Measuring Engagement of the Social Web: ‘07-’09

engagement-trendIt is no surprise to any publisher that many of the conversations about their content and within their audience are occurring off their properties. Increasingly, social hubs such as Facebook, Twitter, and Digg are found in the top 3 traffic drivers, and their influence is only likely to increase over time. At PostRank, we’ve been monitoring and recording all the off-site engagement since mid-2007 when we first launched our service. Since then, we’ve observed a number of interesting trends, which are worth paying attention to, especially if you are a publisher yourself.

In absolute terms, when we sample over the top 1000 of the most engaging feeds for the past three years, we see approximately a 30% year over year growth in engagement –- a sign that more people are participating in the social web. The “share, and like this” phenomenon, which is sweeping through Facebook, Twitter, and dozens of other social hubs, are all facilitating this trend.

On vs. Off-site engagement

Not surprisingly, as the mind-share of these social hubs grows, the conversations and discussions about the content are increasingly happening off the publisher’s property. Between 2007 and 2009, the amount of on-site engagement has dropped by over 50%, while the off-site engagement has skyrocketed!

Larger audiences & highly engaged communities

create-vs-chat

Looking across the different types of activity (PostRank’s 5Cs of Engagement), it is clear that trackbacks, which were previously the primary means of engaging in conversations, are becoming dwarfed by the lighter modes of engagement. From 2007 to 2009, the contribution of trackbacks to total engagement has dropped from 19% to 3%! In the meantime, channels such as Twitter, Friendfeed, and Facebook have gone up from less than 1% to over 29%. For publishers, this is great news: larger audiences, more conversations, and more engaged communities. That is, as long as you have the tools to track this long tail of distributed conversations across the social web (shameless plug: PostRank Analytics).

Content Discovery & Distribution within the Social Web

An interesting side effect of the widespread adoption of more pervasive communication tools (Facebook, Twitter, etc.) is the increased lifespan of a typical story. Back in 2007, we observed that over 94% of all the engagement occurred within the first day of publishing the article. Even more interestingly, the 98% of the engagement on that first day occurred within the first hour. In other words, the half-life of a story was, and still is, less than an hour!

1st-hour-eng

Fast forward to 2008 and 2009, and we’re seeing a steady increase in the lifespan of a story: down to 83% of total engagement for the first hour in 2008, and 64% in 2009. While the real-time web is all about lowering the latency, the pervasive nature and number of people engaged in their communities and conversations (the Social Web) is helping with information discovery. What we’re observing are the effects of strengthening the weak ties.

Engaging with the Social Web

The social web is growing fast. With 30% year over year growth we can expect both the number of people and the amount of their participation to steadily increase, which will translate into more feedback loops for the publishers. At the same time, the “share this” phenomenon shows no signs of stopping, and the deployment of “one click participation” mechanisms such as Twitter’s new retweet functionality, or Facebook’s “like” button, means that each one of us is becoming (if not already) a content curator. In other words, the distribution of content is becoming truly distributed.

Anyone who worries about the monoculture effect of niche networks should breathe a sigh of relief — the data indicates the complete opposite. Even as we organize ourselves into micro-networks of interests, the weak ties that everyone brings are helping the content to travel further and over longer periods of time. Increased lifespan of a story from 2007 to 2009 is case in point.

The challenge: while the consumers are trying to wade their way through the abundance of content, very few publishers understand how to properly engage with the social web. After all, if over 50% of the engagement happens in the first hour, which is also definitely not driven by Google, then how do you optimize and better engage with your audience? Fire and forget is not a strategy, and as random as the behaviour of these social networks may seem, there is a method to the madness.

How do you do it?

  • user engagement has grown a lot over the last year, since we started Bloghology Network a network which is similar to Facebook but for Bloggers and Blogging needs. we have also seen that.

    Both Bloggers and Marketers are engaging to conversations via commenting, status updates and other means .

    it will be very interesting to see how things will evolve from now and on.

    Congrats for the stats/report.
  • drgunn
    The way I understand that is that the mindshare an individual article gets at the moment it's published is reduced, but filtering mechanisms such as likes and reshares bring it back up over days and weeks to those who have good social networks.
  • Hmm, I don't know if we can draw that conclusion exactly. In fact, the overall engagement is going up, as seen by the 30% year over year growth. The trick is, there are lot of conflated variables: more people starting to use social tools and leaving comments, more users overall, etc.

    What you have to keep in mind is that those are % values, not absolute. So it is the case that the absolute numbers are growing as well, it just so happens that the distribution within that absolute number is changing to be predominantly off-site.
  • drgunn
    Yes, Ilya, I was speculating a bit. To expand on this, would it be appropriate to say that people are subscribed to more sources now than they were in 2007, so that particular person's attention is more divided than it was? If so, then it follows that they'll miss more stuff in the hour or day it comes out, right? If they have a good network that reshares stuff they're interested in, they'll also be more likely to see it later.

    Like I said, this is all speculating, but sounds plausible to me.
  • That's an interesting hypothesis. I could do some analysis on the PostRank site to see if that stands (for RSS subscriptions specifically). However, even with that, that obviously wouldn't capture the effects of Twitter, Facebook and other platforms.
  • drgunn
    Yes, and I guess there wouldn't be much point in leaving them out, considering the trends in attention paid to those sources.
  • Great stats Ilya, thank you. It struck me as peculiar the increased life span of a story as this is counter intutive to the buzz around the Real Time web. I would have expected the first hour to be more important, but this may still be the case, as there is a huge net increase in activity across the board. Maybe the first hour was always mission critical, now it has a longer effect, any thoughts?
  • Louis, that was definitely an interesting take away for me as well. First, I think the 'real-time web' phenomenon is still very new, so I don't know if we've actually seen the full effects -- it will be interesting to revisit these same stats in a year or so. Second, and this would be hard to measure, but it would be interesting to analyze the 'first hour' within each niche network. My theory is, before the content would spread virally within a network and then quickly tail off. Now, there are a lot more networks ("we have more connected components", in graph theory speak), which means that information is spread to more distinct subgraphs of people. I would expect that within each network, it is the case that first hour completely dominates the distribution of engagement -- coming back to everyone being a curator.

    In fact, retweet is a great example of this: each time someone retweets a link, they're basically seeding it within their own micro network (same thing for "like" in facebook, friendfeed, etc.).
blog comments powered by Disqus