6 months ago I started streaming as a social outlet during the pandemic. It's been phenomenal for connecting with others and breaking up the monotony of quarantine. Below is collection of observations that would have been useful to me when starting out. (1/29) https://twitter.com/RiotNu/status/1333218934464516096
Making good content is table stakes. There's a plethora of resources/examples out there on creating content: Find a sustainable schedule. Be energized and make cool shit! Have fun!

This thread is about the mechanics of how growth actually works for small streamers. (2/29)
Let's start with metrics. I primarily use Twitch's Channel Analytics and TwitchTracker. All of my data is public at https://twitchtracker.com/riotnu  if you want to compare to your channel. I'll talk about the obvious outlier in this graph deeper in the thread. (3/29)
I stream 60 hours a month. The channel has steadily grown month-over-month. Key metrics for November:
Net Followers: +262
Subs: 57
Avg Viewers: 18
Avg Unique Viewers: 159
Max Viewers: 48
Live Views: 3,964
Raid Views: 0.4%
Chat Messages: 19,925

(4/29)
Observation 1: Three types of viewers compose a channel's audience: visitors, regulars, and patrons. Visitors are folks that stop by and check the channel out. Regulars are the ones that keep coming back; they're the core of the community. Patrons offer financial support. (5/29)
To grow, focus on the regulars. Convert one-time visitors into repeat viewers that eventually become regulars. Authentically engage with viewers that participate in chat. You should genuinely care about the people who choose to spend their time watching your channel. (6/29)
Patrons might be regulars; they might also be generous folks just passing thru. Concretely value the support of patrons (ex. shoutouts, emotes), but don't prioritize patrons above or to the detriment of regulars. Aim to grow a bigger community, not monetize a small one. (7/29)
What does this look like in practice? I'm always excited to meet a new visitor. I've gotten to know most of the regulars in chat, and they also know each other. I get to learn lots of cool stuff from people around the world that I would otherwise not get to interact with. (8/29)
We have a community discord with around 200 people talking video games, game dev, and sharing cool creations (obligatory plug: http://discord.riotnu.com ). I read every message. We play community games a couple times a month. I also ingest a lot of VALORANT bug reports. :) (9/29)
Subscribers get cool perks like emotes and sub badges to show appreciation for their support. I don't think it's useful or appropriate at small scale to gate access on sub status (ex. sub games/streams). Small channels don't need these types of filtering mechanisms. (10/29)
Observation 2: Inevitably some regulars will leave the community. Someone might show up for 20 streams in a row and then never come back. This hurts. You probably won't get feedback on why they left. Try not to let it bother you. (11/29)
The streaming relationship is asymmetrical. Viewers know a lot more about the streamer than the streamer does the viewers. Neither party owes the other anything. When a regular decides they're no longer interested, respect and appreciate the time they spent with you. (12/29)
Observation 3: Metrics across weeks and months are useful for tracking progress. Pay attention and learn what works for your channel!

They're less useful for an individual stream. Fewer viewers might just mean some regulars were busy during the session. (13/29)
Observation 4: Twitch's key metrics are Average Viewers and Followers. They're loosely correlated with one another. Net follower change over the last 30 days is a better predictor of average viewers than total followers. (14/29)
A channel's most recent 200 followers are much more likely to be active viewers than the channel's first 200 followers. Twitch Partner requires growing average viewers to 75; total followers only matters indirectly for the goals Twitch sets. (15/29)
Observation 5: Despite Twitch's insistence otherwise, your channel's "Go Live" push notification message is irrelevant. Based on my channel's metrics, there are approximately 8 people in the world who engage with them. Spend your energy elsewhere. (16/29)
Observation 6: I believe Unique Viewers is the most critical indicator to grow into a medium size channel. This is the big piece I'm trying to figure out for my channel. Unique viewers is the size of the funnel of visitors that can potentially convert to regulars. (17/29)
Assume you have the best content on Twitch. You convert 100% of visitors into regulars. Unique Viewers minus regulars is the number of visitors introduced to your channel each stream. If this number is small, your channel's growth will remain slow. (18/29)
Twitch isn't designed to introduce new viewers to small channels. Very few potential viewers are going to scroll by a bunch of larger channels and visit a new small channel. As I write this, I have to skip 225 English VALORANT streams to find a channel with 5 viewers. (19/29)
Most of my streams have ~150 unique viewers. Analytics are incomplete, but here's what they show as traffic sources:
* 0-4 recommendations
* 4-10 search or browse (scroll by ~50 channels at my scale)
* 20-40 Twitter or Discord
* 40-60 followers

(20/29)
Observation 7: Twitch's recommendations don't help unless you get raided by a much larger streamer. I got raided by a party of 3,000 near the end of a stream in October. Most dropped immediately, but the stream held over 1,000 viewers for the rest of the session. (21/29)
The raid itself generated a small amount of growth, but in the raid's wake Twitch started recommending the channel to lots of new viewers. Unique viewers were 3-10x normal for about a week. The subsequent streams generated a bunch of growth for the channel. (22/29)
Then those recommendations stopped and unique viewers settled at a slightly higher number than before. The raid didn't change the channel's weight class. It was a ton of fun, and it sped through a couple weeks' worth of normal growth. Any big streamers wanna raid me? ;) (23/29)
Observation 8: Accelerating Twitch growth requires bringing in new viewers from elsewhere (ex. bringing in an external following on another platform, repeated support from larger streamers). Grinding on Twitch alone without an external push won't generate much growth. (24/29)
Something has to bring new viewers to the channel. Twitch probably could do something to help, but as far as I can tell there's nothing today. Small streamers have to figure out their own path to generate visitors. It sucks, but it's how it works. Deal with it. (25/29)
Much of my channel's community originated from this Twitter account. This wasn't intentional, but I started tweeting around the time the VALORANT came out. Then I started streaming. My Twitter presence has gradually grown and so has my Twitch channel. (26/29)
I suspect Twitter is inefficient for bringing folks to Twitch when compared to more similar platforms like YouTube. I'm starting to experiment but don't really know anything about growing on YouTube. Give me your tips. :D (27/29)
I was going to write about how the money works at this scale (tl;dr: it's small but still appreciated and helps with costs), but this thread is already too long. Let me know if your interested in money details and I'll do another one. (28/29)
Wrap-up plug: If these types of conversations are your jam, or if you like talking game dev/engineering, or you're just really into decidedly mediocre VALORANT gameplay, I'd love for you to check out the channel. I stream evenings T/Th/Sat/Sun at http://twitch.tv/riotnu . (29/29)
You can follow @RiotNu.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled:

By continuing to use the site, you are consenting to the use of cookies as explained in our Cookie Policy to improve your experience.