This is an interesting change of heart for the “King of Automation” -
Megan Garber / Nieman Journalism Lab:
Google News gets a new human touch, launching publisher-curated Editors’ Picks as a standing section — When Google News launched in 2002, it did so with some declarations: “This page was generated entirely by computer algorithms without human editors.” And: “No humans were harmed or even used in the creation of this page.”
————————————————————
Even Google – the king of automation – has decided that consumers sometimes want the human touch of someone deciding what might be the most interesting or appropriate content.
We are firm believers that there is no one answer for how to best curate content. Sometimes full automation and algorithmic curation is best because it has almost zero costs and often surfaces content no human could ever find (data set too large). However, even the best algorithms have a hard time understanding “appropriateness” or “taste” or the fit with a particular Brand or audience. For this reason, publishers must have access to tools and platforms that provide granular control over exactly what content is published and where. This approach may cost a bit more up front in human labor but it also can provide big returns when the editor/curator finds that perfect fit of content to audience and drives up adoption, page views, consumer satisfaction; and, hopefully lasting consumer loyalty.
While full automation and granular manual curation are good, there is wide gap between the two that can be filled with innovative products that merge the efficiency of automation with the control of manual curation. At VA, we have launched two new products/platforms to serve this “in-between” space.
A couple of months back we introduced the first content “White listing” platform. This enables any publisher to surf the web and with a single click – authorize content for algorithmic selection. In essence, publishers can create their own candidate set for the automated system to choose from. In this way, editor/curators can ensure that every potential piece of content is explicitly approved for use on their site without having to manually place it there every day. Now there are three distinct syndication options: by partner site (algorithm selects from all article on site); by article/video (editor manually places on each page); or by White List (editor/curator gets to create a custom candidate set or long tail which the algorithm chooses from). The cool thing about White Listing is that it is cumulative. Over time the authorized set of approved content grows with each new white listed item.
As we have learned the hard way – one piece of inappropriate content on a publishers site can do more damage to the brand that the good that comes from millions of new page views. While manual curation solves this problem entirely, it adds cost. White listing is better because it leverages all prior white listed articles and allows publishers to build cumulative “long tails” of content they explicitly approve — but is still requires a human to look at each potential page. Since none of these options is perfect, we developed a Rating Engine which fills the gaps between automated and manual curation. At VA, we have team of content analysts that rate content for age appropriateness, sexual content, violence, and/or drugs/alcohol. This platform allows publishers to select partners for content syndication but with a few clicks limit that content to any of the rating criteria.
As even the king of automation has determined (Google) — the ever changing world of online publishing is not a one size fits all world. Sometimes automation is great – other times us humans still have an important role to fill. In either case, new platforms and technologies will continue to be invented that make the process easier and more profitable.
