44 Comments

Summary:

Twitter’s purchase of Summify, which delivered an email summary of interesting links from a user’s social networks, shows Twitter is trying to get smarter about how it filters the flood of information users are exposed to. It’s a challenge that’s only going to grow.

Twitter made an interesting acquisition on Thursday, when it bought a young Canadian startup called Summify, a company whose service (as its name implies) was designed to cut through the noise of all those social-media streams and summarize the content that matters. More than anything, this is perhaps the single biggest hole that exists not just in Twitter but Facebook and other services as well: the need to give users more ways of filtering the massive amounts of information that keep flooding their activity streams and other social-media inboxes. There are so many ways of producing and sharing content but so few good ways of filtering.

As has been reported elsewhere, Summify says it’s mothballing its service (a decision that was not received warmly by many users), and the team of five will join the growing ranks at Twitter’s new headquarters. The two co-founders, who are originally from Romania, moved to Vancouver, B.C. when they were accepted into an incubator program called Bootup Labs and later received angel funding (according to one report, a Summify investor posted a message that suggested the Twitter acquisition was an all-stock transaction, but the tweet was later deleted). Like some other services such as News.me, Summify filtered a user’s activity streams, then used an algorithm to produce a daily email with links to the most-shared content in their social networks.

While the service is being closed down, it seems fairly obvious that Twitter bought the company to try to incorporate that kind of semantic filtering into its offerings — likely by beefing up its new “Discover” tab, which uses trending topics and other features to try and suggest interesting content to users. As it exists now, the Discover option isn’t very comprehensive or well-organized, and it makes sense for Twitter to try to do that better. One of the service’s challenges has always been to figure out how to keep new users engaged, since it isn’t clear to many what exactly Twitter is for.

As we’ve described before, the need for that kind of curation — whether human-driven or algorithmic, or a combination of both — is hardly unique to Twitter. Facebook also suffers from an information overload problem, one likely to be exacerbated by the launch of hundreds of new “social sharing” apps that will fill users’ feeds with oceans of “news” about things their friends and acquaintances bought or read or ate. Even Google hasn’t been able to solve the information-overload problem, despite its launch of Circles for its Google+ network: a feature many find cumbersome.

In some ways, services like Summify seem like a throwback to the early days of the web, when people wrote blogs that consisted solely of half a dozen links to interesting posts or news articles or websites. In fact, one of the earliest examples of a blog — Jorn Barger’s site Robot Wisdom, which he called a weblog and later shortened to “blog” — was exactly that, and the tradition continues today with link-blogs such as Kottke.org or Brainpickings. The principle is the same as it was when Barger started doing it: to filter the massive amounts of information on the web in some usable way, to make sense of the flood.

In a blog post on the Summify acquisition, Mike Davidson — founder and CEO of an early social-news community called Newsvine, which was later acquired by MSNBC– said his ideal news site would look very much like a Summify email, but on the web: half a dozen links to the most interesting or relevant content available, changing daily. And Summify and News.me are hardly the only ones trying to fill this need. Others, including individual efforts such as Dave Pell’s excellent NextDraft newsletter, are aimed at solving the same problem.

The same impulse is behind apps like Flipboard, which recently started to incorporate more curation elements, both human and algorithmic — and it’s the driving force behind new startups such as Pinterest, as Om pointed out in his coverage of the red-hot newcomer. Smart aggregation and curation that makes sense of the flood of social data we have all around us is the Holy Grail, and it’s also one of the few remaining justifications for traditional media, since journalists have been doing that kind of curation and filtering for decades. But who will ultimately win this race — and how — is still an open question.

Post and thumbnail photos courtesy of Flickr users Luc Legay and See-ming Lee

  1. Simon Le Pine Friday, January 20, 2012

    I recently wrote a blog post outlining how I believe the next big “thing” will be filtering. We have big data (and content overload), now we need to filter it in order to use it. What I suggested in the blog post is a filter API, that way a user can pick specific words/phrases and filter them from social media, news/blogs, email, etc.

    For example, I have no interest in the latest gang shooting locally. I would like to filter all my news and social media to remove any posts related to shootings.

    Unfortunately for something like this to work the social networks and/or social media apps would have to sign up and you’d have to develop Outlook plugins.

