Recommending Content and Automated Sharing of Content
September 14, 2011 14 Comments
You find a great blog, read regularly, and tweet or share everything posted. So why not automate sharing what you already share, and get a few minutes back each day?
On the surface, it seems logical. However, if you are using automation, it likely is having side effects.
- Automation has made your algorithm, primarily your selection of sources, more important than the content itself.
- You share more than you used to. I follow people that, with the addition of automation tools, now share more content than I have time to read, even if they were my only content source.
- It has disconnected you from your best content sources. Your time savings are from spending less time reading, considering or commenting on content from the best sources, those you are willing to automate. The simple act of sharing is a minimal time savings.
- Your sharing does not include a comment or note that adds context for your audience. The comment or note improves the recommendation, helping your audience see what content is right for them.
These side effects mean your content sharing is no longer a valuable content filter for your audience. Although much of it may be worth reading, we are flooded with content from multiple sources today. A content feed cannot compete with the recommendation value of carefully curated and reviewed content shared with additional comments.
If sharing content that you don’t recommend your audience reads, why do you share content?
- Are you creating a dedicated audience by sharing content you wouldn’t recommend? If so, how valuable can that audience be to you?
- Are you looking to increase engagement through increasing sharing? If so, how effective can you be striking up a conversation about content you haven’t read?
Alternatively, is your sharing algorithm good enough that you are sharing great content and building a valuable audience, despite not being able to individually endorse each piece of content?
Your Turn
What is your view on automated sharing? If you are against automation, what side effects bother you? If you support automation, how do you feel about the side effects listed here? I would love to hear your view in the comments below, or share it with me on Twitter at @wittlake.
About Eric Wittlake
I am a digital and B2B marketer with a background in online media and analytics. I work with B2B clients on media and integrated marketing programs. You can connect with me on Twitter at @wittlake or in the comments here on my Digital B2B Marketing blog.
I am a digital and B2B marketer with a background in analytics and online media. Based in Portland, Oregon, I run the media group at 
Digital B2B Marketing 
Amen. Great summary an I agree!
Thanks Mark!
Hi Eric… Timely post. I am in my third week of testing Triberr, an automated content sharing tool. At first, I was sharing everything from my Tribe and felt like I was spamming my followers. At the same time, it was super cool when 40 people shared my blog posts.
By the end of the second week, I turned off the Auto Sharing mechanism for most of my Tribe mates, which queues up tweets and you can decide which ones to share. So far, this seems like a win-win approach. I monitor the pending tweets and only accept the content I like. And, it does save time because the tweets are already written.
I wrote a short post on this topic on my blog: Should You Join Triberr To Boost Blog Traffic? Maybe.. Maybe Not
http://thesocialmediadiary.com/2011/09/02/should-you-join-triberr-to-boost-blog-traffic-maybe-maybe-not/
Thanks for getting the conversation started!
Marci, thanks for the comment. I use Triberr also, I am now 100% on manual. With apologies to my tribemates, I view it more as an RSS reader with a better integrated tweet functionality. For that purpose, it works well, and the quid pro quo is that we put each other in our readers and take the time to at least briefly review the content.
Mark, who made a brief comment on this post, is someone I look up to, and through Triberr I’m now on his reading list. I don’t believe he read this blog often before we landed in the same tribe, gaining readers like Mark and yourself has been a benefit of Triberr, but notably, that benefit isn’t dependent on automation.
Thanks for taking the time to read and comment, I appreciate it!
While your arguments are on general level, I’d like to point out that there are at least two ways that automated content curation can work: source-based (like Triberr) and algorithm-based (like paper.li).
There are also important differences between the two. If you fully automate your content curation based on keywords, the results can be pretty much anything, really. On the other hand, if you fully automate it based on source, this should ensure better quality as you have a much more accurate idea on what to expect from a particular source as opposed to a particular keyword.
I’m afraid your questions to readers are formulated in a way that makes them mere straw men: sharing content one does not read and does not recommend, these are more of populistic moves than serious questions.
I will try to ignore that part and answer the basic, essential questions:
1. Can one create a dedicated audience? Yes, why not. The content is selected based on some criteria, and sharing content based on a set of criteria can produce value, and therefore interest an audience.
2. Can one discuss the content shared this way? Yes, why not. There is nothing to prevent the one who shares content automatically from also reading all of it. In fact, if the curation is successful, and the content is of value, the algorithm-master would naturally want to read all of it as well.
