscopic network design - the overall design, structural activity and use of a SNS

The focus of my upcoming paper outlines in more detail these value components and the next set of measurements I believe we should start exploring in order to have a better understanding of where and how value is generated within SNS's. While this will be an ever evolving project, I would greatly encourage and appreciate the contributions of others as I share my thoughts and findings on this blog.

What are your thoughts on the recent valutations of Twitter and the history of valuations of SNS's to date?