How I got to 1,000 subscribers without having any clue what I'm doing

On 11th August 2022 I got the magic email saying that I, Timberwolf, was now a YouTube partner. Meaning, of course, that some time slightly earlier I had passed both the 4,000 public watch hours and 1,000 subscribers milestones.

Some people set out to achieve this goal with great care and attention, watching every bit of creator advice they can find, keeping up with the latest rumours about what the algorithm is doing this week and sifting through their analytics to figure out which of the things they're doing are the ingredients for a successful formula.

I, on the other hand, had largely given up on any metrics and was just making videos on things that interested me, for a small group of people I'm mostly in touch with daily via Discord.

Cheating? The Last 100

Despite this lack of attention to metrics my channel was slowly crawling to 1,000 subscribers, adding maybe 4-5 viewers a week while only losing 1-2, with the rate of the former slowly increasing and the rate of the latter slowly decreasing. I was at 900 or so when I released what was supposed to be a stopgap video while I spent a lot of time failing to reverse-engineer Test Drive 3; Remastered and Definitive Editions: The Sierra And LucasArts way. This kind of did what I expected for the first day or two, racking up 100-200 views on a gradually diminishing curve that'd usually top out around 300-400.

Then it exploded.

From the second day it went on this mad near-vertical spike which eventually topped out, at the time of writing, around 10k views. More importantly, a small but critical proportion of those people were watching other videos and subscribing to the channel. It felt like cheating; I was supposed to be on a long, gradual haul that'd eventually tick up to the magic four-figure subscriptions number, and now I'd just been handed it with one single video.

This was actually quite a miserable experience. The explosion in popularity brought trolls, people who didn't "get it" and a general feeling that the channel had become more work and less fun. I spent most of that boom waiting for the numbers to die down and things to go back to normal.

In retrospect, this was not so much cheating as a combination of positive factors:

Old graphical adventures are a popular topic
I was one of the few people to cover the subject in a balanced way rather than the typical "Sierra sucked, LucasArts could do no wrong"
Knowing this was a "filler" video in terms of preparatory work, I made a deliberate effort to push the editing quality much higher than usual
It came after a run of similarly engaging videos, meaning viewers could see this was not a one-off and there was something worth subscribing for

Other than the first one, which in my case is just random luck from wandering around different topics, you can group these under one big banner: when people were recommended the video and saw the channel it gave a clear impression that effort has gone into this.

Which was not always true of Timberwolf's Stuff.

600-900: Finding A Voice

There are a few themes you'll see turn up consistently in my more recent videos:

Not being afraid to get technical at a low level.
Discussing how things were perceived at the time, not just how they are viewed now.
A lot of subtle, blink-and-you'll-miss-it humour

This is not a cheat sheet, unless you collate it into a single bullet point of being yourself. The technical stuff works for me because it is my day job. I can talk about contemporary perceptions because I was there at the time and those were my perceptions. The subtle, quick humour is because I don't do big "yuk yuk" jokes and the videos where I tried them it's obvious how false they are.

What's perhaps notable is how early these themes start turning up, but how late I actually spot this for myself. As far back as the early Let's Play stuff I have long diversions into the excitement of playing a fully-voiced CD-ROM adventure for the first time or how DOS game effects worked but it took me a long time to realise this was the thing, I could just make videos about it straight up without needing to sneak them into the world via the medium of Let's Play.

From X-COM: The Wilderness Years to about Pizza Tycoon there's this early instalment weirdness which is sort of "Timberwolf trying to do Slopes/AVGN/LGR/etc." rather than "Timberwolf being Timberwolf". It's not bad per se but it is clearly someone still doing YouTube with a deliberate goal of gaining clicks and subscribers by sticking to a well-established format, rather than just amusing themselves and a few other people. If I did X-COM again now it'd probably end up being 30 minutes of me messing around in level editors and going into painstaking detail about how the building collapse physics in Apocalypse work. Kinda tempted to make that video.

