The "Feedization" Of The Web (continued)
Back in March 2006 I wrote a post about the "feedization of the web UI" where I observed that I was seeing more and more web UIs that reminded me of the news feeds that are common in the financial markets. I predicted that we’d see more of that in the coming years.
And as the NY Times Magazine points out today in a great piece by Clive Thompson:
He [Zuckerberg] developed something he called News Feed, a built-in service that
would actively broadcast changes in a user’s page to every one of his
or her friends. Students would no longer need to spend their time
zipping around to examine each friend’s page, checking to see if there
was any new information. Instead, they would just log into Facebook,
and News Feed would appear: a single page that — like a social gazette
from the 18th century — delivered a long list of up-to-the-minute
gossip about their friends, around the clock, all in one place. “A
stream of everything that’s going on in their lives,” as Zuckerberg put
The Facebook news feed appeared in September 2006, initially to outrage about "too much information", but as Clive Thompson tells in the Times, users got used to it and now it’s the heart of the Facebook service.
It’s also the heart of the twitter service, the friendfeed service, and a host of other services. And we’ve now got AIR clients like Twhirl that bring all these social feeds to the desktop and iPhone apps like Twinkle and Twitterific that bring them to the phone (please build Twitterific or Twinkle for bberry!).
This aggregation of information into a news feed has been adopted by many of the new web services that we get to see in our office every week. I’d say its the most common web UI/home page we see these days. Services like outside.in which launched with a more newspaper like front page a couple years ago have adopted the feed UI as well.
Here’s a screen shot of my outside.in radar this morning.
Radar is the Facebook News Feed for your neighborhood. What really excites me about this neighborhood feed is that all of the items in it are interesting and relevant to me. The tweet about the gaucamole at Los Dados was sent to my friend Alex Lines by someone I don’t know but will now find out who they are (just like twitter or facebook). The story about the Rag and Bone show is interesting because the Gotham Gal and our girls went to that show. The Spotted Pig and Magnolia are local haunts. And The Standard Hotel which is under construction looms large over our neighborhood.
As I said in my post in the spring of 2006,
Most people who have grown up designing magazines or newspapers
probably look at this user interface and think of it as ugly and boring but I think its super efficient
The web continues to grow and browsing is getting less and less efficient. Search was the first solution to that problem and it’s a huge part of the web experience now. Feeds are the second solution to that problem and I believe they will be an equally big part of the web experience. If you look at my google analytics data, you see that search drives only 29% of all visits to this blog and referring links drive over 40%.
And if you drill down into the referring link traffic, you see that of the top ten referring links to this blog, eight are feed oriented UIs (all but this blog and stumbleupon) and they drive almost 20% of the traffic to this blog.
So feeds are a powerful way for users to navigate the web and get to the information they need. I expect them to get more powerful over time as more users adopt them. Clearly you all are early adopters and the traffic that feeds are driving to this blog is much greater than a more mainstream website would experience.
But think about the Facebook generation. My kids are growing up with the news feed as their start page. Not Yahoo’s portal approach and NOT google’s search box approach. In time, its entirely possible that feeds will be more powerful than search.
So to all the people that say social nets can’t be monetized, just look hard at the feed and think of the possibilities. There’s money in them thar hills.
Totally right on, Fred.
great post Fred and thanks for quoting Twhirl. We would like to add the Facebook feed to Twhirl but unfortunately they do not allow that for now.
That’s because they want to be the only feed in town, amigo. They’ll happily suck in all your data but they won’t export it to other places. I’m sure that will eventually change.
Once again, I appreciate your perspective.I also regularly enjoy adding feeds for particular keywords and/or usersusing Search.Twitter.com.By deciding how to recognize a theme, you can configurefeeds into a [sometimes dedicated] aggregator account, and have an excellent grasp on a subject very quickly, often in real time. (Hats off to Ev for seeing that miles away)
I’m not sure if I buy the “because feeds are becoming a great way to drive traffic, people will eventually figure out how to monetize them” logic. Text/banner ads might eventually become a method of monetizing feeds, but I don’t think they’ll be nearly as effective as search advertising.The beautiful thing about search and what Overture/Google did, was they basically said “Searching” is a proactive display of intent to immediately consume. Whether that consumption be in the form of content or product, users were proactively seeking out information that was important to them at that very moment.I use my feeds as a way to passively keep up to speed on people/topics that interest me all the time, not just when I’m seeking a specific piece of content (or a product). So feeds don’t really capture my immediate intentions.I’m sure there are arguments to the contrary but at the end of the day, when I’m typing “great places for brunch in New York” into a search engine this morning, versus seeing a post about a great brunch spot at midnight on wednesday, I’ll be more likely to click an ad in the search results as opposed to the ads that appear in a news feed.
