Our RSS experiment

For the past week, we've been conducting a little experiment with Cognitive Daily. In the past, we've had several readers complain that we don't include the full post in the RSS feed for CogDaily, so last week we published every post in its entirety on RSS (if you don't know what RSS is, I explain it below). Today I'm going to share the results of that experiment.

There are two primary philosophies about RSS feeds. The first one says that the point of RSS is just to alert users to new content. Users don't want to read content using their RSS reader; they'll visit the original web site to do that. They just use RSS to get updated on what's new.

The second philosophy says that site owners shouldn't limit how users access their content. If users prefer to use RSS readers, then site owners should provide full content in RSS feeds so that users can view content using the method they prefer.

For years, CogDaily has been operating under the first philosophy for three major reasons. First, we only make money when readers visit our site, so by just providing a "teaser" in the RSS feed, we encourage readers to visit and see our ads. Second, some of the features of our site aren't available via RSS: polls, comments, videos, and other content don't work on RSS, so there's no point in putting entire posts there anyway. Third, we feel that offering part of each post really does offer a service to readers: we provide enough information in RSS for them to decide whether it's worthwhile to visit the site each day and read the entire post, thus saving them time if an article isn't interesting to them.

But the second philosophy has a lot to offer as well. Some users like to download RSS feeds in advance and then read them offline. If all the content is in the feed, this gives them more freedom. Some users have slow dialup connections, so they can download a bunch of content at once while they're doing something else, then read at their own pace. Some users simply prefer the convenience of having all their reading in the same virtual space. It's even possible that offering these appealing features might attract more visitors to our site than partial feeds do. By attracting more RSS subscribers with full feeds, we might end up with more overall visitors to our site. And offering full feeds could attract "power users" who link to our site from their blogs, thus generating even more interest in our site.

Since both philosophies offer potential benefits, we decided to experiment with full feeds for a week and see how our traffic compared to the previous weeks. Here are the results:

i-02492c6ce97bd28e993f75fdfe060bac-rss1.gif

It looks like traffic's gone down and RSS subscriptions have gone up. So perhaps this new plan is just good for subscribers but bad for us, since fewer people are viewing our pages and looking at our ads. But it's also possible that our content simply wasn't as interesting during the week of the study. Also, we've noticed in the past that subscriptions tend to just go up and up. People apparently are more likely to subscribe to a new RSS feed than unsubscribe to an old one. So perhaps rather than look at whether RSS subscribers are increase, we should look at the rate of increase. And let's take a look at some other measures as well -- for example, how many people are commenting and linking to our site.

Unfortunately, all these measures have very different scales -- comments and links might be in the tens or hundreds, while page views can be in the hundreds of thousands. So the following graph converts several different stats into z-scores, allowing us to compare very different measures:

i-3876f2d729bd9205aedbc6a604917115-rss2.gif

As you can see, we had a very good week April 23-29. Links to CogDaily seem to correspond quite closely to page views, as do comments. But RSS subscriber growth was largest during our full RSS week. So perhaps simply including the full RSS feed was responsible for the larger-than-average increase in RSS subscribers. But traffic on CogDaily can vary so much from week to week that it's really difficult to say.

It's also possible that since our change in RSS policy was unannounced, it didn't actually affect readership much one way or another. So let's take another look at the numbers in a week, now that we've publicly announced the full RSS feeds.

[I'm including an explanation of RSS feeds below the fold]

What is RSS?
RSS, which stands for "Really Simple Syndication," is a way for readers to view the content of a blog or web site without visiting the actual site. Instead, they can use a reader such as Bloglines or Google Reader to view selected content from that site.

The site owner determines what to put in the RSS feed -- it could be anything from headlines to complete articles.

The most powerful feature of RSS is this: When you "subscribe" to several different RSS feeds through an RSS reader, you can see all those feeds in one place, without visiting each site individually. In this way you can be updated on literally hundreds of posts in a matter of minutes.

Some RSS readers allow you to download RSS feeds in advance and then read them later, when you may not have access to the Internet. Some readers allow you to mark articles you've read, or keep notes on articles.

To subscribe to CogDaily or any of the blogs from ScienceBlogs.com, click on the "RSS" tab above (or just click here).

Your opinion?
After a week of full RSS feeds, are you finding Cognitive Daily more convenient / easier to use? Or did you prefer things the old way? Let us know in the comments.

More like this

Last week we reported on our site statistics after going to a full RSS feed. The results were disappointing; our numbers went down. We said we'd continue the experiment for another week to see if the trend was reversed once more people heard about the option of viewing all CogDaily content in RSS…
One of the most hated practices on the Internet is the breaking of articles into pages. Jason Kottke swearingly rants against it here, and Mike Davidson denounces the practice here. I don't much like the practice either, especially when a short, pointless article is broken into four or more pages (…
Last week, the number of subscribers to this blog's RSS feed passed the 2,000 mark, after teetering just below that number for a couple of months. It's very gratifying to know that so many people enjoy my writing, and although this increase in subscriptions is tiny compared to the total number of…
ScienceBlogs is, without question, the largest online conversation about science. We have 71 blogs, almost 70,000 posts and 850,000 comments. How does one reader keep up?! One of the easiest ways is to subscribe to the ScienceBlogs Weekly Recap, a fun email newsletter that summarizes the previous…

I usually do read stuff in my feed reader, but I don't mind clicking to read more. It's only annoying on the blogs that update 5+ times a day and you have to click through for all of them. I think you guys were fine before (although I did notice when you changed, I thought it was a bit odd when I didn't have to click through any more).

