Customer Login  |  

by Kevin Dodds

Large media companies are moving away from paying in-house staff to create proprietary online content and favoring two lower-cost models for content creation — user-generated content (UGC) and crowdsourcing-generated content (CGC). Since these two approaches are often confused with each other, it’s worth looking more closely at the real differences between them.

UGC is just that: ancillary content (ratings, reviews, etc.) provided free of charge to content-driven sites by users. CGC is where paid crowd workers create the core information product itself. [When users supply content for free through user polls, contests, etc., this is often described as crowdsourced. It’s more accurately called UGC.]

So which approach is “better”? Well, they each have their benefits and drawbacks.

UGC appears to be “free,” but it carries substantial costs for curation via filters, editors, and/or moderators because:

  • Users exaggerate or are simply dishonest
  • People are more apt to write about negative experiences than positive ones
  • Users often violate copyright laws

Well-known commercial examples of UGC include:

  • Yelp, where users post their experiences and opinions about retail stores, services, and restaurants.
  • Local news sites where citizens with mobile phones post photos of a local events for republication.
  • LinkedIn, where users write content about themselves that often reflects only their positive qualities.

CGC is often cheaper than in-house content creation, but its costs are often wildly underestimated. Some of these costs include:

  • Highly specialized training and credentialing before a project starts
  • Sub-routines to ensure consistent quality (i.e., editorial/QA reviews)
  • Moderators to prevent dissemination of inappropriate content
  • Software to prevent “gaming the system” in various ways
  • An experienced crowdsourcing manager to design processes and to train and qualify a group of crowd workers exactly suited for the specific task at hand.

The choice between the potentially unwieldy crowd or amateur user-generated content can be tricky. One thing’s for sure, though. Content creation still requires managers who can handle complex information processes, even if the content’s provenance is out-of-house.

{ 0 comments }

posted by Shyamali Ghosh on November 19, 2014

by Matt Manning

A recent article in the SF Chronicle gave a vivid, very visual, demonstration of the power access to “public” data can have in municipal policy debates. The story covered companies that match buyers and sellers of services often heavily taxed by municipalities: taxis and hotels. Specifically, it looked at their impact on municipal tax coffers and public policy goals.

San Francisco residents, like those in many major metro areas, are concerned about their neighbors turning into de facto hotels. This has sparked the mining and analysis of data on short-term rentals. Making sense of the situation required the deployment of Connotate’s data extraction technology to gather data on “public” rental listings. The results were overlaid with government data on licensed local hotels and B&Bs.

This shows that technology the private sector uses for competitive intelligence on market share and pricing is destined to cross over into the public sector. The ROI for just one city to compile of a list of tax dodgers for local tax assessors in this fashion is huge. An investment of $100K, for instance, could easily identify thousands of individuals owing millions in taxes.

This isn’t a first, either. Few may remember Pictometry, started in the 1980s. That company used aerial photography and the storage capacity of then-revolutionary CD-ROMs to help local tax authorities identify taxable construction projects not noticeable at street level but clearly visible from above. Again, an emerging technology found an eager government market because the returns were so compelling.

Never known for being able to move quickly, local governments have taken a few years to react. Still, with the returns on their investments in data analysis potentially massive, we can expect them to be hungry for technology that helps them to do their jobs better, cheaper, and faster.

{ 0 comments }

posted by Shyamali Ghosh on October 23, 2014