Feeds will likely replace significant amounts of websearch.While scanning your feeds, you’ll “ignore” a friend’s review of a brunch place. However, when you decide later that you want to go to brunch, that memory will pop out and you’ll search your feed for the reference.
I agree, but will advertisers pay to be an afterthought?That’s my main point.
Yes, I think they will as new standards emerge for what interaction and intention mean. They’ll realize that there is a lot of value in many, many nibbles over time . . . and it will be clear what they have to pay for it.
I’m confused. Which advertisers will pay to be one place that I’m seeking information about products but not another? (Assuming, of course, that the latter has critical mass.)The CPM may be different, but the gross amount of money should pretty much track the amount of “attention”, or rather, intention.
Advertisers will pay to be wherever they can get the highest return on their investment. I’d argue that they’d get higher returns on their advertising dollars by going to mediums that best capture a customer’s intent. It’s all about conversion rates.
Do you really think that an advertiser will pay more for a thousand impressions that produce ten purchases than said advertiser would pay for a million impressions that produce a hundred purchases?That’s 100x lower conversion rates but 10x more sales – which is worth more to an advertiser?If the million impressions cost the same as the thousand, the highest ROI has lower conversion rates….So, which is it – ROI or conversion rates? They’re different….
Andy, I think the problem with our little back & forth is that we’re ignoring the other end of this equation — the publishers.If Google’s conversion rates were 100% less effective, would it matter to the advertiser if they were getting the same ROI? Obviously not.But would it matter to Google? Of course it would.That’s why search is effective from a consumer, advertiser and publisher end. I can’t see feeds doing the same.
I think that you’re assuming something about Google’s costs that isn’t true.I’m pretty sure that google’s costs to serve an ad are dominated by NREs and account servicing, that the cost to process an insertion are in the noise.If that’s true, Google’s costs to 10-100x as many ads would not be significantly higher than they are now.Yes, feed ads might require additional NREs. And, there might be more advertiser accounts. Still, as long as the marginal cost of ad insertion is low, Google can profitablly handle very low conversion rate ads.Note that the costs described above do not include the costs of running a search engine. Why? Because the search engine is “just” a way to attract eyeballs. A feedbased system would have analogous costs, independent of advertising. Maybe, like search, they can be paid by advertising, maybe they can’t. (I suspect that running a feed system is cheaper than running a search engine. It certainly should have lower NREs.)
I wasn’t making assumptions about Google’s costs, I’m talking more about inventory.It’s the inability to effectively monetize that has left Yahoo!’s revenue growth far behind Google’s when it comes to search ads. Google’s more effective at conversions and ad monetization.Cost structure aside, if Google can make $100 serving ad 100 impressions, and we take your example of a site that is 100x less effective at converting, than that would mean that site would only make $1 per 100 impressions. What do you think that would do the company’s overall revenue and profitability (assuming identical traffic across both sites)?This is a great discussion by the way, feel free to email me direct if you’d like to talk about it further: wayne at tickerhound dot com.
[In keeping with the history of this discussion, I’m going to continue to ignore brand advertising.]It’s unclear how we can ignore costs when discussing profitability.I think that the ad serving cost is low enough that we should be talking in terms of the number (and value) of conversions per user, not per ad. The number of ads can be set to maximize revenue.Feeds aren’t sites. Many feeds come from sites, but the ads in the feeds may, or may not, come from those sites.I think that feed ads will have slightly better conversion totals (per user) than ads on the site. The base is the same because feeds are merely a different way to read a site. (Instead of converting from an on-site ad, a feed-user converts from an in-feed ad.)I think that feeds will do better than sited ads because a feed lends itself to “I saw something interesting a while back in my feeds” searches, replacing other information sources. I think that web search advertising will take a small hit but the big hit will be to off-line advertising.If feed ads from a site have lower per-user conversion totals than ads on the site, said site, and Google (at least as far as that site is concerned), will get lower total revenue. However, if that’s the case, where are the missing conversions happening?Note that users may well move to feeds even if the per-user conversions goes down. Some sites will drop feeds, but they’ll still lose users and some revenue to sites that go for volume via feeds.anamax at earthlink.net
Maybe I misunderstood but I guess I thought you were talking about my original post regarding search advertising. In no way was my discussion related to regular site ads vs. feed ads. So I’m a little confused by the site ads vs. feed ads discussion — but if that’s what we’re talking about then I don’t think I could make a compelling argument one way or the other.
