We have been aggregating comments from around 2000 blogs for several months now and I thought I would share some statistics.

Daily Total Comments

We currently aggregate on average 13 thousand comments a day but that drops to around 7-8000 on a Saturday and Sunday. Everyone obviously has better things to do on a weekend than comment on blogs. I dug into the stats and plotted (in green) the number of posts against the number of comments.

It seems the ratio of comments to posts is pretty steady (the graph does show a fluctuation but its pretty small) and that the readership still exists at the weekend but the content level in the blogosphere goes down and therefore so does the commenting.

Distribution

I twittered a week back that Mike Arrington was one of the top commentors on his own blog (this is what blogging is about!) – but I wanted to understand how many people returned to the same blog and left comments. I have plotted a distribution graph for a few different blogs.

TechCrunch

The first is TechCrunch – It is very clear that over 14 days the vast majority of people only comment once (2015 of them), then 466 left 2 comments, 171 left 3 comments. Now we must take these results with slight caution because the feed from TechCrunch is not authenticated – which means that anyone can leave a different name each time they comment.

Mashable

The second is Mashable – With 1506 commentors only leaving 1 comment we still see that commenting is still seen as a one off throw away event and not a conversation.

Mashable does also have their own social networking style setup plus threaded comments which certainly have an impact on people returning to follow up comments left.

Chris Brogan

And lastly we have Chris Brogan’s blog. Smaller than Mashable + TechCrunch but it is clear that he has a following who do come back and comment more than most, Chris also participates a LOT, he is in fact the top commentor on his own blog (lesson to learn here for new bloggers)

Even with the technical issue of non-authenticated names it is still clear overall that commenting in general is seen as a reaction to the the post content rather than an ongoing conversation. I think this is still largely due to the barriers involved in leaving comments and the lack of generic comment tracking.

Brand Tracking

When I mentioned some of these statistics on twitter a few people asked if we could track brand names in the comments that people were leaving. The answer? sort of.. it is not very easy right now but I fashioned a couple of SQL requests just to see the count per day. Below are the graphs for Dell & Apple (two big brands which on some of our blogs do get talked about a lot)

What we don’t know however is the context in which these comments were made, were they complaints, recommendations or just general discussion. If we could understand this then the value of the tracking would be much higher. Watch this space.

Going Forward

Given peoples fascination with these sorts of statistics I will be putting some effort into automating the process of building some of these charts. If anyone has any ideas on what you would like to look at drop me a line. Use the contact form or you can stalk me on twitter.