Category: Topic Exploration

Using KipCast to build content depth

By admin, May 12, 2010 2:00 pm

I’ve been working recently with a group out of Italy called KipCast.  They have their origins in the dark days of syndication technology – back when FeedBurner seemed to be the only game in town.  In the last few years they have repurposed their crawling and scraping technology – which is highly precise – to support content republishing.  Unlike commodity web crawlers out there, they don’t just spider and crawl to build a general-purpose or vertical search index.  KipCast is used to crawl a targeted list of sites, extract the core content, clean it up, enrich it with consistent meta-data and then publish it somewhere else – usually in aggregation and usually just a summary. 

This approach certainly flies in the face of the notion of the linked data cloud – wherein content stays where it started and is never copied but, rather, linked-to.  Acknowledging that, I still think that republishing is a better approach when it comes to curation.  Deep stories and even links to related content need more context in a curated story in order for it to truly engage the reader.  Simple links just won’t do the trick imho. 

Story curation is just one example of how KipCast can be used to engage audiences.  In Europe, they have done a lot of work with listings and directories to, for example, build aggregated catalogues of items for sale from various far-flung, long-tail e-retailers.  I suppose this kind of aggregation around a purchase decision can be easily seen as curation too…

Reblog this post [with Zemanta]

Zemanta + LiveWriter = Brilliant

By admin, February 5, 2010 6:39 pm

My oh my.  I am really impressed.  I just started using Zemanta – and I can’t say I know much about it’s history, but I like where it is going.  You don’t need to use it with LiveWriter – it works with most blogging platforms, but if you use Windows, I suggest LiveWriter.  I truly hate posting content in the tiny not-so-WYSIWYG editors found in most CMS systems and blogging tools.  LiveWriter is clean and simple and, when paired with Zemanta, it becomes a pretty decent tool for curating stories.

I certainly use OpenCalais within Wordpress and have used it in some custom applications, but OpenCalais only goes so far in helping me find decent related content.  Inform and Sphere are also good alternatives, but their model seems more closed and proprietary – things may have changed, but Inform, for one, didn’t give me a chance to select the related links myself.  It just select all the related links for me, which was rarely good enough.

Zemanta sits there on the right side of LiveWriter (or your blogging tool’s edit screen), analyzes what I’m writing and goes out to find related content suggestions on the fly.  Brilliant – and it will probably just get better.  I noticed a few oddities – at least in the LiveWriter plugin.  It doesn’t always refresh the list when you start writing a new post.  Relics are left behind from the previous post, but they are only suggestions anyways and can be ignored.  The LiveWriter plugin also provides suggested in-line links for your content, usually linking off to Wikipedia for definitions.

It also offers a media gallery that, presumably, goes out to media repositories and grabs readily-licensed or public domain images that are relevant to your topic.  Given my subject-matter, I haven’t seen much in the way of relevant images – yet. 

You can also feed Zemanta with your own Flickr account and provide it with a list of your favorite feeds.  I assume this will allow me to see more related content and images from my preferred sources.  Very nice.   I’m not entirely sure what their business model is yet, but I like where they are going.

I think they need to improve a few things – and maybe this is where their business model comes in:

  • tier the suggested links by: my site, my network of sites, my content partners, trusted sources and then the broader web
  • allow me to roll over the suggested links to get a summary of content
  • provide sentiment analysis on the content showing me indicators of whether the content is positive, negative or balanced
  • allow me to help it along by selecting topic phrases out of my content; automated topic extraction is hard to do well – we believe in manual assistance, so let us assist.
Reblog this post [with Zemanta]

Does your CMS suck?

By admin, February 5, 2010 10:36 am

Seriously, I haven’t seen a good one yet – so I am forced to ask: have you found a good CMS that really opens up the potential for rich content in a digital medium?  Personally, I think we have all just been beaten into submission by the conventions set by CMS systems from their earliest incarnations.  With only slight variations, these tools give us a text editor, usually TinyMCE and unusable forms for attaching meta-data to articles.

CMS vendors seem to have been drawn into focusing on functionality related to workflow, multi-lingual publishing, scalable page serving and the like.  I think they tend to care more about corporate clients, marketers and e-commerce plays than they do about media companies.  Even if there are exceptions, they have done little to advance the functionality that helps us create great, well-curated stories.

Most people would argue that blogging tools help us build deeper stories with their widgetry and overall “freeness” and ease of use.  I agree that blogging tools have done a lot to bring storytelling to the masses, but they won’t really help publishers create stories that can be effectively monetized.