I was and am.You claimed that advertisers wouldn’t buy in-feed ads.I asked why not and have now provided one reasons why. I also suggested one reason why feeds might be a more effective advertising vehicle than sites.You suggested that the conversion rates (which are per ad) would be different. I pointed out that CPM doesn’t matter, the only metric for advertisers is ROI, and that the cost of serving an add is low enough that we can adjust the number of ads to maximize the number of conversions.
For direct marketers its all about conversionsFor brand marketers its all about impressions
Fred, in your opinion would a feed-based ad model be more for brand marketers or direct marketers?
I think it would work for both but the execution would have to be different
I’ve noticed that feeds on Twitter and Facebook go back only so far. Unlike Google, a Twitter or Facebook feed doesn’t bring up items from months or years ago. This means I can’t go into the feed to find that item I vaguely recall and now want to locate. (My Twitter archive, as far as I can tell, stops somewhere around the end of July.)I understand (I think) that Twitter and similar services are just one example from the feedization universe, but that limitation really bugs me.On the other hand, would it even be desirable to have every feed be like Google, persistent (& searchable) for years? Would you then end up with so much noise that you need a feed for the feeds? 😉
since moving to Mac and Ubuntu, the thing I miss most about Windows is the Google sidebar headline news feed.I pointed Google reader to all my feeds and got notifications throughout the day from RSS, headline news, email.Have been looking for one of those OSX using growl – no luck.A good universal news feed from all your info sources and messaging platforms, that learns what you want to be notified about is a killer app.the Achilles heel of most of these is relevance, and Google seems to have figured that out pretty well for search.
Excellent insight. it would be nice to see more of an auto-refresh in a portion of the screen, (maybe this is built-in to some of these already(?)) with a user option on the frequency of the refresh, so you could have more of a sense of realtime flow, along with a static “most recent”. The real time version would allow you to glance along as you read through something in more detail. In a sense, the screen would be divided up into time intervals.
Like I said to @davewiner, I dont want a “river of news” feedreader, I want a goddamn “river of events” operating system…– MV
Fred back in 2006 I created a video demo http://hk.youtube.com/watch… with a friend called “liquid lists” to show how information would appear as a river of information would be shared with friends and then via social networks, attention and a recommendation engine, information could be refined and even monetised based on reputation.Sadly the various UK VC’s I demo’d this too did not get the product and most still don’t but ot be fair it was pre-twitter and friendfeed.I would now argue that this “rivers of news” (kudos to Dave Winer) view of the world is the wrong way for us to rapidly digest contextual information. The model I am working on uses XMPP, bi-directional atom and a new reputation algo to help users “discover” relevant information in real-time which is closer to your trader-desk analogy.The service is called iPALS (identity, presence, attention, location, services). If you want to know more ps ping me.
good thinking. good luck. it is what you would know if you already knew it. kicking out the time delay as much as possible.
“In time, its entirely possible that feeds will be more powerful than search.”no doubt. this is why i thought feedburner was going to be the greatest company in the world, one with the potential to disrupt google. of course the borg took care of that, though i still think the disruptive potential of feeds and web syndication technology in general is largely untapped.
It would be beneficial for developers to consider the timing of the feeds, and the division of the screen in cinematic terms; which would allow for a syntactical and liguistic structures to be developed. An old article, with very fundamental insights into this, is Gene Youngbloods, “Cinema and the Code”, http://www.jstor.org/pss/15…
I just love your point about the seach engines being the “first” step to find the information on the web, and feeds a second step. But I think there is a misconception here. Search engines are “products” in a sense that they can be used “as is” by the average “end user”, while, according to me there is still a great lack of tools to exploit the RSS feeds for the “masses”.Of course, Google Reader, Netvibes, Bloglines… are doing a great job, but most of the time they are “oversized” for the average user who only “follows” 3 or 5 sources. Also, all of them are just “another tool” and not allowing web user to use what they already use! Finally, I think the feeds where created to change “pull” into “push”. Meaning that the information would come to the end user without him doing anything! It’s definetely not the case, because if I don’t put netvibes of Google Reader as my home page, I would never know that you just published something great!As a “conclusion” : yeah, feeds are great, but they are mostly used as a final product (which they’re not) when using them as a tool (technology) in better tools could really change the way people surf and access information!
when i say “feeds” I don’t really mean RSS/google reader/blogines/netvibesthey are part of what i am talking about, but just a part of itthe facebook news feed is the biggest thing out there today in terms of aggregating info you’ll likely care about and that’s a very mainstream service.twitter and the smart aggregators like techmeme and hacker news are also a big part of what i am talking about and they are also easier to consume than RSS for the mainstream useri am sorry i wasn’t more clear about that in my post
Feeds are great but there is a certain hit-or-miss / serendipitous quality to them – great for discovering stuff at random, not so much for finding something in particular as we do in search. At least not yet. I’m sure more powerful tools are on their way. Also not sure if its quite ready for prime time the way search is. You and I know to search for #dnc08 or #rnc08 on twitter, does average joe?