I actually find that I am far more inclined to click through and comment when I can read a full article in my feed reader.

I do much of my RSS tracking from work and I don't like to visits sites which are clearly not work related. RSS strips CSS and presents articles in work friendly format for me.

Avoiding the ads is a side benefit. However, given the choice I'd opt for full RSS feeds with embedded ads over partial feeds. The Register does this, as does the Rampant Coyote blog.

Tasteful ads, please, or I won't be visiting the site or reading the feed on a regular basis!

It's a bit of a complicated issue. Often, if I don't see a full story in my RSS reader, I just move along to the next one unless the "teaser" is particularly engaging. On the other hand, if the full story is there I'm a bit more likely to keep reading (as I did for this post). If the end question is "what is more likely to get the reader to click through to the page", then I think that has to be looked at on a per-post level. Certain posts are interesting enough so that just a teaser will pull people in, while others need to have the full text out there.

Just my $0.02

One technique I have seen used successfully is to use a fragment of an RSS feed as an ad. For example, in my RSS reader, I could get "cognitive daily [advertisement]" along with the regular "cognitive daily" feed. Several sites I subscribe to use this method and I find it non-obtrusive and helpful for the site which relies upon page views for ad support.

Alternatively, you could provide summarizing material in the RSS feed to stimulate a page view.

Keep up the good work.

I like the full RSS feed. I check Google Reader at work, so I don't want to have to visit each site I'm subscribed to every day. But if the subject interests me, I'll click through regardless. I suppose it's all up to the teaser. Perhaps the main point of the study, a few graphs, then more in depth under the fold?

Keep it up, CD.

For CogDaily, which doesn't have too many articles appearing per day, I think the smartest thing would be full length rss articles, with ads.

By Raman VikramAdith (not verified) on 14 May 2007 #permalink

I think feeds should always be complete. The reader-software should offer the option to show only the summary and the subscriber should have the option decide whether/when to see the complete post, wherever they choose.

As for page views, news/blog/information web sites are passé. Unless you offer an interactive experience at your site, don't expect visitors. If you need ad revenue, include ads in the feeds.

Please publish the full content in the feed. I read 140 feeds and it would be insane if I actually had to visit all these sites every day. Ads make negligible money anyway. You're doing it again, with this "RSS explanation below the fold" stuff. No hidden content! It takes me a millisecond to skim over something if I'm not interested, but takes several seconds to click, wait for page to load, switch tabs, then close the tab, etc. Have mercy and publish your full content in the feed!

I will admit I'm not at all likely to visit the actual site if you put the whole thing in the RSS feed. There's no longer a reason for me to do so, and I'm usually pressed for time so don't bother reading any comments if the whole story shows up in my Bloglines.

You're doing it again, with this "RSS explanation below the fold" stuff

Why would someone reading an RSS feed need to know what an RSS feed is?

I prefer having to click through a million times better than having an ad in the feed. Particularly if you have graphs and charts, I find it much better the way it was, with a couple of paragraphs to determine if it's interesting, then a click for the rest. Don't listen to the arrogant "I'm much too busy to be bothered" crowd.

If you redirect your RSS feed through Feedburner you can track not just how many people are subscribing to your feed but how many of them actually click on your feed to visit the site and that's the figure you really need to know.

I may be in the minority here, but I like getting only the summary in my reader. I can tell right away if it's something I want to read more about, and if so I will click and visit the page. Otherwise, on to the next item. If everything had the full content, it would take longer to get through all the feeds.

First let me vote for a full RSS feed as well. Usually feeds that have truncated entries get kicked out of my subscription list very quickly, just because it is annyoing to always click and leave the reader. Often enough feeds that use a smart "barrier" in their entries have the chance to provide enaough of an abstract and teaser for me to click through.

For my own blogs I chose the full RSS feed, as this is, IMHO, best for the readers. But then I don't have the traffic that would make me whine about lost ad revenue -- which I do have a bit on that other blog. And there most ad-related traffic comes from googlers hitting the archives.

It might just be my personal pereference: feed and main page are ad-free and not truncated.

Also: great blog! keep up the quality!

I like shorter RSS feeds myself -- I try to spend as little time as possible browsing through irrelevant things, and a paragraph or two is generally sufficient to let me decide if something is of interest and worth clicking through for. I don't know the tech behind RSS very well, but surely it is easier for the human publisher than the reading software to decide how much content is needed for the reader to know if the rest of the article is likely to be interesting?

I only started reading your blog after the complete RSS feed was implemented. And here I am having clicked onto your site. :)

Popularity and scale is good for bloggers. Continue the full feed and I think you that your internet presence will increase.

Maybe you should be looking at the total number of readers over time? (site visit and rss).