Faced with information overflow, we will need to increasingly automate our content discovery in order to find interesting content from amidst all the noise. Automating the sharing part of content curation is just one more step from there.
However, it should be noted that both Triberr and paper.li have had to add more editorial features to their services, and it does not look like fully automated content curation is a viable option at present.
In my opinion, for it to become viable, it has to include both source-based and algorithm-based elements, just one or the other is not sufficient. I’m still not sure if it can be made perfect even by fine-tuning both of these.
I recently discussed this same issue in more detail in my blog: http://www.kilkku.com/blog/2011/08/klout-triberr-paper-li-and-the-future-of-content-curation/
I think you are confusing several concepts.
For one, you say “automating the sharing part of content curation”. If you are relying on an algorithm to share content, you are not curating the content. Curating content is about amplifying a message. If you amplify the noise along with the message, the message is not clearer. Curating is as much about filtering out noise as it is about amplifying the signal.
For another, you say “we will need to increasingly automate our content discovery.” Auto-sharing content is not content discovery (it’s the opposite) and doesn’t aid content discovery. In fact, it clouds the field. Automated resharing is a large part of the reason we now NEED automated discovery. Or we don’t need it. I don’t NEED to be force-fed “content.” When I have a question, I ask it and I can typically find an answer quickly and cheaply. I don’t NEED a constant stream of answers to questions I haven’t asked.
By resharing, what are YOU adding to the content? What do you provide your audience by using Triberr that they cannot provide for themselves by using Triberr? Why would an audience be dedicated to YOUR information stream? Are your selection critereria so novel that they could not have thought of them? How does your algorithm specifically address each of their personal desires for knowledge better than an algorithm they might create? What are YOU adding to this already too noisy channel?
Curating content is like tending a garden: you cannot make your flowers stronger than nature has, so you must kill the weeds. Automated resharing is like the wind: it distributes all the seeds all together without thought or plan. Be the gardener, not the wind. It takes more work. If it takes too much work, do not be a curator of content. You don’t have to be, you know. There is plenty of content out there growing like dandelions in every crevice. It really doesn’t need to be encouraged or nurtured, it does pretty well on its own.
I do realize that some definitions of content curation rule out automation. However, I do not agree with such a definition.
Simply put, I think content curation is the searching, selecting, and sharing of valuable content on a specific subject. Making the definition itself exclude automation would be unjustified. Now, whether automation can fill these criteria is a fertile subject for discussion.
Perhaps my sentence on automating content discovery was not clear enough, although I did mention that sharing is just one step from it, not identical to it. Automated sharing is not content discovery, obviously.
Regarding your questions, I find that they address the key issues with automated curation quite well. Yes, I believe that it may be possible sometime in the future to create algorithms that discover and share relevant content. I also believe that the combinations of author data, keywords, and negative keywords to achieve the best results are complex and need to change over time. Therefore, it is perfectly imaginable that people who are experts in a given field and interested in content curation could spend considerable amount of time tuning such criteria, and that by so doing produce an automated stream that, indeed, could not be produced by just anyone. And, as I already mentioned, I don’t think Triberr as it is today is on such a level (but it could become such a tool, depending on the direction it is developed).
If we take your garden analogy, would you say that a robot could never tend a garden? If you do, there is an obvious philosophical difference between our views. My view is that yes, a robot could tend a garden. We just haven’t built such a robot yet.
Ville, great point. The definition of curation should not exclude automation. LinkedIn Today curates purely based on popularity (which doesn’t necessarily mean it is the most relevant content to me yet).
I will disagree a bit on your earlier point about Triberr, or any other curation methods based on source. If you are using Triberr on manual (disclosure: I do), then it is just a workflow, it streamlines reviewing content to share. Reader2Twitter will let you do this within Google Reader, but it doesn’t have any of the prepopulation benefits. Similarly, Paper.li I see as a way to streamline the process, to surface content for “potential” curation, but not a tool that can really automate curation yet. But I wholeheartedly agree, it doesn’t mean curation and automation must be mutually exclusive in definition.
On your original point two, if the algorithm is good, and therefore I want to read all of the content, what is the real benefit of automating the sharing, versus building a custom feed with a limited amount of content? If I want to spark conversation by sharing, I can almost always spark more conversation by adding a question, comment or perspective when I share the content. That said, to your point, that doesn’t exclude it is a viable option for someone who develops a great algorithm and wants to become a key content source, versus start conversations and develop two-way dialogue.
