Feed Stats - What Matters

I know that a lot of people watch the little FeedBurner chicklets that show how many subs any given blog has.

Here is mine:

Here is TechCrunch's:

So what does this mean? Not as much as everyone thinks. FeedBurner has a really good post up on their blog explaining why these numbers have been going up a lot recently.

The main reason is that Google has started to report their feed subs to FeedBurner, both Google Homepage (their MyYahoo/Netvibes like service) and Google Reader. Since Google has about 20% of the reader market, everyone's subs went up on average 20% recently.

The other thing that goes on is some of the aggregators auto subscribe their users to the top blogs. So some blogs get huge boosts in subscribers. My blog has ~5000 subs from Rojo. I don't think they are real but they get included in my total of 19,714.

So what to do? It's really simple. Focus on "reach". FeedBurner calculates a reach number every day for every feed it manages. Here's a screenshot of the popup you get when you click on "what is reach" next to your reach number in the FeedBurner dashboard:

Here are my reach numbers for the past week and my site visitors in parentheses for comparison:

Monday - 2,540 (3,079)

Tuesday - 2,736 (3,738)

Wednesday - 2,987 (3,180)

Thursday - 2,878 (3,042)

Friday - 2,534 (2,733)

So you can see the number of people who read this blog each day via feeds is pretty close to the number who read it on the web. And there is certainly some overlap (that would be a great number for FeedBurner to calculate and I am sure they can do that).

So on any given week, this blog gets between 5,500 and 6,500 readers, about 40% of which are via the feed and 60% are via the web.

So what is the relevance of the 19,714 number? Not much actually. It's the number of people who at one point have subscribed to this blog. If you subtract the daily reach of approximately 3,000 people, you see that over 16,000 subscriptions are out there but not particularly active.

I suspect the same is true of Techcrunch's 279,000 readers as well. If their feed subscribers are like mine, Techcrunch's daily reach would be about 50,000. That's still a huge number of people reading a feed every day.

Comments

[Green Girl]

I find your blog so informative! I look forward to learning more!!

[John Dodds]

Does your calculation allow any discount for people such as me who tend to check posts on RSS, then click through to the website to check comments and/or images that don't appear in the feed and who then occassionally return at the end of the day to see how the conversation in the comments has developed? I couldn't estimate how many people do this, but on the basis of a conversation amongst bloggers last week, it is not an uncommon approach.

[not John Dodds]

What John above said! I too read feeds until I find a post interesting enough that I either want to read comments or make my own comment.

[Andrew]

I love learning about new engagement metrics, very interesting. Like pretty much any engagement metric, it has some inherent drawbacks... I suspect that this reach metric could be gamed by switching to partial feeds. I could double or triple my "reach" numbers simply by moving from a full feed to a partial/excerpted feed.

That being said, as long as the underlying feed structure doesn't change, this metric would be internally consistent (though you couldn't necessarily compare to other feeds of a different structure).

[Nate Westheimer]

An interesting metric would be, "how much traffic are you driving to other sites." It's like "Reach" but follows up on those clicks.

[P-Air]

Do these numbers take dupe subscriptions into account? For example, until recently I was using Bloglines, but then I moved to Google Reader. I haven't unsubscribed fm Bloglines, I just don't use it much any more. Do you know if I'm being counted as two subscribers?

[John Zeratsky]

"I suspect that this reach metric could be gamed by switching to partial feeds. I could double or triple my "reach" numbers simply by moving from a full feed to a partial/excerpted feed."

Andrew, in the case of Reach we factor in Views and Clicks, so even if you switched to partial feeds, the people "forced" to click would've already seen the post anyway. They only get counted once.

[Stuart Watson]

Fred,

It's important to read the definition of "Reach". It's the "total number of people" not subscribers "who view or click on content of the feed".

I blogged about the impact:
http://www.syndicationgumption.com/web/rssense/talkback/no_changes_in_rss_metrics/

So not sure if this is true:
"the number of people who read this blog each day via feeds".

Feedburner "engagement" also seems a little fuzzy and ultimately does not matter (working on that post now).

Thanks

Stuart