A Good Editor: Quality Assurance for Language & Meaning

In July, I linked to an article about an experiment conducted at IBM to measure the impact of an editor on content. That study found a quantifiable improvement: a 30% increase in reader response to content that had been edited versus raw content.

Besides overseeing and improving the quality of content, editors are also responsible for keeping the trains running, so to speak. In Real Editors Ship, Paul Ford explains, with examples, why projects need editors:

I recently left zineland and did a bunch of freelance work and hooboy do people not know how to ship. A three-year project that yielded only 90-second page load; or $1.5 million down the drain with only a few microsites to show. And I’ve started to find myself going, God, these projects need editors. Editors are really valuable, and, the way things are going, undervalued. These are people who are good at process. They think about calendars, schedules, checklists, and get freaked out when schedules slip. Their jobs are to aggregate information, parse it, restructure it, and make sure it meets standards. They are basically QA for language and meaning.

Ford explains what goes into the daily job of “shipping” All Things Considered, NPR’s flagship news program:

I remember when I used to write for All Things Considered, my editor there sent me a few pictures from the whiteboard they used to put together the show. It changed constantly throughout the day; they kept a webcam trained on it (this was a few years ago; maybe they use websockets and node.js now). There were an insane number of variables that went into creating that big hunk of nightly audio: Recordings created months ago or two hours ago; people working together in a dozen time zones; contracts, permissions, fact-checking. It had to fit together technically; it had to be transmitted efficiently at a high bitrate to maintain quality (but may be sped up or slowed down to the limits of Fourier transforms); it had to be edited to match certain durations; it had to have a certain consistency and flow; and so on. It requires the human equivalent of map-reduce to manage it. And they—meaning editors and producers—managed a release every night, with 12 million users.

Apologies, for this post the comments are closed.