Pinned Tweet
WE WON A @shortyawards AWARD 🏆
A huge thank you to @LeroyterBraak , who works behind the scenes on the channel, always goes above and beyond including doing the speech, I hate public speaking so he really saved me here 🤣🙏
A thank you of course to my wife Ella, for putting up with all the madness I bring into the house (she also invites the madness in btw 😂)
And saving the best to last, a massive thank you to all of you guys! We couldn’t haven’t done any of this without you, everyone who watches, supports, engages… we can’t thank you enough 🙌
Oh and lastly, ADOPT DON’T SHOP. This is for all the rescue dogs 💙🐶
4
14
2,101
WE WON!!! 🏆🐶
The Golden Kobe Family just won Best Use of YouTube at the UK Social Media Awards 2026, up against Autotrader, Transport for London, Specsavers, Virgin Media O2 and P&O Cruises 🤯
When I first started filming rescues with Kobe and Moki I never imagined we’d be winning awards, I just wanted to help as many dogs as I could.
Massive thank you to @LeroyterBraak, honestly we couldn’t have done it without him. And of course Ella, who’s in this with me every single day ❤️
And a huge thank you to all of you who watch, comment, donate and adopt. This one’s for all the rescue dogs 🐾 Adopt don’t shop!
Thank you UK Social Media Awards and the judges 🙏
JJ Knight retweeted
If you truly want to become a big player in the YouTube industry:
Never take credit for shit you shouldn't.
Let me give you a very recent example of this:
We are in the process of finding another editor for Golden Kobe Family.
One editor didn't just take credit for a whole edit; he also went through the pain of recording himself reacting to the video he was most proud of and explaining his thought process.
JJ reached out to the creator. The creator had never heard of him. Meanwhile, the editor boasts about work for some big creators in their X bio and on their website. Insanity.
I am too busy being an actual strategist for my creators and being a father to my kids. If I wasn't, I would call out this type of behavior a lot more.
And it's not just editors or thumbnail designers.
Even strategists who talk big game and carefully craft a digital persona that makes creators trust them have simultaneously been in my DMs asking how to improve their business or why their A/B test ended early for a client.
Sometimes I'm confused about the state of this industry.
It feels almost embarrassing to say "I don't know" nowadays.
Whereas I love the editor emailing me saying, "I don't know shit yet, and I don't have a fancy resume of viral videos, but give me a shot, and I will amaze you."
I'd pick that person over the guy taking credit for billions of views he never earned.
People pay me to be right more than the average person. That's it. Strategists aren't magicians who flick a wand and make every creator go viral.
Yet sometimes, people in this industry think that is the case based on what I see these people put out here.
Let me tell you this: lying or taking credit for shit you shouldn't will bite you in the ass. The most arrogant and boastful people on your timeline are the ones working hardest to keep the house of cards from falling.
Question everything and everyone in this industry, including myself.
JJ Knight retweeted
🚨 Looking for a YouTube long form video editor!
Opportunity to work with the fastest growing dog rescue channel on YouTube (The Golden Kobe Family) and edit videos that actually make a difference 🐾
💰 $1,000 per video + $500 bonus
🎬 Story driven, 45m-1h+ long rescue edits
📅 2-3 a month, ongoing
Who we’re looking for 👇
🚀 Outperformers - we don’t coast here
📈 Always levelling up their craft without being asked
💪 Happy to go the extra mile to make something extraordinary
⏰ 3+ years editing long form
🐶 Love dogs of course
Apply link in the thread below 👇
*DO NOT DM OR EMAIL*
🚨 Looking for a YouTube long form video editor!
Opportunity to work with the fastest growing dog rescue channel on YouTube (The Golden Kobe Family) and edit videos that actually make a difference 🐾
💰 $1,000 per video + $500 bonus
🎬 Story driven, 45m-1h+ long rescue edits
📅 2-3 a month, ongoing
Who we’re looking for 👇
🚀 Outperformers - we don’t coast here
📈 Always levelling up their craft without being asked
💪 Happy to go the extra mile to make something extraordinary
⏰ 3+ years editing long form
🐶 Love dogs of course
Apply link in the thread below 👇
*DO NOT DM OR EMAIL*
Apply here 👉 forms.gle/Tq1d1px2oUkEDVbj8
2
9
1,146
Tried inkome.ai on my channel. It "analysed" it in about a minute and suggested... an ebook.
To make this even more fun, I got inkome.ai to come up with ideas for Fern, one of the biggest and fastest growing documentary channels on YouTube. What did it suggest? An ebook on how to compete with Fern...
If it had actually analysed the channel, it would know Fern recently released a video about AI slop channels copying their format. And inkome's answer is to release an AI slop ebook on how to compete with them.