There's one video in this early, weird period which really stood out for me, and that's PC Speaker Wizardry. This was the first one where two important things happened:

I tagged someone I actually know to say, "this is a topic I'm aware you find interesting and I have made a good thing about it"
I got the feeling I'd made the kind of video that I actually wanted to watch

And while it was a slow burner at the time, it's done way better than anything else from that period. Me carefully explaining an esoteric topic in achingly dull terms of 8253 timers did nearly 4x better than me doing the "popular" thing of angry game reviews. In retrospect, it's obvious: one is genuine and unique, while the other is a knock-off of something thousands of people have tried before.

The weird thing with the growth from this period is most of it came after I "gave up"; when I started making videos for myself and a tiny corner of the retro gaming community. There's probably a lesson in that.

Although of course there was always the tutorial.

500-600: That Tutorial

Oh yes. Getting Started With OpenTTD 12.1. The single most popular thing I've ever recorded.

I had already accepted hitting a hard limit on how much OpenTTD was going to grow subscriber numbers by the point I recorded the tutorial, but there were a few things which annoyed me into action:

People being confused by duplicate store listings charging money for a free game
Existing tutorials missing the newer UI and game features
The amount of bad faith advice being given to new players that was more about people confirming their own biases than being helpful
Wanting more of a bridge between my older OpenTTD content and newer documentary content

My expectation was one of the bigger OpenTTD YouTubers would record their own version sooner or later and I could bow out of doing OpenTTD content knowing I'd done at least one nice thing for the community, but still had never really hit whatever zeitgeist said community was looking for. Of course, none of this happened and it became The Tutorial for a good 12 months or so. I guess... don't complain about the watch hours?

One strange thing that happened with the tutorial is the more documentaries about other games I put out, the more subscribers seemed to come through the tutorial. There's obviously an overlap between people who are just getting started with OpenTTD, and those who want to watch general videos about old games and how they worked.

Which is weird, as there definitely wasn't that overlap for players who already know how to play the game.

500-500: RCT2 and The End Of Let's Play

If you're going for historical accuracy the end of Let's Play on Timberwolf's Stuff was trying to launch an abortive Timberborn series and seeing it trash all my metrics, but the seeds of that were sown at the end of the OpenTTD era.

I was reluctant to move away from OpenTTD early on, for two reasons:

I didn't feel confident enough to "carry" an audience on my personality alone.
I felt a sense of obligation that 500 people had subscribed to me for a particular thing, I should continue doing that thing.

Let me point out with the benefit of hindsight this is an excellent way to kill your channel by running out of enthusiasm for it. I could feel this lack of enthusiasm building up and knew I needed to do something else. As well as my own lack of enthusiasm I was seeing diminishing returns from the Let's Play format; most new OpenTTD episodes were struggling to break 100 views and I was making only the tinest gains in subscriber numbers. Being nervous about losing all those subscribers I'd built up, I chose the smallest possible changes to what I was putting out:

Change the Let's Play from OpenTTD to OpenRCT2
Change OpenTTD content from Let's Play to investigating game mechanics

OpenRCT2 started out really well. The first Crazy Castle episode jumped up into the 200-300 view range almost immediately on release, I had positive comments from existing viewers, and then... it just went off a cliff and never recovered. RCT2 would sit at about 30-40 views per video, despite things like Gravity Gardens With Cursed Coasters being some of my favourite Let's Plays both to record and to watch. I still can't prevent myself from singing the Place Dinosaurs Song when I play RCT2.

OpenTTD Science, on the other hand, got a ton of views. This was my game mechanics series, and I rapidly discovered something fascinating: almost all of the advice circulating Reddit, from signalling to station design, were misconceptions or in same cases outright wrong. Which was a problem.

One thing I dislike about the modern Internet is the obsession with appearing smarter than everyone else. People who can't be seen to learn something in public, or to allow someone else to have figured something out. OpenTTD Science brought this crowd out in droves. If I did something as innocuous as say, "let's see what happens... well, that's a surprise" I would wake the next morning to a comments section full of, "clearly your surprise at this result is a lack of understanding of game mechanics" or pointless quibbles about signalling or whatever other way someone could put across that they were smarter than your host.