I have no problems with small and tasteful ads in RSS feeds. Anything with flash, noise or movement would have me unsubscribing very quickly.

"Second, some of the features of our site aren't available via RSS: polls, comments, videos, and other content don't work on RSS, so there's no point in putting entire posts there anyway."

This seems like faulty reasoning. How is posting the whole article on RSS minus the extra content pointless? If anything, this is part of the solution to get users to click through. I know it works for me, quite well. I'm here posting a comment for the first time, aren't I?

I find sites with one line teasers in their feeds I simply scan through, to be honest. Thus I don't get much info from them and I definitely click through less.

Thanks to everyone for your marvelous input. We're taking it all in, and we're continuing the full RSS experiment for another week. Then we'll take these issues (and our hard data) back to the management at ScienceBlogs.com and see if we can find an RSS solution that works well for everyone.

You may wish to check out systems that allow you to add advertisements to your RSS feeds. Google does, as do many others that pay better. Some sites, change all links to point to their site, and then if someone wishes to follow up on a story, they must first click to view your site (you get ad revenue), then click to actually get to what they wanted. (I dislike it, but it does help the content provider make money).

Lastly, some feeds eliminate pictures from the RSS feeds, but give you the whole story. To see images, you must click to view on the site.

Please keep your RSS feeds full. If you want to make money there are ways to put ads, even google ads at the bottom. I really like to stay in one place. It's got to be really interesting for me to visit the site. Also it is WAY easier to e-mail from my reader so I do that more often if it's in my reader.

So that's why you've been filling my Scienceblogs combined feed with overlong blocks of text. Please stop. If you want to offer a full length feed, please do so separately so those of us aggregating a lot of stuff don't have you shout in our faces.

I'm one of the RSS readers but my reader is configured to show the entire webpage. Thus, I see the ads.

I have gone back and forth on the partial/full thing with my own blog and I finally decided on full posts.

While it's not a very big data set, I would like to point out that your comments went up even as your links and page-views went down. That may (and boy is that a big "may") indicate that the RSS readers are just as interested in your content as the regular visitors. Which would bode well for advertising.

You should consider carving out visitors who arrived by links from those who arrived by bookmarks or other methods. With the z-score chart it's impossible to tell if the links are down because fewer people are reading your blog, or if fewer people are reading your blog because there are fewer links to bring them here. Separating the populations may help answer that important question. If an RSS feed decreases the number of links you get then it doesn't much matter that I'll wail and moan if you stop including the whole article in the RSS feed.

One week isn't alot of time to test over, but the results are very interesting!

When it comes to rss feeds I think there's really two significant reasons to click through... either to read the article... or to read/post comments.

While I'm probably less likely to click through for the article alone if I already have the full text... the more text I get to read, the more hooks I get into thinking the comments might be good... or that I might have something to contribute to.

Which (in my case, at least) amounts to me being more likely to click through on full-text articles than on partial feeds. But thats just me.

I like an earlier comment that some basic in-feed advertising could be used to support the full text feed. That way you get a higher subscribed volume of people looking at the ads you serve.

Well, I like to have full content in my feeds and have subscribed to your feed and taken the time to comment here to give weight to that side of the argument. Regardless, love your content. Keep up the great and thoughtful work!

I don't use RSS, so I have a slightly different kind of comment.

When I visit the page, I liked having the full article below the fold since it made it easier for me to scroll down the page and find new posts. If there was one I wanted to read later, I'd usually click on the link to open a new tab in the background that I could come back to later. But with the full article on the main page I find it more annoying to do this, since now I have to scroll a lot further to find anything.

My old policy used to be read articles with "Read more" later and ones without now. For at least one reader, you've made a serious change to the way he can read your site.

I would propose making the following change: put the full article in RSS, but on the main page have only the teaser. Or, better yet, offer both a teaser RSS and full RSS feed. Then people can choose which they want to use.

I support the shorter feed, especially when one is reading many feeds at one place. I believe that's how we're habituated with, like scanning headlines of several newspapers in the morning, cruising through search results in SERPs and suchlike.
As for the topic of this essay, I guess one week is too short a span to come to firm conclusion. Now that you've made an announcement, it has to be seen how things develop over at least a month.
It'll be interesting to know the final outcome of your survey.

Hi Dave,
I just thought of something... the people who still visit the site are maybe more likely to be the ones who would click on your ads and outgoing links anyway.

The people who read in RSS are fairly likely to never have clicked on any of your ads.

Best,
S.

Senia--

Can you tell that to the folks at Seed (who pay us based on page views)?

Seriously, though, that would be an interesting experiment. Anyone have any ideas on how conduct it? After all, the people who don't visit the site don't have an opportunity to ignore our ads.

I'm a HUGE fan of the full post feed. I do much of my reading on airplanes and, while I copy posts that look potentially interesting to a "follow-up online" folder, I skip many others because I can see the full feed. You can monitize your RSS feed through FeedBurner pretty effectively - www.feedburner.com. This doesn't solve the comments problem, although presumably if someone is passionate enough to want to comment on something, they'll find a way to do so when they are back on-line.

Thanks for the great blog - I really enjoy it.

seth levine