The June 2013 revelations about the NSA and its surveillance programs soured the general public on the idea of “metadata” as anything the average person might find beneficial. Alexis Madrigal restored some glory to the concept with his recent piece on Netflix and its use of metadata, but we still have a ways to go to redefine its context for the mainstream.
I’d like to elaborate on the point my colleague Hyoun Park made earlier this week about the modern mainstream perception of metadata as “evil.” Metadata isn’t inherently evil. We used to do good with it …
Back in the mid-2000s, we labeled things on the Internet, and we called these labels “tags.” Blog platforms such as WordPress and LiveJournal had recently introduced the ability to label your posts by “tagging” them with relevant keywords. Flickr encouraged its users to tag their photos in a similar manner. Del.icio.us, a then-popular social bookmarking website, urged its users to tag their links as well. These tags? Are all data about data; they’re metadata. And the websites who provided this feature ended up with databases of crowdsourced information about the data they were hosting, conveniently categorized by their metadata.
This metadata empowered web discovery to a previously unknown degree. Search engines were not an unproven factor at this point; certainly Google dominated. But how did Google decide what results would be relevant? Who was making these decisions? Compare this to a del.icio.us search: the results would tell you how many people had bookmarked a specific link, how they had tagged it, and what they had to say about it. You could click through to see a specific person’s bookmarks in the same category, and if you decided that what they had curated was valuable, discover even more relevant information on a subject. Being able to search a database with nothing but these curated, tagged links felt like you had the inside track on information gathering. Instead of what seemed like infinite pages of possibly-relevant results, you saw quality links float to the top, vouched for by others invested in collecting the best results.
Marshall Kirkpatrick wrote, in a 2010 elegy to del.icio.us:
“Tell an everyday person they can put their bookmarks online, making them accessible from any computer via a service like Delicious, and they are often amazed. Tell them they can then see other bookmarks that other people have tagged with the same categories – and they begin to see another world, a world where the Web is social and interconnected, where we all benefit from the trails of data created by one another’s everyday use of the Web.”
But this tagging took effort on the part of numerous users. Whether you were tagging a photo, a blog post, or a link, it took time and attention. And adding these tags was not something a mainstream internet user in the mid 2000s would consistently do; it required at least a little commitment, and belief that the effort to add these tags was worth the results. Though this tagging was supported and encouraged, and even seen as part of your duty to the community in some Internet subcultures, this attitude did not percolate into the mainstream. Casual internet users perceived this tenet of tagging as onerous and unnecessary, if they were even made aware of it.
Not long after, microblogging with status updates in Twitter and Facebook started gaining popularity. It grew to the point that the mainstream internet user thought of Twitter and Facebook as the whole of the social web. Tweets and Facebook posts became the defining elements of social media. Because posting a status update required jumping a lower hurdle than posting a blog post, it made having an online voice even easier. And as social media became more popular, even previously avid taggers found their commitment to useful metadata slipping, because it was easier to just share or save links for later without stopping to categorize them. And useful, usable, human-sourced metadata withered away.
Which brings us to today, where the NSA has made the concept of “metadata” evil to the average person, yet businesses still look to Big Data as a savior. Despite the negative connotations surrounding “metadata” at present, we need to use metadata because metadata makes it easier to discover and analyze data in the first place. Netflix broke Hollywood down into comprehensible pieces by applying metadata to movies. Photo mashups today still use Flickr more frequently than Instagram because Flickr has metadata and a culture supporting its use, and Instagram does not. The challenge for the future is in promoting ethical, beneficial uses of metadata. Metadata is value-neutral. It’s what you do with it that is good or evil.