(What I have since learnt is pedantic, fun-sapping comments breed. If you reject/delete the first couple you tend not to see any more once the video's been out a while.)

I might have lived with this if it was growing the channel, but the people turning up to debunk my research were also pointedly not subscribing. Which by this point had started to feel like a constant frustration with OpenTTD content.

50-500: Let's Play OpenTTD

It's time to deal with a room elephant. I always got a sense of resentment from the wider OpenTTD community. Almost like there was this cosy world where all graphics were flat-shaded 1x zoom, all videos were about serious and optimised gameplay, there were well-accepted rules about signalling and then I came in and kicked it all apart with these sprawling 2x zoom vehicle sets, videos where I had fun and injected a load of my own personality, and then capped it off by going, "oh hey, those reverse-facing signals in front of platforms, completely pointless, here's the proof".

A perception which is not entirely true.

I had a lot of support from the community at first, especially the developers, people who were heavily involved in getting the game onto Steam, and people who'd made some of the biggest and most successful graphics sets. Which brings me to a good point:

If you're starting a channel, do it for an audience who support you, even after you start growing.

The retro gaming scene has its own problems, but in my little corner the more my channel has grown the more I've found people share my stuff, support what I'm doing and invite me to collaborations. We know that we can all grow together.

Which is the problem I had before. What I said about resentment isn't entirely true, but is also not entirely false either. There was always this undercurrent of unpleasantness around my experience in OpenTTD, particularly around it being okay to straight-up insult 2x zoom pixel work when even the crudest MSPaint sprites would get gentle, constructive help. You can't grow a channel when your only champions in a gigantic community are a couple of dozen developers and addon authors.

I got to this frustrating point where no matter what I did, what I improved, I wasn't able to grow beyond 500 subscribers doing OpenTTD. Worse, some of those subscribers seemed to be subscribed solely for the purpose of unsubscribing whenever I did anything slightly different. Which is a terrible place to be for a channel; you're not growing doing what you're doing, but as soon as you try to do anything different your existing audience punish you so hard for it that YouTube will refuse to show it to anyone.

(Worth noting: of the people who actually subscribed to me this group was a vocal minority, and I allowed them to have far too much influence. The majority of subscribers from my OpenTTD days are still with me and enjoying the current content, as far as I can tell.)

It didn't help that after over 100 episodes across multiple series it was all a bit dull. I loved some of the clever bits and the discussions about music and whatever but you're still talking about having maybe 5 minutes out of every 20 that could be jammed together and assembled into a rambling, unfocused but sporadically entertaining podcast. The rest is really just me clicking on trains for hours. I never intended for it to be anything more than a series you put on in the background just to have someone talking, but that's a tiny niche. Also an indicator how little confidence I had in those early days: put my videos on in the background and ignore them!

This lack of satisfaction in the content and frustration at lack of growth was gradually getting me down, and then it hit that final straw. Digitiser had always been a massive influence on my Let's Play style with the rambling diversions, bizarre ideas and gentle surrealism. But when I tried to play my small part in publicising the Level 2 Kickstarter, the response was, "nobody cares about this, we just want to see OpenTTD". This was the catalyst for the end; knowing I had an audience who didn't care in the slightest for things I loved. When I got to the end of my West Country playthrough I had a couple of beers, recorded the wonderfully rambling Finale where I got it all off my chest, then published X-COM and Isometric Dreams and kept delisting stuff until YouTube started showing them to people because it had nothing else from me that wasn't unlisted.

In retrospect I should have just started a new channel at this point. I feel a bit that I was trying to have it both ways by doing something completely different, but keeping all the existing subscribers. Less so now, given how many have stuck around.

Leaving Let's Play in this way was a shame because for 100 episodes of 1815 Start I did have a brilliant time. So let's move on from the downer bit and have a positive chapter.