Thanks for taking the time to comment, I appreciate the perspective you have added here.
Eric, regarding Triberr, going manual is actually one of the editorial features they have added. In the beginning, Triberr was full automatic, no options. In manual mode, it is, indeed, not an automated content curation tool.
Actually, one key question for Triberr is how they are able to reconcile the split in their community between manual users and automatic users, as that has to be the most hotly debated issue around the tool.
Regarding commenting when sharing, it is a practice that I am also quite fond of, and I like to know what people who share items think about them. It really sets the entire stage for clicking that link and starting to read.
I haven’t had any measurable success while doing it myself though, I don’t think anyone has ever commented on my comment. Perhaps my comments are just dull. Anyway, based on this experience, I don’t see sharing without comments as a big issue, and also see some value in the small amount of time saved by automating the process.
BTW, is the comment system limited to three levels deep? At least I couldn’t find any Reply buttons on the third level. Not really encouraging discussion, that.
Eric,
I have not used automated tools for sharing, and when I do share something, I have read the article, and I do post something insightful about it. To do this means that the content has to be pretty darn good.
More importantly, it doesn’t have to be marketing related (my industry). I think about my own reading habits and know that I like original, fresh comment that makes me think or even better, laugh. If I like the content, then I’m pretty sure my followers will, too.
I have a number of Twitter followers who automatically tweet out my content, something that bothers me. In turn, some of their followers automatically tweet out their automated content. So what you end up with is spam and noise, plain and simple. I made a vow a long time ago to not add to the noise — which means I’ve cut down quite a bit on what I share.
Note to self: stop tweeting Dianna’s great content so quickly. 🙂
Thanks for the comment, and the perspective on diversity of content as well. It is a great point, if we are looking to establish relationships via social media, we need to allow some insights into elements of our lives beyond marketing, content, social media, or whatever pure “professional” focus we have.
Thanks again for taking the time to read, comment and share, I appreciate it!
Great post, and great dialogue. Thank you Eric…and everyone who has chimed in on the debate. I am personally still on the fence. Cheers!
Eric,
So I think I started this and I would like to weigh in. As you know I wrote about this over on my blog. The question I was asking was really whether I pissed off my core of valued Twitter followers as opposed to the hordes of followers I picked up as a result of the automated sharing.
A couple of points:
1. I would argue that although I am not curating the content, I have filtered the authors who I have already vetted based on their valuable content. Does a bad post slip in? Sure. Just as I write a bomb now and again. But for the most part, I am sharing great content from great authors who I respect. I am basically sharing the content from all the bloggers I read although I don’t always read all the posts.
2. The small amount of automation has helped me reach my larger objectives of growing my audience and being able to spend my time interacting vs. curating.
3. The comments and tweets have been strongly weighted on the positive. Does this make you somewhat of a “purist?” Maybe 😉 I totally get and value your perspective. I think I’m viewing my tweets as somewhat less important individually (each tweet as a personal recommendation) than the sum of all the parts (total audience plus engagement plus interaction).
I mean, in the end, I get all the negatives but to me they out weighed the positives. Now, if you and a whole bunch of others started unfollowing me or hatin’ on me, I would stop immediately. Promise. Just please don’t start a hatin’ campaign…
Seriously, I do appreciate the insights and honesty.
Now you call me a purist? I’m tempted to say “Thanks!” 😉
Great points Michael, and yes, your blog post at http://www.b2bmarketinginsider.com/social-media/twitter-by-the-numbers-are-you-listening-to-100-million-voices sparked this post. As you know, as you start writing about an idea, it tends to shift some as well.
Similar to my comment on your post, I started with the idea that we are not really recommending content, or we are recommending more content than our followers can possibly read, as automation not only removes the filter but also increases volume. On the flipside, traditional marketing says frequency is important, and most people don’t see every tweet in their stream. Dark content days are not great for engagement on Twitter.
If I sum it up, your position (at least for your current trial of automating more sharing) is that sharing a few things you would not share otherwise is an acceptable price for the additional time it frees up that you can spend engaging and discussing. Is that a fair summary? If so, won’t refute it. It is an interesting position, and I think a healthy reminder to me and everyone else that what we do on our blogs, Twitter and elsewhere in social media isn’t an end in itself, it is a means to an end. When we get caught up in dictating a single path (as I sometimes do), we lose perspective on the purpose we have in the first place.
No “hatin’ on Michael” campaign will come from me, I appreciate your challenges, support and inspiration!