    Share
    1. Hey Simon. totally agreeing with you, the next big thing will be filtering. We are a silicon valley online travel startup called Guestmob, and one of our core innovation is curation. For example, there are 140 hotels in central NYC, We filter down to about 6-7 of only 4 stars with similar quality/amenities/location/price. Did I mention our price is unbeatable? That’s besides the point. However, just wanted to share that with you. Glad to see someone who believes in curation.

      Share
      1. The NEXT BIG SCARY thing is filtering! I can possibly imagine anything worse then filtering… imagine all information controlled by what a computer thinks is important to you till it narrows the information you receive down to a single demension and your a person from the movie “idiocracy”

        Share
  2. Bettina Padurano Friday, January 20, 2012

    Thought it useless.

    Share
  3. Matt its not just content curation is the future

    1) content curation with API-oriented approach for shared Web content flow model
    2) mining and integration of social graph with interest graph with better understanding of relationships and user behavior
    3) more sophisticated aggregation and federation of data sources using semantic linked data style structuring

    Share
    1. Richard Creamer Saturday, January 21, 2012

      Steve, great ideas/suggestions. It’s encouraging to encounter people with a deeper understanding of these issues.

      Share
      1. Thx Richard and so you know I consult across my 1-3 points. Also weary of peeps who think it’s just curation or filtering as the future ;)

        Share
    2. It’s a model thats set for failure! Think about how wrong this can go so fast… http://www.ted.com/talks/eli_pariser_beware_online_filter_bubbles.html

      Share
      1. Muchas gracis por el link, buenas ideas : Thank you very much for the link, great stuff

        Share
  4. Joshua Ledgard Friday, January 20, 2012

    I loved this awesome service.

    Great for the founders. But what sucks about these “talent grabs” is that it leaves users hung out to dry.

    We’re working on a great alternative called SiftSocial that won’t be shut down any time soon. Sign up for our beta today! http://www.siftsocial.com/?s=1SKC.

    Our focus is to help you engage more efficiently in conversations and content online.

    Share
  5. I am truly inspired by this article. I do agree that curation is the future. For example, in the travel business, no one wants to sit for hours and hours to go through 200 hotels in just one neighborhood. One of Guestmob’s core innovation is curation, we help travelers to cut through the noise of advertisement, false reviews and get to the hotel that you want. Our hotel curation system saves you the time&trouble of finding quality hotels with good reviews. We despise poor quality hotels with nasty stains or broken glasses.
    Great piece! It’s shared on our company’s facebook page. :)

    Share
    1. Isn’t one comment promoting your company enough?

      Share
      1. Not promoting, but making a connection with curation.

        Share
  6. William Mougayar Friday, January 20, 2012

    Hold on a second. Crowd-sourced social filtering is not really curation, unless one is really loose with that word’s meaning.

    Share
    1. Why do you say that, William? I think there are many different kinds of curation, some human and some not — the best systems have some of each, I think.

      Share
  7. Kevin Marshall Friday, January 20, 2012

    The acquisition is more about an email service/feature that Twitter wants to offer than it is about the filtering or algorithms that Summify had.

    They are getting a really good and talented team that has spent a lot of time and energy building an email service to scale…and they are getting them pretty cheap…it’s really a great fit/deal for both parties I think.

    Disclosure: I built http://knowabout.it (my own horse in this general space) and we are likely poised to also benefit greatly from this news (though we offer up something slightly different, I believe many Summify users will likely find what we offer fills the void very nicely).

    Share
    1. Do you know for sure that Twitter wasn’t interested in the filtering methods or algorithms? An email service doesn’t seem like the kind of thing you would have to buy a company to figure out.

      Share
      1. Kevin Marshall Friday, February 17, 2012

        Sorry for the delay in response…they may eventually use the filtering or algorithms, but Twitter actually acquired a company called Julpan last summer that was also in the same basic space…that team is in the NYC offices, and as far as I know they are the ones focused on search and discovery at Twitter…

        So I think if they were going to integrate the Summify tech, they would most likely have put them in the NYC offices rather than the CA offices (where I believe the email efforts are being built up)…

        Share
    2. I was very happy with using Summify as an app in HootSuite. Please integrate Knowabout.it as a HootSuite app! It’s so much easier to monitor and reshare right from the HootSuite interface.