0-50: Learning the basics

The great thing about YouTube for me is that I started knowing literally nothing more than how to capture a game using OBS and how to plug a microphone into a PC. Technically I didn't even plug in a microphone at first if you go back to the first voxel model demos I did, although I quickly realised a headset sounded terrible and picked up more of people revving motorbikes on a nearby road than it did my voice.

Which may be the silliest thing about all of this: everything I've done on YouTube comes from me filling a bit of dead air in a voxel modelling demonstration with some nonsense about imaginary goth bands, and since nobody complained about it I thought I may as well make a Let's Play series where I do more of it.

I should talk a bit about Let's Play, because I think it's got a bad name in some corners of the creator community and that's not entirely fair. Yes,

There is a lot of it.
It doesn't take much effort to film or edit.
A lot of what's out there is done badly.

But unless you're in charge of storage procurement at YouTube, what does it matter? Let's Play is the thing which gave me the confidence that people would listen to my voice, that I could be me in front of a microphone and there was an audience for it. Besides, a good Let's Play does take a fair amount of effort outside of the actual recording coming up with gameplay concepts, things to talk about and so on. Not as much as a carefully researched and edited documentary, but it's still there.

Speaking of microphones, it also taught me a lot about audio:

People listen in really bad environments and you need to master with this in mind. I usually test on a phone speaker in the bathroom with the extractor fan going.
A microphone with a directional pickup pattern helps a lot in a noisy environment, if you get used to how it responds to different placements.
Compressor, limiter and noise gate are vital.
Reverb and clothing rustle are almost impossible to remove.
Breathing noise is possible to remove but takes bloody ages.
Low level static and electrical noise is OK to remove with a spectral denoiser but too much of it leaves artifacts.
You need to apply compressor/limiter/etc. to your voice before it gets mixed with game audio.
Avoid unpleasant noises by staying hydrated and being conscious of mouth movements.

The gear acquisition cycle is tempting but even with a Shure SM58, Focusrite Scarlett and FetHead on hand I can get plenty good enough results using a Røde SmartLav and basic TRRS-to-USB adapter into my phone if I position it carefully and do a bit of post-processing. For video I messed around with a DSLR and other nice cameras I had to hand early on, but with the pain of getting the focal plane correct I find it's easier to just stick a good phone on a tripod and make sure the lighting is good enough it's not too noisy.

Let's Play taught me all this in an environment where it would just be 1 or 2 people saying they had trouble hearing me or they were getting distracted by breathing or mouth noise. I still cringe at some of the audio mistakes on current stuff (Test Drive 3 has really bad clothing rustle because I put the main mic in a bad location and it ended up with unrecoverable reverb problems, resulting in me having to use the backup) but it's usually me noticing it first rather than the audience.

Conclusion: The Town Hall

Is this useful advice? Probably not. In the era where I was actively trying to grow the channel I was making bad choices, in the era where the channel has actually been growing it's been an incidental side effect of making videos I enjoy.

Except: maybe that's the trick.

Every day thousands of people upload videos designed to game the algorithm, capture popular searches and bring growth to their channel. Of those, a couple may succeed. Many more won't. The ones who do will be part of an industry in selling the "secrets". Many of the ones who don't will also, believing that they only failed due to not applying enough of the secrets. Ultimately the result is a lot of joyless, repetitive content. More of a waste of disk space than a creative and fun Let's Play, in my opinion.

You cannot control which videos the algorithm is going to pick up, how many likes or subs you get, or who will show up to comment.

You can control whether you had a good time making a video and produced something enjoyable.

Something I do when a view count or other metric looks a bit disappointing is visualise the room with all of these people in it. 5 people? That's a living room full of people watching your stuff. 15 isn't a bad party. 50 is a decent pub. 250 is a busy town hall. 1,000 is a huge number of people. If all of them turned up at once you'd have a hard time finding a venue large enough. With that in mind, "have a good time" is a much more sensible objective than "win the approval of enough people to fill a decent-sized concert hall".

Be genuine, put effort into it, have fun and don't worry about the numbers is probably unfashionable advice and it's certainly not a shortcut to even partner let alone YouTube stardom, but I'm having a much better time now I've learnt to follow it.