Great post. The reason Google make so much money is because so many people use their main search engine to find content, and they show relevant adverts down the side of the page. As more and more people use feeds for their daily web experience (and I’d say that 75-80% of all web content I see on a daily basis outside of direct destinations is from feeds – a mix of my own feed reader, twitter links, and Facebook links), then the ability to place relevant adverts next to that feed will be just as effective for monetisation as Google is with search.
Most (all?) examples of “feedization” are coming from entertainment-oriented websites such as those you quoted above. I wonder if the same concept can’t be applied to productivity software – e.g. to make progress reports easier, communication more seamless, reduce management overhead, etc.
Is anyone doing anything like this already?I would definitely make use of a live feed of my Analytics, AdSense and AdWords data: – site x beat its weekly pageview high – site y cracked $100 today – site z has a CTR of 5% on 3,000 views for $70 – site y’s incoming traffic from Google is down 25% from last weekRight now, I have to log in to three different Google sub-sites and review data from 25-50 domains and, on top of that, try to pick up some of those less obvious items (such as traffic from a particular source having changed dramatically).Anyone else think there is a market there worth exploring?
For sure, feeds are a new way to consume content, event-driven. Precisely, the pain point for feeds is consumption. We don’t have time to read everything along a feed. Which threat should drive attention ?First thing I do is look at the author, and if he is on the radar, I’ll read. But maybe he talks about unworthy things.To know how many people of your group/network consume one thread would be a first step to unclutter feeds.To build on your financial analogy, there could be quotations on threads ;-)eg : (a vc feedization 12.34 + 0.34)
Agreed. The reverse chronological feed is a great UI for browsing the web, but we need better tools for filtering, prioritizing, and personalizing feed items. Without good filters, the “river of news” format suffers diminishing returns as feed volume increases. An ideal system would be one that encourages publishers to share as much information as possible while also providing adequate tools for subscribers to focus on the pieces of content that interest them most as individuals.
JoeI think the filters are there, but they are being built by various webservicesFacebook your friendsTwitter people you find interesting enough to follow (+ your friends)Outside.in your locationLinkedIn your business relationshipsThe list goes on and onWhat we need is ways of mashing these feeds up. Twhirl and Twitterific’sdesktop clients would be an interesting place to try that.
hyper-connection will eliminate the distinctions between centralized and decentralized. mashups are certainly a step towards that higher order of functioning.
I’m picturing something more advanced than the filters provided by thosevarious services (Facebook, Twitter, Outside.in, etc). For example, I mightfollow your feed on AVC or on Twitter because I like your taste in music andtechnology, but I might want to filter out your posts on politics, NYC, andyour personal life. Another subscriber might just want your posts onpolitics and music.There’s an audience for nearly any content that an individual or companyposts to any of these services, but each subscriber is interested in aunique view of the feed. As a feed subscriber, I find myself removing feedsfrom people where the signal to noise ratio is off, when really what I wantto do is opt-out of certain post types, not the feed in it’s entirety. As afeed publisher, I often find myself holding back on publishing certaininformation since I’m worried that only a small percentage of my subscriberswill find that post interesting… and that the others may unsubscribe dueto the noise. The ideal system would encourage me to post information onany topic that I’m willing to share and then route each post to subscribersthrough some personalized “interestingness” filter.
This is so right onStart a company Joe!
feed/stream aggregators like FriendFeed seem like they’re in the best position to do this.As a first step, they don’t have to “push” things into buckets for new readers, as much as prioritize what you see at the top of your feed.
notice that we have to tell the “machine” what we want and how we want it. which means we already know. we are “preaching to the choir” of ourselves, a kind of endless doubting of what we already know.sorry to be philosophical, it is a disease i have.take fred’s advice, and start a business!
When are we going to see an ad format that fits into the river? Avatar = advertiser logo, User = advertiser, headline = link, content = ad. Google doesn’t yet offer this and I don’t know of anybody besides Facebook that is even experimenting with this. If folks are becoming increasingly blind to sidebar content as they consume their personalized rivers(which I think they are), us publishers need to find a way to insert meaningful, relevant ads into this same format.
ads have to be as close to invisible as possible or else you will make enemies, not money. they cannot steal attention or cause interruption. they have to be indistinguishable from content. which means they have to be purely relevant, of great service, valuable, germane.the best way to accomplish that is to change motivation. instead of trying to get something from people, try to give something. and i don’t mean freebies. literally try to improve people’s lives exactly as good content is trying to do.the golden rule is not religious morality, ….. it is a kick-ass business plan.
