In the midst of the chaos of the Dallas police shootings, at 12:52 a.m. Eastern on Friday, the Dallas Police Department tweeted a picture of a man they called a “suspect” and asked the public to help find him. It soon became clear he was innocent, but the tweet remained, and people continued to circulate his picture by the thousands. The Dallas police finally deleted the tweet Friday evening, after it had been retweeted more than 40,000 times and embedded in news stories around the Web. Yet the man in the picture is reportedly still in hiding, fearing for his life.
Deleting the tweet was probably the right thing to do, under the circumstances, even though it should have been done much earlier. Another option would have been for @dallaspd to reply to or quote its own tweet with a correction, declaring the man in the photo innocent. But that would have left the original tweet unaltered, for people to go on retweeting and viewing on news sites. Deleting the tweet at least halts its spread.
Yet deleting an inaccurate or otherwise problematic tweet is a seriously flawed solution, especially in cases where the original tweet was newsworthy, or when it has already circulated widely. The tweet simply disappears from the Web, replaced by an error page that tells you nothing about where it went or why.
Replies to the tweet, and retweets that quoted it, are stripped of their context.
News stories that embedded the tweet are similarly affected, rendering them retroactively confusing or even nonsensical.
The worst part of deleting an important tweet is that it precludes the possibility of replying to it with corrections, clarifications, or even apologies. Not only does it cut off discussion of the tweet by third parties, but the person or account that tweeted it can no longer leave a threaded reply with the correct information.
The solution to this problem is not a mystery. People have been asking for it for years: an editing function.
Allowing users to edit a tweet would solve multiple problems. In cases like the Dallas PD tweet, it would allow the police department not only to stop the spread of a damaging tweet, but to correct the record, clarifying that the man in the picture had nothing to do with the slaughter of police officers. Ideally, the edited tweet would replace the original, not only on Twitter, but everywhere that it had been embedded across the Web. It still wouldn’t reach all of the people who saw the original—corrections rarely do—but it would reach some of them, and a follow-up tweet could put the word out more widely.
There are also, of course, plenty of less fraught reasons why people might want to edit a tweet, such as correcting a typo, adding a link, or swapping in a different photo.
An editing feature makes so much sense that it’s almost hard to believe the company has existed for over a decade without one. A year ago this month, Twitter CEO Jack Dorsey himself seemed to endorse the idea, in response to a request from Kim Kardashian West.
Yet as Twitter has introduced one controversial service change after another under Dorsey’s leadership, the edit button has not been among them. What gives?
We got a partial explanation last fall from Kevin Weil, then the company’s head of product, who said at a Recode conference in October that nothing was imminent. “There are real challenges to editing tweets after you post them,” he said. For instance, an edit that changes the meaning of a tweet could present problems for those who had rebroadcast it, because their commentary around it might no longer make sense. In the worst case, a user could theoretically tweet something benign, get it circulated widely, then change it to something offensive or graphic.
OK, so: There are challenges. If only Twitter had a staff of hundreds of highly skilled engineers, designers, and product managers, plus $3.5 billion in cash on hand, perhaps it could solve them!
An obvious first step would be to build the edit feature in such a way that an edited tweet is marked as such, and includes an option to view the original so that you can see exactly what was changed. If you really wanted to minimize the risk of abuse, an edited tweet could be grayed out on sites where it’s embedded until the reader clicks to reveal it. You could even delete the tweet itself and replace it with a message that the tweet has been edited, along with a link to view both the original and the edited version.
Some of these proposals are admittedly a little clunky. But there’s little doubt that Twitter could find a way to allow edited tweets if it considered it a priority.
The real question, then, is why Twitter doesn’t consider it a priority. Twitter itself won’t say, and the company had no comment when I inquired. But I have a theory.
My theory is that an edit feature is the type of feature Twitter would have implemented long ago if its main goal were to make the service better for the people who use it. Of course Twitter does care about making the service better for the people who use it, and, to some extent, about the public interest as well. But ever since it went public in 2013 at a valuation of some $25 billion, the company has been under intense pressure from investors—and, oddly, the media—to grow much larger than it is today. One CEO and a slew of executives have already either stepped down or been ushered out for failing to overhaul the service in pursuit of rapid expansion.
The result is that Twitter can no longer afford to prioritize the interests of the people who use it. In order to vastly expand its active user base, the company must instead prioritize the putative interests of people who do not use it. For people who do not use Twitter today, an edit function is highly unlikely to be the change that lures them in at last. And that is why I believe we have not seen it, and probably won’t see it anytime soon.