Twitch Takes Traffic Hit Following Streamer Boycott Over Harassment

Twitch Takes Traffic Hit Following Streamer Boycott Over Harassment
Photo: Martin Bureau, Getty Images

Yesterday, thousands of streamers abandoned Twitch for 24 hours in protest of its ongoing facilitation of harassment campaigns known as hate raids. While the Amazon-owned streaming platform has yet to implement proposed changes to the controversial raid system, it’s clear what some previously dismissed as a small movement has had a big impact, potentially costing Twitch roughly 22% of its peak concurrent traffic for the day.

That number comes from TwitchTracker which provides outside data on platform metrics like how many people are streaming and how many are watching. While Twitch normally peaks at 4.5 million concurrent viewers on any given day, on Wednesday it hung at about 3.5 million instead. That drop came as the number of content creators streaming on the platform dropped by over 10,000 as they rallied around the hashtag #ADayOffTwitch.

Estimates by data firm GameSight Analytics showed a drop-off from the prior Wednesday as well. According to it, there were 897,745 channels streaming last Wednesday on August 25. That resulted in 2.415 million hours streamed and 65.167 million hours viewed. Yesterday, those numbers dropped to 834,107 total channels, 2.201 million hours streamed, and 54.877 million hours viewed. That’s around 16% fewer hours of content created and consumed overall. The hashtag also trended on Twitter throughout the day with over 100,000 mentions.

Twitch Takes Traffic Hit Following Streamer Boycott Over Harassment

“We support our streamers’ rights to express themselves and bring attention to important issues across our service,” a spokesperson for Twitch said in a statement to Eurogamer. “No one should have to experience malicious and hateful attacks based on who they are or what they stand for and we are working hard on improved channel-level ban evasion detection and additional account improvements to help make Twitch a safer place for creators.”

#ADayOffTwitch came about because of growing frustration over Twitch’s unwillingness to crackdown more strictly on harassment and bigotry on the platform. In light of a growing number of hate raids targeting marginalised streamers and streamers of colour in which bots spam channels with vitriol and slurs, content creators started putting together practical guides to help victims and calling on Twitch to overhaul the raid feature.

#TwitchDoBetter, a hashtag created by streamer Rek It, Raven! built up steam throughout August around a 2018 tweet by Twitch telling people to hold it accountable. In light of the latest wave of criticism, the streaming platform rolled out improved chat filters and announced plans for a better ban evasion detection system sometime in the future. But streamers want swifter, more decisive action, which culminated in plans for a September 1 boycott organised by Raven, fellow streamers LuciaEverblack and ShineyPen, and others.

Twitch Takes Traffic Hit Following Streamer Boycott Over Harassment

“I think we need to normalize being able to really make change on our own,” Raven told Wired following yesterday’s protest. The streamer said that at least initially, the day-long blackout and accompanying social media campaigns appear to be moving the needle. “People are talking about this all over the world. We have created a sense of solidarity. Twitch has responded and met with me.”

Social media platforms have been criticised across the board for facilitating the spread of misinformation, hate speech, and harassment, but Twitch has come under special scrutiny for how it directly profits off of content creators it’s done little to protect. The streaming platform takes a whopping 50% cut of their subscription revenue by default, a fact that rival platforms on Facebook and YouTube have tried to leverage by offering better deals in exchange for streamers jumping ship.

And in some cases it’s working. Just this week DrLupo and TimTheTatMan announced they were leaving Twitch for YouTube. Washington Post reporter Nathan Grayson noted on Twitter that so far the results have been mixed. While DrLupo has seen a slight drop-off in his average viewership, TimTheTatMan’s audience ballooned to over 100,000 concurrents for his inaugural steam on Google’s platform.

But mixed might be just enough to convince some other streamers to try their luck outside of Twitch at least the big ones with a built-in audience who can help pave alternate routes to success. And in the meantime it might help put pressure on Twitch to be more responsive to the demands of those content creators who have remained.


  • The result of the boycott is still nothing overall… You can guarantee they’ll take a bigger hit from the handful of streamers that just went to Youtube than the supposed thousands of streamers that took the day off.

    This sort of thing would have to be sustained, or basically straight up tank the site for the day, to get any genuine reaction from Twitch beyond what they were already planning.

    Right now it’s a blip on the radar for one ONE day of a month. And the thing people are really blind to is that it is also a day where Twitch can pinpoint in half a second the reason WHY activity dropped and go, “Well that explains it, nothing more to discuss. Carry on.”

  • People keep refering this as a boycott and not a strike, is there a reason for this? I know in the case of the NBA ‘boycott’ the wording was carefully chosen to avoid collective bargaining agreement restrictions. Does Twitch have a similar legal restriction in place over it’s streamers?

    • Because using the word ‘strike’ to describe anything other than industrial action being taken by staff engaged under standard employment contracts would be weird.

  • im sorry, but what evidence do you have that the drop is directly related to the supposed boycott? Because from the graph there doesn’t appear to be anything statistically unusual nor can you directly correlate anything with the boycott.

    • The article explains that there was a drop of around 1 million peak concurrent viewers compared to previous Wednesdays. What were you expecting, to receive 1 million letters confirming everyone’s motivations? By any reasonable measure a 22% drop is statistically significant.

      Sure, a deep-sea cable might have broken down somewhere, or Aliens might be glitching the system, but in the absence of a better explanation there’s a pretty clear correlation here. You’re just being contrarian.

      • They compared 1 Wednesday. only one.

        This statistical analysis and inferring the result of such a small set of data is a joke.

        • The dip comes across pretty clear in the second table mate. There’s a whole week of data there, unless you’re suggesting that there’s something particularly unique about Wednesdays compared to, say, Tuesdays and Thursdays?

          In any case, this ain’t an academic essay. Not every article requires PhD levels of analysis, and certainly not this one unless you have some other axe to grind here?

Show more comments

Log in to comment on this story!