How the Retweet Button Changed the Internet

Before the button

The retweet didn't start as a feature. It started as a workaround.

On April 17, 2007, a user named Eric Rice did something nobody had a name for yet. He rebroadcast a reply from another user, Jesse Malthus, to his own followers. The next day, Rice coined the term "ReTweet". There was no standardized format. People just copied text and credited the source however they felt like it.

By January 2008, a user called @TDavid posted what became the template: the "RT @username" format, sharing breaking news about a fire in Las Vegas. The convention spread fast. "ReTweet" got shortened to "RT" because Twitter's 140-character limit left no room for ceremony.

Twitter co-founder Biz Stone watched all this happen and wrote on the company blog in August 2009 that the manual RT method was "cumbersome" but also "a great example of Twitter teaching us what it wants to be." He previewed a native feature that would replace the copy-paste ritual with a single click.

That single click would change everything.

Building the button

Wetherell had been working on the retweet project as a contractor for about nine months before Twitter engineer Ryan King joined in March 2009. Matt Knox later came on and took over implementation. Mike Limon was the sole DBA handling the database side.

The technical work was not trivial. They added a source_status_idfield to the tweet storage schema so each retweet could point back to the original. Knox built a feature-rollout system called "Decider" because feature flags were not common practice yet. It let them control the rollout in stages: internal team first, then all of engineering, then every employee, then a graph-walk targeting highly connected users, and finally a percentage-based public release.

On November 5, 2009, the button went live. One click to amplify any tweet to your entire audience. No copying, no pasting, no character budget spent on attribution. Just a tap.

Wetherell believed it would elevate underrepresented voices. That people who deserved a bigger audience would finally get one. For a while, it seemed like that was exactly what happened.

The loaded weapon

Five years after launch, Wetherell noticed something strange. He had a saved Twitter search for the word "journalism." Suddenly it was flooding with a phrase he didn't recognize: "ethics in game journalism."

This was Gamergate. A coordinated harassment campaign against women in the gaming industry that weaponized the retweet button to brigade targets, spread misinformation, and amplify outrage faster than anyone could respond. When actor Adam Baldwin retweeted a YouTube video with the hashtag #GamerGate, the celebrity amplification made the movement explode. Victims were forced to leave their homes.

"It dawned on me that this was not some small subset of people acting aberrantly," Wetherell told BuzzFeed News in 2019. "This might be how people behave. And that scared me to death."

Gamergate was a "creeping horror story," he said. The button he built to give people a voice was being used to drown voices out.

Then came the research that made the gut feeling scientific.

In 2018, a team at MIT published what became the most cited study on misinformation and social media. They analyzed 126,000 cascades of news stories across 4.5 million tweets by 3 million users, spanning 2006 to 2017. The findings were brutal. False news was 70% more likely to be retweetedthan accurate information. Falsehoods reached 1,500 people roughly six times faster than true stories. False political news traveled "farther, faster, deeper, and more broadly than any other type."

The reason was not bots. Humans were primarily responsible. The researchers found that false information tended to be more novel, triggering surprise and disgust, which are exactly the emotions that drive people to hit the share button.

The death of "retweet"

Here is what the slow death of a word looks like. "Retweet" was always a Twitter-specific term. "Repost" is the generic version that every platform ended up adopting. By Q3 2023, when Twitter officially became X, the word "retweet" had already been fading for years.

The spike in "repost" during Q3 2025 coincides with Instagram launching its native repost feature in August 2025. The mechanic that Wetherell agonized over is now so universal that the biggest photo-sharing platform in the world added it as a standard feature, sixteen years after Twitter made it a button.

The experiment that failed

Two weeks before the 2020 US presidential election, Twitter ran an experiment. Starting October 20, anyone who tried to retweet was nudged toward composing a quote tweet instead. The idea was that forcing people to add their own words would slow down mindless amplification of misinformation.

Twitter stated the goal was to "encourage everyone to not only consider why they are amplifying a Tweet, but also increase the likelihood that people add their own thoughts."

Here is what actually happened.

On the surface, it worked. Sharing dropped. But a closer look at the quality of what replaced the retweets told a different story. Of the new quote tweets the experiment generated, 45% contained just a single word. And 70% had fewer than 25 characters. People were not adding thoughtful commentary. They were typing "this" or "lol" or a single emoji and moving on.

"We don't believe that this happened, in practice," Twitter admitted about their goal of encouraging thoughtful amplification. On December 16, 2020, they reverted the change and brought back the original one-click retweet.

The experiment proved something uncomfortable. Adding friction does reduce total sharing. But it does not make the sharing that remains any more thoughtful. People will jump through a hoop to amplify content. They just will not jump through two.

Everyone copied it anyway

While Twitter was soul-searching about the monster it had created, every other platform was busy building their own version.

Tumblr actually got there first. David Karp launched the reblog feature in March 2007, just one month after Tumblr itself launched. The concept originated at Eyebeam Art and Technology Center, led by Jonah Peretti, who would later found BuzzFeed and HuffPost. Karp deliberately excluded commenting and disliking. Reblogging was the whole point.

Facebook added a Share button around 2009-2012. Internal documents leaked by Frances Haugen in 2021 revealed what Facebook knew about resharing. Users were 4x more likely to see misinformation via a reshare-of-a-reshare versus a normal post. With more shares in the chain, users were 5 to 10x more likely to encounter misinformation. In India, "deep reshares" made users 20x more likely to see false content. An internal memo stated bluntly: "Misinformation, toxicity, and violent content are inordinately prevalent among reshares."

Proposals to limit the reshare button or kill it entirely were reportedly resisted over engagement concerns.

TikTokbuilt its own repost that shares videos to friends' For You feeds with a "Reposted by @username" label. Bluesky included repost and quote post functionality from day one. And in August 2025, Instagram finally launched a native repost button, adding a dedicated "Reposts" tab to user profiles. Sixteen years after Twitter shipped the feature, the last major holdout gave in.

The mechanic that one developer called a loaded weapon is now a standard feature of every social platform on earth.

Retweet becomes repost

In late July 2023, Twitter became X. Elon Musk, who had purchased the company for $44 billion the previous year, started renaming everything. On July 31, "retweets" became "reposts" in the UI. New Terms of Service on September 29 made it official: "tweet" was now "post," "retweet" was now "repost."

The irony is that the X rebrand killed the word but kept the mechanic completely intact. In fact, when Twitter open-sourced its recommendation algorithm in March 2023, it revealed that a repost carries 20x the algorithmic weight of a like. For comparison, a reply is worth 13.5x, a profile click 12x, and a bookmark 10x. The repost is the single most powerful action a user can take on X. The button Wetherell agonized over is now more influential than ever. It just has a different name.

The question nobody answered

In 2023, a study published in Science added one more twist. Researchers found that while reshares massively amplify the reach of political news, they "do not detectably affect beliefs or opinions." People see more content through reshares, but that exposure does not appear to change what they actually think.

That finding complicates everything. If resharing amplifies reach but does not change minds, what exactly is the harm? If the 2020 experiment proved that adding friction reduces sharing but not thoughtfulness, what intervention would actually work? If every platform has now copied the mechanic despite knowing the risks, is the retweet button a design flaw or just a mirror?

Wetherell built a button that let anyone broadcast to everyone. Sixteen years later, we are still arguing about what that means. The word "retweet" is dead. The 🔁 icon lives on in every app on your phone.

The weapon was never the button. It was always the impulse behind the click.