      Share
      1. Kevin Marshall Friday, February 17, 2012

        Will look into it for sure…thanks for the idea!

        Share
  8. Stop using curation. You are all using it wrong, give that word back to the curators at actual museums.

    Share
  9. Gopala Krishnan Friday, January 20, 2012

    It’s content 101 duh! It is not a throwback to the early days of the WEB. It is a simple return, to a common sense content filtering and presentation mechanism that has existed in the traditional publishing industry for decades (print etc.) .”curation”? It’s the return of the content “Editor”, as understood in traditional publishing. She makes sense of the noise and publishes only the relevant stuff. People have biological limitations – they can’t handle vast amounts of information (noise) thrown at them. Atomated filtration/curation does not work either. Humans are still way more capable than machine algorithms to make “meaning” . Editors with expertise and knowledge will be the people who make “meaning”.

    Share
    1. Gopala Krishnan Friday, January 20, 2012

      That’s “Automated” and not ” Atomated”

      Share
  10. Filtering/content curation/relevance…..I think that all of these offer great opportunities for development. But, as observers of the stream of info, would too much content creation stymie creative thought/leads/ideas? Or would it just get plain boring after a while? Utter irrelevance sometimes sparks ideas!!

    Share
  11. I meant “content curation” in the last comment!!

    Share
  12. For a kick, take a look at my 15 year old model for adding value above the social layer http://www.thefrequency.tv

    Share
  13. Mathew, you said “There are so many ways of producing and sharing content but so few good ways of filtering.”

    FYI, I’ve used Summify. I also currently use several “useful” filtering tools — Trove, Percolate, Eqentia and Twittertim.es — and each has their “good” qualities.

    That said, I’m thinking that the evolutionary path is to take the “best” of those qualities and integrate them into something like Feedly (yet another useful tool).

    IMHO, selective feed aggregation (digest), combined with auto-filtering (intelligence) and refined presentation (summary) of content are all must-have qualities for the “ultimate” productivity enhancing tool — yet to be created.

    Share
  14. Steve Lubetkin, MBA, APR, Fellow, PRSA Saturday, January 21, 2012

    This reminds me of the matching technology that Lotus long ago incorporated into its DOS based Agenda product. You could import structured text files and Agenda would match it to categories in your tree. Never exploited to its potential.

    Share
  15. The future of Twitter is being less social with more people? That what I got out of the article. And another red flag is the statement in the article…”since it isn’t clear to many what exactly Twitter is for”.

    Share
  16. How about Storify.com ? Missing from this article, yet a rapidly growing player in the “curation” playing field, and used by big names as The Wall Street Journal, The Whitehouse, Le Soir, The Weather Channel, … I guess the question becomes: who of the big players will pick them up ? (storify.com, http://www.jurgenlison.com/storify_startup)

    Share
  17. Richard Creamer Saturday, January 21, 2012

    I like Steve Ardire’s comment the best.

    In general, I agree that curation will help, but this is only part of the solution.

    Implementing true semantics is still a challenge, but some form of human-assisted semantic classification and “hinting” is possible today, but would require extensions to the UX.

    But of course, many (most) users are happy with the status quo. Thus, it may well be we need two types of social networks: 1) One for those who seek substantive content and are willing to invest some effort configuring and using a richer, elegant UX (a la Apple), and 2) another for those who are happy with current offerings. Or…

    Certainly there is a lot of available content, and doing a good job of filtering out only the per-user salient content is not a trivial problem.

    But another problem exists which I have not seen anyone raise before: Even with good filtering, there is still too much relevant information to digest. Thus, I think that information distillation is a critical need (less is more). Semantics can play a role here…

    By the way, Twitter as-is, is already a wonderful source of interesting information feeds if you follow the right people. I can spend 3 minutes scanning my feed and find more relevant, interesting content/links than I have time to read!

    I am working on a whitepaper to summarize how I would improve social networking but it will be weeks before I can finish it…

    Share
    1. Richard nice thoughts and interested in taking a look at your whitepaper when you have it.

      Share
  18. Social Curation is going to be critical going forward. I have a small startup which has built a curation tool for the Twitter Timeline. Its in limited beta, so if this is of interest to anyone as a tool to use on a daily basis, please contact me for access. I’d love to get some early feedback.