You are right to be looking at it this way. This is the challenge and the opportunity
I move as many sites’ rivers as possible into my Netvibes page, and love FB’s feed via BBerry. For those seeking a less-than-obvious, solid RSS reader on the Blackberry, I recommend Viigo (http://viigo.com/download), and would welcome any suggestions or views that there are better (local app) ways to view RSS. All of which is to say “I’m there” and am a huge feed fan — but I take issue with two allusions of Fred’s post.Issue 1: Feeds “are a powerful way for users to navigate…”Issue 2: Feeds “will be more powerful than search…”Pew released 80 pages on news consumption habits (online & offline), and they go back 10 years with some of their poll’s Q&A — rare perspective for the interweb — and I’m gonna lean on that data to make two points:#1: “Feeds are Powerful” : Among U.S. consumers, 57% 18-24 “feel overloaded with the amount of news,” and that number is 65% for those 25-49. This is data from the current calendar quarter. Feeds are not “powerful.” Feeds offer near-unlimited choice and flexibility, but overload is the near-supermajority consensus.#2: Among those who “read news daily,” 82% use search engines for news story discovery (odd to me), while only 12% use RSS — so that is a big, big gap to close.The other implicit assumption as Fred mentions his kids’ feed-based focus is that his kids will continue to care about the same topics and/or consume in the same manner over the longer-term. This is inconsistent with Pew’s results, where 18-24 year-olds care way more about “getting news on topics of interest,” but as we age interests shift increasingly to “getting an overview” of the news — likely as life gets busier. That said, those over age 25 didn’t really “grow up” with the social web, so it’s hard to say if anything “changed” over time.I see RSS (like WIFI) has a half-done standard that needs quality-of-service wrapped around it, and I see all of Fred’s top site referral sources (two of which I use) as requiring the equivalent of a pilot license to use unless you’re all up in the industry. News is the #3 activity after email and search, there is solid gold in those hills.
“So feeds are a powerful way for users to navigate the web and get to the information they need. I expect them to get more powerful over time as more users adopt them.”Can’t agree more. What about a toolbar full of feeds?
Not sure I want a toolbar full of feeds. More likey a desktop and phone client
What I am interested in is how do I, as a user, discover and consume content as easily as possible.Search was Discovery 1.0, Feeds are Discovery 2.0
Great post; great thread.Feeds will become the critical element for everyone to discover information from the web, because they generally come from a source that’s relevant or trusted… and they can filter out the 99.9% of the web that’s irrelevant to me. Marketers will become a part of people’s feeds as a means to maintain continued relationships with their customers and to develop new ones. This is how social media will be monetized. Not through “ads” but through the exchange of relevant and useful information.It’s just that right now, most feeds operate at the same efficiency that search did in 1998 when you always had to “click here for the next 20 matches.”That will surely change.
The combination of “feedization” with ever-growing info fire hose (exactly as @Joe Lazarus noted) is precisely our raison d’etre. (AideRSS started as a weekend project by our founder who got tired of being unable to ever get through all his RSS feeds.) Of course, as @Fred Wilson also commented, Google Reader, Bloglines, etc. is only one part of the picture. The list of what *can’t* be fed into a feed grows ever smaller.Good filtering is still a work in progress, for us and any anyone else tackling it, since there’s still a lack of standardization out there. Blogs use different platforms for publishing and comments. Not all social media apps have public APIs… the list goes on. A lot of the evolution is still ad hoc, which can have questionable ROI sometimes, as any software/web service company can tell you.What @stephanelee mentioned is a huge area of potential, too, I think — relationship-based filtering — and something I’ve talked a lot about with our developers. Word of mouth is powerful stuff, so it only makes sense to harness it for managing information, professionally as well as personally, same as we use it for meeting people, finding businesses to deal with, getting recommendations for anything and everything we buy/eat/visit.I’ve been pondering the idea of being able to analyze/filter one’s network on social sites, allowing you to see not only who’s there, but how they engage with the site and others, and how relevant that person’s presence in your network is over time.Solving these issues is still certainly embryonic, but absolutely fascinating given the technological, social, and psychological levels the solutions exist on (I think).
Fred,I’d be interested to know what you think of iGoogle’s new dashboard – it allows users to aggregate RSS with feeds from Facebook, widgets from GoogleTalk, Gmail and others – all as a “homepage”. Here’s a link to see what I mean – http://tinyurl.com/3qdbfd