Both blogging tools and CMS systems are far too focused on the individual article or post – and not the art of how they can be aggregated into something more useful and engaging.  Sure, you can plug in widgets that support rich media and you can easily create links to other stories, but this isn’t nearly enough in our view.  Blogging tools – WordPress in particular – excel at using meta-data to link related content, but, again, that isn’t true curation.  Curation, in our view, is about anticipating how users want to explore topics and then using your expertise to assemble the right content or, at least, pointers to the right content.  In our experience, better monetization depends on creating experiences that will be explored.

I’ve yet to see a publishing system or CMS that does this.  If you know of one, let us know.

Story Analytics

By admin, January 31, 2010 7:59 pm

We have all worked with web analytics tools at one time or another to understand how well the site is performing.  Google Analytics,  Omniture, WebTrends – all of them are great tools – but we are not convinced we use them as well as we could.

The media properties that we have been associated with tend to look at total site traffic in aggregate over long periods or, if they ever get into the details, they tend to look at the site section by section.  Often, they will even look at the performance of an individual article.  In fact, it is at the most granular level that analytics become the most helpful and drive the most insight.  Sadly, we have rarely seen much insight drawn from these numbers.

We propose that media companies need to turn web analytics on its ear.  They need to take a bit of a different cut on the data to make it really meaningful.  Also, no one tool can currently deliver the type of data that good curators will need to manage their portfolio of topics.  We haven’t seen a perfect combination of tools but, looking across sources, here is what we should measure:

  • leading indicators of interest in a topic: even before the first article gets written, use data to help identify what we should be writing about.  Demand Media, arguably THE biggest media success story of the last few years, bases their business on this practice.
  • story type and investment profile: patterns in data related to a story or topic can suggest a certain trajectory or story type; e.g. evergreen with growth potential vs explosive but short-lived types.  Different story types deserve different levels of investment in time and resources and, likely, create different monetization opportunities.
  • performance of story elements and multi-variate tests: determine what story elements work the best; this will likely differ by the type of story.  Use A/B or multi-variate testing if you have access to the required tools.  This will tell you if you are assembling your stories in a compelling and usable way.
  • depth of visit/engagement in the topic: at a glance, measure how many pages deep your users go in a specific story.  Sure, topics can be linked and the lines between stories can be blurred, but do your best as this is a critical measure.  While depth of visit on the overall site is an important traditional measure, engagement in a single story more directly correlates to the performance of contextual ad units – which will become  a critical measure in its own right.
  • contextual ad unit performance: while not all ads can be measured this way, those with a clear call to action and a conversion objective ought to be measured within each individual story to see if the rich context of a well-curated story makes a difference.  This is a critical selling point of story curation.  For this to be meaningful, the story topic must have some contextual link to the ad itself, i.e. the ad should relate to the story.  We expect that some types of stories will perform better than others, but, overall, good curation should yield the kind of ad unit performance numbers that you will want to take to media buyers and advertisers.

These new categories of measurement, plus all of the other traditional analytics data, are crucial to effective curation.  In fact, it goes beyond just helping tune the story, it also helps prove the model.  Supportive data across these areas of measurement is ultimately critical to proving the benefits of story curation – including the monetization of properly curated content.

As far as publishing goes, gone are the days of post it, forget it, move on.

Topic Exploration versus Exploitation

By admin, January 31, 2010 2:11 pm

Curating for Topic Exploitation

Google and other search engines, including emerging semantic search, do topic exploitation very well.  You give the search engine a topic keyword or phrase, and it will do its best to show you relevant articles for that match the topic.  For curators, exploiting a topic according to a keyword is important, but Google will always do it better.  We think good curation considers exploitation, but techniques for enabling exploration are more important.

Curating for Topic Exploration

A great curator has a sense of how a topic ought to be explored.  That is the art in the science of publishing to the digital channel.  Google will never do this very well.  Exploration can branch in many directions, depending on the user, and providing useful paths for exploration is not easy but, if done well, your users won’t go back to Google right away and your all important engagement metrics will improve dramatically.  Advertisers and sponsors will learn to appreciate topic engagement because (and depending on the topic) it provides a great context for promoting brands, products and everything else that advertisers care about.  Ironically, the best example of curating for exploration is Wikipedia – a web property that can’t even be monetized.

Panorama Theme by Themocracy