Okay engagement is the metric for feeds, how do we get a number?

Rick Klau pinged me this morning about his post on Burning Questions this morning.  Given that I think Rick is a sharp guy, I love FeedBurner, and I’m a metrics junky, it was an immediate click.  Obviously the genesis of the post was GoogleReader finally including subscriber data back to publishers.  Overnight feed subscriptions lept (on my personal blog it was 107 subscribers overnight).  This vaulted GoogleReader over Bloglines (as many of us suspected was what was going on).  As Rick points out, however, the number of subs you have doesn’t equate to market share:

Total Feed Subscriptions ≠ Market Share
With the feedometer now at 604,533 feeds on behalf of 347,000 bloggers, podcasters and commercial publishers, we’ve got a great sample size to gauge the impact of an aggregator’s participation in the market. On any given day, FeedBurner sees more than 3,000 clients capable of reading all kinds of feeds, including podcasts and video feeds. Each of these feed readers reports data differently, so comparing subscriber market share alone can lead to several challenges:

  • Not all subscribers are alike. Yahoo reports active subscribers over a rolling 30-day period. Most other web-based readers report the total number of individuals who’ve subscribed, regardless of whether they have actually logged in recently.
  • Default feeds are popular. (Yes, this is an early frontrunner in 2007 “Painfully Obvious Bullet” balloting.) Said differently: many aggregators offer a set of default feeds for every new account, or provide “bundles” of feeds by category. These feeds will get disproportionately high subscriber numbers at specific aggregators.

This prelude sets up what we think is a better statistic for measuring market share: Engagement. Audience engagement, which is to say, people reading feeds and people clicking on feeds – is how we’ve increasingly been interpreting feed subscription numbers to better understand market penetration.
Source: Burning Questions • FeedBurner’s View of the Feed Market

I think the “default feeds” point most interesting.  I wonder how many “uber-blogs” have a large portion of their “readers” in this category.  I wonder how many of them really read the posts or, to use the word of the moment, engage.

Regardless, Rick is right, it is engagement that counts.  Now, how do we make (and agree upon) a metric that we can track?  Is it an average of views to clicks?  Do you factor in recentcy of feed access (say your average subscriber checks the feed 3 times a week or something)?

Marketers, web geeks, online publishers, we all like to have a metric.  A number we can track to see how things are evolving over time.  That’s what we’re all waiting for.  Maybe Rick still has that still up his sleeve.

Technorati tags: , RSS metrics, RSS engagement

Comments

  1. Tris … I’ve decided to try the formula in this post of mine to measure my engagement. It shows my “community index” is decreasing as my blog matures.

    http://minethatdata.blogspot.com/2007/02/measuring-community.html

    Thanks,
    Kevin

  2. I’m still trying to make heads or tails out of my feed stats on FeedBurner. Some days I drop 10-15% of my feeds. Then it goes up again. Then down again. I know it’s also counting bots, I just haven’t yet gotten my mind around an accurate way to measure how many ‘people’ are subscribed to my feed.

  3. Traci says:

    Dawud – Feed subscriptions numbers can be feisty due to the way that the thousands of feed-reading clients check your feed for new content. But FeedBurner has catalogued the behaviors of all of these in order to give you the best estimate of subscribers. You can read more about how all of these metrics are calculated in our TechCrunch case study and well as this What’s up with that? Vol. 1 post.

  4. Tris Hussey says:

    @Kevin … thanks for the link! Oooh and XL spreadsheet to play with!

    @Traci, thanks for getting Dawud’s question while I was (eek!) offline yesterday.

    @Dawud, you’l see a natural fluctuation in your subs. Since FB counts feed requests, it dips on the weekends and peaks over the week. Totally normal and it gives you insight into your readers. So if you subs peak on weekends, you know you have more weekend readers than weekday readers!

Speak Your Mind