Dear Google, Fix Your Algorithms—They’re Killing Journalism
By Minda Zetlin | From the June 2013 Issue
You and I have had our differences over the years, most notably when ASJA helped scuttle the settlement between you and the Authors Guild over the illegal scanning and dissemination of more than 30 million books. I continue to believe that your scanning amounts to theft, but this letter has nothing to do with that. It has to do with your core business, the thing your company does best: Search.
Your insight that you could crowdsource website ranking literally launched an empire. You judged that a site with many “inbound” links is one that people find valuable, and so the people using your search would be likely to prefer sites with lots of them. Your judgment proved spectacularly right, and users the world over, myself included, learned to trust Google results and use Google as our default search engine. Today, despite Microsoft’s and Yahoo’s best efforts, more than 65 percent of Internet searches go through your search engine.
But your algorithms don’t only rank for inbound links—there’s a lot more to them. They also rank for the most recently refreshed content, and that’s what I’m writing you about today. Google fellow Amit Singhal explained in a 2011 blog post that you rank for freshness of content because users are likely to want the most up-to-date information possible. I respectfully disagree.
Someone searching a sports event minutes after the end of the game is likely looking for the final score. Someone searching “Boston bomber” during the manhunt for Dzhokhar Tsarnaev probably wanted to know if he’d been caught yet. But other times, maybe even most times, users want the best information, not the most recently posted. Someone searching “How to safely remove embedded Lyme tick” doesn’t want a post by a 15-year-old blogger about how his girlfriend pulled a tick out. He or she wants a doctor-tested set of step-by-step instructions—even if that information has been online for more than a year and the 15-year-old posted his five seconds ago.
So weighting fresh content as much as you do means that at least some people aren’t finding what they want. But that’s only part of what’s bad about it.
High ranking in Google searches is incredibly valuable, and website owners (including ASJA) go to a lot of trouble and expense in pursuit of improving theirs. Knowing you rank for freshest content, websites far and wide are in a death race to add new material as often as they can. In other words, your algorithms strongly prefer quantity of content over quality. That’s what’s killing journalism. I know, because I’m watching it die.
“Content mills” have sprung up to supply huge numbers of articles very quickly so that sites can constantly have fresh content. These articles are written by poorly paid people who have little expertise and can only afford to spend half an hour or so crafting their pieces. That usually leaves no time for research, but since there’s no fact checking for these pieces, it’s no problem if they make stuff up.
What research they do have time for is strictly Web research—finding information on other sites that they can refer to. Web research is great—I do a lot of it myself—but if everyone’s research consists only of finding stuff other people have posted online, that’s a problem.
For a while, I was writing for a major business website that, in search of fresh content, published an aggregation blog. New items had to be posted 20 or 30 times a day. The way we came up with these items was to find interesting stories on other sites, write a few paragraphs about them, and put in a link. Sometimes, looking for interesting stories, I’d click on a link only to find the website I opened was doing its own aggregating, posting, and had published a post about a story on yet another site. Sometimes I’d follow this breadcrumb trail three or four steps before I finally found a news organization that had done some actual reporting.
That put them at an economic disadvantage. Here they were actually paying reporters to go out and find out stuff when they could have had more fresh content and a higher Google ranking simply by linking to someone else’s reporting. On the other hand, if everyone was linking and no one was actually doing interviews or traveling to where news occurred, the whole system would break down.
The problem for journalists, and for journalism, is that sites needing new content 20 times a day aren’t likely to pay well for it. Pay for professional writing of all kinds is dropping. As a result widely published journalists have been moving into completely different professions, or turning to writing for corporate clients so as to continue making a living. While it may benefit corporations such as Google to have access to their talents, it’s not good for journalism, or for the reading public. The drop in pay for writing isn’t entirely due to your search engine, of course—but it is a big part of the problem.
It could be part of the solution. Two years ago, an executive at the public relations firm Edelman made a splash when he predicted a future in which a readership overwhelmed with constantly refreshed content would inevitably become much choosier. That same year, it was widely reported that your “Panda” update was intended to disfavor content mills, though you never officially confirmed this report. Still, for a while I hoped I was seeing the dawn of a new age.
But I’m still waiting for search that ranks quality of content over quantity. I know it can be done. You could rank for the reputation of the content’s author (does his/her byline appear in major newspapers and magazines?) the number of sources quoted, for charts and diagrams. You could rank for how often that content, and/or its author, is clicked or shared, in keeping with your original philosophy of crowdsourcing. I’m sure there are many more, and better, methods I haven’t thought of. Despite Apple’s claim, your company employs true technological geniuses. You can figure it out, if you choose to.
I hope you choose to. Because right now, thanks to your algorithms, not only are the best writers I know quitting journalism, an endless number of websites are cluttering up the Web with an endless supply of nearly worthless content. And most people don’t want to drink from a fire hydrant of the freshest verbiage. They want well-researched, well-written content they can trust.
Columns, From the President's Desk. Bookmark the permalink.