Newest Content
Latest NYC Coverage
NY Web Tech Directory
Most Popular Content
Freshly Baked Jobs
Press Releases
New Interviews
Recent comments
- Re: Did Delicious Lose Its Chance To Be FriendFeed?
2 hours 36 min ago - Re: Did Delicious Lose Its Chance To Be FriendFeed?
6 hours 19 sec ago - Re: Did Delicious Lose Its Chance To Be FriendFeed?
6 hours 49 min ago - Re: TinyURL Adds Custom URLs; Is This Exciting or What?
10 hours 6 min ago - Re: TinyURL Adds Custom URLs; Is This Exciting or What?
10 hours 54 min ago
Why RSS is "Broken"
Written by Allen Stern - October 15, 2007
I've been thinking about RSS a lot over the last few weeks/months and with all the hub-bub about RSS stats this past weekend, I thought it is a great time to talk about why I think RSS is "broken" and in it's current form isn't sustainable over the long-term. Before I begin let me state that I like what RSS offers a lot and do read several feeds. Let's look at RSS from two different "broken" angles: subscribers/metrics and marketing. I put broken in quotes because my guess is that RSS was never intended to be what I will discuss but we are here now and it's time to fix these issues so RSS can move forward as the vehicle of choice. Note: this conversation is mainly intended for a business audience, not for a personal blog's feed.
Subscribers/Metrics
The big discussion this past weekend was around subscriber numbers and Google Reader. The discussion has crossed the spectrum from begging to be added to a spreadsheet (which is already out of date) on TechCrunch to Pete at Mashable claiming that "Google Reader Stats are Bullshit". I wrote about RSS subscriber numbers being equivalent to the "hits" metric of the mid-90's. It's so easy to game, and even more than with hits, those that are defaults (or bundles) on any network get additional subscribers that could produce no end-value. A simple example - check out most of the start pages - upon entry, you are have automatically become a subscriber to the defaults they offer.
Let's compare RSS subscribers to E-mail list subscribers. I managed several marketing lists in my former role with millions of users, we had a policy of purging any user who didn't engage with a newsletter in six months (we sent several warnings before the purge). Why did we do this? Two reasons: email cost and annoyance/spam levels. Are we at a point now where RSS lists should be purged? If engagement with the feed doesn't happen, should that user be purged? Perhaps then we can get to real numbers. With the CN feed, I know that every single subscriber has physically selected that they want in. This is because CN is not a default on any network. Rick Klau from FeedBurner/Google noted the following today, "a Yahoo "subscriber" is not necessarily the same as a Bloglines "subscriber" or a Google "subscriber". Is this for real?
It appears that other sites are also looking at removing their RSS subscriber counts, and I applaud them. Until we can get to a metric base that is agreed upon, shouting from the mountain top about how many RSS numbers you have is meaningless.
Since we are starting with a clean slate, what metrics would you be interested in? Some of my very basic ones are:
- gender
- income
- technical specs (some of this is available today)
- past engagement (which stories appealed to x user group)
- time spent in feed
- comment to read ratio
- physical location
- feed referrer (where did they pick up the feed from)
- forward a story - did they do anything with the story after reading it in the reader?
- story rating
- ad clicks/interaction
- network involvement (facebook, myspace, etc.)
Marketing/Monetization
This is the area in which I believe RSS is severely "broken". Let's assume I have 100 people on my feed, what in the bloodly hell do I know about all 100 of them? Nothing. Zip. Nada. Ok sure, I know they are using Google Reader or Snarfer. And I know if they click a story, that's it. Great, how does that help me as a marketer?
If we compare again RSS to email, with email I can force a user to provide the data points I want before they subscribe. With RSS, I have no ability to do that. As RSS grows in popularity across the mainstream, the ability to customize a feed will become more crucial.
From advertising to tracking, the current RSS model just doesn't handle what's needed. Why should I show a blanket ad to everyone, wouldn't it make more sense to customize per user? With some sites showing potentially millions of ad impressions daily, isn't it time for a real RSS monetization engine?
There is such an excellent opportunity here for all of us to capitalize on the revenue which RSS can (and will need to) provide to our blogs and Web sites. We can either seize the opportunity and create an advertising system for RSS or we can leave the money on the table. Imagine how much easier it would be to sell a targeted ad versus a blanket ad like we are forced to today. I don't know about you but I need to pay my NYC rent.
Conclusion
I don't know what version number we are on with RSS, but it's time for a new version. One that addresses the need for marketing data, monetization customization and metrics that we can all agree to. Otherwise, let's just start publishing hits numbers again, I will start with 20 trillion.
Our Partners
those aren't bugs, those are features of RSS!
LOL
RSS may indeed be some revisions. Honestly, I haven't thought much about it. But the metrics you describe wouldn't make it on my list of changes either from a publisher or subscriber's point of view.
From a subscriber's view, one of the things I love about RSS is it is one of the last places where there is still some anonymity on the web. If I wanted to disclose a bunch of demographic data, I would just subscribe to a newsletter, etc. Also, RSS can be consumed by and for things like mashups where the demographics you describe could be pretty meaningless as the page could be owned by a group of people with diverse demographics.
You go on to compare email and RSS especially when it comes to a publisher's need to purge subscribers. There are a lot of good reasons for list hygiene with email for example, sending email has a cost associated with it and the lack of list hygiene can negatively impact your delivery rates, and so on. None of these apply with RSS. If a user "subscribes" to your RSS feed but never visits, there is zero costs to the publisher of the feed. There is no such thing as RSS spam--if a user doesn't want the feed any more, they unsubscribe.