    Share
  19. Useless I think. One more filter that needs validating weather it had missed critical data or not is just one more filter to monitor. When people put up crud (Facebook users in india tagging themselves in photos if models and actor and supplying fake backgrounds and educational data, for example. I’m sure it it happens all over but I can only speak for what I have seen) there is not a system in the world that validate that. It’s ok if you buy it, there just isn’t. Of course unless you let simplify twitter google+ ect ect ect access all of the info on all servers around the world they just can’t validate what they don’t know. If some of you may almost no body was excited to have google access their medical records. Even though it was an opt in thing. Why would anyone let these services access all of your data everywhere. People minimize their browsers if the see their friends walking up behind them. Get real people, we just expect things to be automatic that’s we have so much useless unmerited data out there let the app auto post this and that.

    Oh and I don’t have link anymore but former CEO of a company that already does filtering google maybe you know of them, had addressed this problem long ago before it was a problem. He said and I don’t quote now you can’t just post data you need to post data about the data. The computer does not understand jokes. End not quoting. I don’t remember 100% of the words is why I can’t quote.

    But if you ask me it’s all about how you present the data. Just driving to work I consume more data then my router can down load but I can filter it. But that’s just my option. So who got the next filter? We will have filter the filtered data once voice recognition becomes more popular. I would been done typing this in about 30 secs. Not three mins.

    Share
  20. Michael Tippett Saturday, January 21, 2012

    It is interesting to see the role that Vancouver is playing in defining the future of social news. I posted about this earlier this week: http://www.tippett.org/2012/01/20/is-vancouver-becoming-a-social-news-hotspot/

    Share
    1. You have a very good point with Zite ( by CNN ) and now Summify ( Twitter )…..Who’s next ;)

      Share
  21. >> produce a daily email with links to the most-shared content in their social networks. << Let ME control what the filter provides. There's a value in having an easier way to view the posts that are LEAST Tweeted, too. And I really, really don't want someone else fitering my content for me or have someone censoring my stream to others. My family members and I aren't volume Tweeters but we DO want to see what the other person posts.

    Share
  22. An interesting headline; but not enough data to make a statistically supported statement

    Share
  23. well, we at in-a-gist (http://inagist.com) have also been trying our algorithms to reduce this information overload. hope to find commercially interesting things from the result set.

    Share
  24. Why so much fuzz about social media? The real information lies in the Closed professional networks, shaped like social media but with business critical apps inside. That’s where you really need to organize information. I think the old intranet is dead.

    Share
  25. Suzanne McDonald Monday, January 23, 2012

    I suppose there is a danger of manipulation by algorithm. Consider mainstream media … There’s a wire service and actual humans making decisions about what should be covered for a particular audience, a la broadcast channels: CNN vs FOX.

    Share
  26. Pawan Deshpande Tuesday, January 24, 2012

    I agree that there is a void in the social media industry for services that help organize the large flow of content. Filtering, is one potential solution, but many find it to be limiting because it only cuts down the content, but doesn’t help with presenting it any meaningful way. Another, more beneficial way to address information overload without putting limits on access is content curation – the process of finding, organizing and sharing content online. Summify is a great example of a content curation tool for readers.

    There is also a growing need for curation for business, especially as the roles of marketers and publishers increasingly merge. Robust curation tools have emerged—such as Curata (www.getcurata.com), from my company HiveFire—that allow marketers to choose which topics are most important to their businesses and then create a one-stop site for prospects and customers to access the most timely, relevant content.

    Your readers may find my eBook, “Taming the Flood of Online Content” especially resourceful as it provides some best practices on curation and how it can be utilized by businesses to calm the chaos of online content. It can be accessed here: http://info.getcurata.com/ebooktamingtheflood.html

    Share
  27. Thanks for this insightful article, Matthew.

    For those looking for an alternative to Summify, you might want to check out http://focus.io/

    Every day, you get a summary of the most popular articles in your Twitter timeline. This allows you to save time, while making sure you’re not missing the important stuff.

    Give it a try, and tell us what you think!

    http://focus.io/

    Share
  28. I am a Chinese translator, and I can hardly find a proper translation for the word “Curation” in this context. It seems to mean “editing” or “filtering” of data. What on earth is its accurate meaning?

    Share

Comments have been disabled for this post