News that Twitch might introduce a ‘Brand Safety Score’ spooks LGBT+ streamers
Streaming platform Twitch could be about to introduce a Brand Safety Score to assess streamers based on how “brand friendly” they are.
The news was unearthed in Twitch’s API by cybersecurity student Daylam Tayari, who revealed streamers could be graded on attributes like chat behaviour, ban history, games played, Twitch partnership status and more.
This score will then pair the streamer with suitable advertisers on their stream and may also be used for Twitch’s Bounty Board where creators can accept paid sponsorship opportunities, with the relationship between brand and streamer handled by Twitch themselves.
Twitch has added an automatic Brand Safety Score which grades how brand friendly every streamer is based on things like chat behavior, ban history, manual ratings by Twitch staff, games played, age, automod and more (See below).
— Daylam 'tayari' Tayari (@tayariCS) March 9, 2021
Twitch are yet to officially acknowledge the existence of the Brand Safety Score, though they did respond to Tayari’s tweet saying “we’re exploring ways to make sure ads are appropriately matched to the right communities on Twitch, looking at a number of different factors. Nothing has launched yet, and no personal information was shared.”
Hey tayari, we’re exploring ways to make sure ads are appropriately matched to the right communities on Twitch, looking at a number of different factors. Nothing has launched yet, and no personal information was shared.
— Twitch Support (@TwitchSupport) March 9, 2021
Twitch did offer a comment to Engadget. It reads:
“We are exploring ways to improve the experience on Twitch for viewers and creators, including efforts to better match the appropriate ads to the right communities. User privacy is critical on Twitch, and, as we refine this process, we will not pursue plans that compromise that priority. Nothing has launched yet, no personal information was shared, and we will keep our community informed of any updates along the way.”
In theory, this score is a sound way to ensure advertisements on streams are appropriate. It appears similar to existing ad-rating systems used by Twitter and YouTube.
However, one person recalled an incident where all LGBT+ videos on YouTube were flagged as “not family friendly” after a similar update.
Remember when youtube did this and every lgbt video on the website was flagged as not family friendly https://t.co/qIQ4FUqfoG
— Naomi Σ (@perfectsIumbers) March 9, 2021
Another critic tweeted that the move would “f**k up LGBT+ streamers hard, because apparently LGBT+ is still advertiser unfriendly according to these brands. Seriously. This will not last too long unless Twitch wants to f**k up people even further.”
Many streamers were similarly concerned, with calls to ensure the attributes are transparent. There are also concerns that streamers who are particularly outspoken about political and social issues may be given a lower safety score.
"Brand Safety Score" okay????? I'm wondering how transparent Twitch will be with this (probably not at all??) and how fair it is to be evaluated based on a biased system (e.g., if you're an outspoken streamer about social issues, does that make you less brand safe???) 🥴 https://t.co/Njm9g2zR8T
— ☕️ ELA 🌨 (@TheCafeEla) March 9, 2021
What scares me the most is "manual ratings by Twitch staff" because that is a great way to ensure bias in how ratings are applied. Marginalized communities will always be loud about issues that affect them and I can see those same voices getting negative ratings as a result. https://t.co/nfIwoazJ0s
— Lucia Everblack💀🏳️🌈⛧🏳️⚧️🔮 (@LuciaEverblack) March 9, 2021
And when Twitch’s own conduct has been called out – most recently controversy over their Women’s History Month content – there are concerns as to whether they’re equipped to run such a programme.
I’m more than capable of showing brands what sorts of things we stand for.
I do not need Twitch staff to give me a rating when they can’t even figure out how to use social media without offending entire groups of people several times a month.
Work on your own staff first. https://t.co/sjAhp4X3Hc
— ✨ Crev ✨ (@CrevLMTV) March 9, 2021
Manual ratings by Twitch staff?? That's a scandal just waiting to happen. I'll have my popcorn ready
— Dalmar Wiltshire (@sirwiltshire) March 9, 2021
This is why brands should & always should have someone manually vet people for opportunities. This goes for any company. Not sure how the tech of this will run but yeah if someone has been banned on twitch several times for misconduct, probably won’t be someone I am working with. https://t.co/ereJZA8pWW
— iamBrandon🏳️🌈 (@iamBrandonTV) March 9, 2021