So I won't argue that RSS couldn't be improved, but I don't see that you've made a strong case for what should be changed or why.
You make some good points about anonymity - but would you agree that you have to give up something to take x blog/site content and do with it as you wish? Good point about the mashups as well - my 4am thought would be that this is a special use case and would have it's own feed address of some sort for tracking purposes.
I would rather have no access to information in a way I'd have to surrender my anonymity.
As far as I know more than 99% of internet users don't buy online, I do my best to stay in this 99% and I'm surely not a by-impulse buyer. I avoid doing business with people and firms who advertise by mail or e-mail, as I find it to be very inappropriate.
People have right to privacy, even if marketers say no.
"...but would you agree that you have to give up something to take x blog/site content and do with it as you wish?"
You've got it backward, IMO. The point of RSS is that it gives content producers a way to reach influentials (bloggers) and extended distribution systems (Planet-type portals) that would otherwise ignore those producers. In other words, you don't sell a subscriber your content in exchange for marketing data... you *purchase* a subscriber's attention by providing content.
No, I don't think I should. Or perhaps something, but not my demographic data. Because you benefit from my RSS subscription too. It has to be a pretty damn good blog for me to keep reading it without having a low cost subscription option (i.e. 'click and I can watch it') because I have better things to do than open up a dozen sites to see if they updated today with an article which interests me.
If there is no RSS feed, I'll click through, read a few stories, forget about the site, remember in a month, rinse, repeat. If I have to go through the faff of filling in another form, verifying my email once more, odds are I won't unless I think you're very very good.
If you do want to offer personalised feeds, then there is already the (admittedly poorly supported) option for things such as http authentication to access feeds. But again, this is poor supported despite being in the standard from the start. How long do think it might take for your much more extensive changes to reach a useful level of support?
A second option is even simpler, as they do over at InfoQ (and I am sure many other sites), when you subscribe, you get a personal feed url like feeds.rss?token=some_string_here . That can be pasted into any feedreader, is unique and (more or less) private.
In addition, the feed you offer tells me nothing the site doesn't. While the site is publicly accessible, nothing stops me scripting a simple scrapper to produce my own feed anyway ...
Which was a far longer comment than I first intended, but these things happen.
You can't bring out a new version of RSS. It's stuck at 2.0 by harvard.
You can produce customised RSS and you have to do this through a programming language - but why not produce several RSS feeds for customisation (controlled customisation).
and you i'm british - ergo the s instead of the z.
I think the other commenters have covered my thoughts about gathering metrics and many of their points also apply to monetization when publishers want to use highly targetted adverts.
RSS is designed for syndication. Whether you syndicate your content to Google (via Reader) or to a desktop viewer, those displaying your content get to dictate what adverts are shown and they know their users' demographics. The online readers (quite sensibly) disable any javascript in RSS content, most will remove Flash too and they can remove your adverts just as easily. Sites that show syndicated content as if it were their own can be denied the feed unless you've already reached an agreement.
That said, I do think that RSS is being used in loads of exciting ways that the original inventors never imagined, many of which are just ways of shoe-horning XML data into inappropriate structures.
So the crux of my argument is that RSS seems to be working well for syndication but that we need better standards for data exchange. This is an excellent post that's got me thinking.
This post should be entitled, "What RSS is not". The shortcomings you listed are strictly speaking not relevant in this area - RSS: Really Simple Syndication. Not, Really Simple Profiling And (Also) Syndication (RSPAAS).
And I am glad that all this data is not included. In my opinion, e.g. Google knows already too much about us - combine data from Gmail, Analytics, Feedburner and the Toolbar and you got one comprehensive profile.
it is impossible to explain RSS to my mother!
JM - check this - it's by Stephanie who decoded RSS and explains it the way Oprah would.
http://cravingideas.blogs.com/backinskinnyjeans/2006/09/how_to_explain_.html
uh, no...this is how RSS should be explained...
http://www.youtube.com/watch?v=0klgLsSxGsU
YOU are an IDIOT.
RSS is NOT and never had been a marketing vehicle and the second it becomes one, it will die a thousand deaths and be replaced by another technology.
Get this into your thick, neandertahl-like marketing skull... people *HATE* marketing, and no technology should be subverted by the likes of you.
No need for name calling here, it's a friendly place. I guess the Web was created with the intent to be the marketing machine it is? Yea, that's right, I forgot.
RSS is a marketing tool today - at its most basic explanation RSS is about getting content from a to b and thereby you are marketing it to b.
"broken" and "makes it hard for me to stuff it full of ads and get personal information from viewers" are two very different things. that's like saying that paper-back books are "broken" because they can't scan the drivers licenses of everyone reading them. i'm sure that at some point in time ATOM will evolve to allow hosts to gather more data on their users, but as most people are using google reader, bloglines, etc., it's not a very data-rich avenue to begin with. and keep in mind that the higher your barrier to entry (in this case, how much personal information is required), the lower your readership - if i went to see a movie and was asked for my SSN, i'd pick another theater.
That's why I put "broken" in quotes.
You mean, "evolve", not evolve. :)
Ich verstehe nicht.
The evolution of RSS would more likely be on the user-side and also for the user. I can hardly call it evolution when all of the sudden RSS becomes a tool to profile the readers. :)
I guess it's hard to make money in Web 2.0. But you don't always have to. I see how people need to make money to pay rent, food etc., but there gotta be other ways.