Not only that, their website says "inkome scans your videos & comments to find product ideas your audience actually wants."
Firstly, a real analysis of your channel isn't a 60 second job. I’ve been lucky enough to get access to AI platforms that aren’t available to the public yet, and when I had just 20 of my videos properly analysed, it took 3 days, and that wasn't even the whole channel. You can see from the Fern result that whatever inkome is doing, it’s only understanding your channel at the absolute baseline.
Secondly, even a proper analysis of your comments won't necessarily give you the right product. Before MrBeast launched Feastables, do you think his comment section was asking him for a chocolate bar? Great products come from understanding your brand and your audience, not from summarising what people have typed.
I also ran inkome.ai on my own channel, The Golden Kobe Family, where we take in the dogs nobody else will because they're usually so shut down. It’s suggestion? An ebook on how to rehabilitate these shut down dogs.
I'm not a qualified behaviourist. Selling that would be incredibly unethical, could cause far more harm than good, and would destroy all the trust I've spent years building with my audience. Imagine selling an AI slop ebook that anyone could generate using AI to millions of viewers. It's a recipe for disaster.
To do this properly, you’d be spending hundreds to thousands of dollars a month just for AI to genuinely analyse your channel. Even then, there’s so much nuance and audience research involved that it needs to be done really carefully.
I don't want to create drama on here, and I don't enjoy talking down a new business. I'm all for innovation and AI tools that are actually useful to creators. But I had to speak up, because inkome is getting so much positive noise in the YouTube space and I don’t understand why. Used as intended, it's a quick way to destroy trust with your audience.
So who is this actually for? AI slop channels… the ones in it for a quick buck who don't care about their audience.
I hope the founder sees this and takes it as constructive criticism. But as it stands, it's unusable if you have a real YouTube channel and are trying to run an actual business.
Introducing inkome ⚡️
Launch products your viewers actually want. In minutes.
We find the ideas, create your product, build your store & help you scale.
Start Selling Today 👉 inkome.ai
Fern’s competitive advantage is the research behind their videos, their ideas and their storytelling.
So I got inkome to actually create Fern’s ebook. It produced a 10 chapter book on how they built a multi million dollar brand? All without asking a single question.
Fern puts serious research into every video. inkome wrote a book about their process in minutes, without knowing what that process is.
That’s not a product, it’s generic advice with Fern’s name on it. Sorry, but this just isn’t it.
6
802
Not to toot my own horn but this is how you take inspiration from an outlier, you don’t simply just copy… you take the idea, you put your own twist on it, you up the stakes, or you can change a variable. It’s still early days for our video, and this was actually our wildcard thumbnail, which is currently ranked #2 on our A/B test
2
256
I really struggle to be on this platform now because it’s so hard to read tweets, threads, replies, ALL written by AI. I can spot it almost instantly now and I will stop reading as soon as I notice it. For all of you out there actually writing your own shit, and only using AI to improve on what you’ve written… thank you
P.s I actually used my brain to write this, probably why it’s so grammatically poor but that’s what makes it human 😵💫
2
217
JJ Knight retweeted
YouTubers and YouTube employees.
You might want to read this article before it is too late.
(ps: Accepting first applicants for strategist school before the end of the week. I wanted to finish this article first)
How to win 100% of your YouTube A/B tests (I've cracked it, read to the end)
Quick context: YouTube lets you run A/B testing on a video, so you upload 2 or 3 titles or thumbnails and it splits your audience between them and tells you which one got the most watch time. I run them on both my channels, the main one (The Golden Kobe Family) and the second, more vloggy one.
Every time a test finishes I screenshot the results, paste it into Claude and type "main" or "second" so it knows which channel. That's the whole habit.
Claude reads the screenshot, works out whether it was a title test, thumbnail test or both, describes each variant (rescue dog on its own vs dog with Kobe, scared vs calm, text vs no text, what the title's actually promising), logs the result into a Notion database and tags it so it's searchable.
26 tests in so far. It also sits on top of a Notion knowledge base built from all my strategy calls with Leroy, so when I ask it something about the channel it's answering from what we've actually said and done, not generic YouTube advice.
Then I asked it what the data says. And the honest answer was: not much, and here’s why…
A watch time split looks like a clean number. It isn't. It's packaging plus a load of stuff you can't see in a screenshot:
The video itself. Great packaging on a boring video still loses.
Audience phase. If YouTube decides a video shows good satisfaction signals, it tests it to a wider audience, and those people click differently to our core viewers. A test running during that phase is measuring a different crowd.
Demographics. Whoever clicks early can shape who gets shown it next. If 18-35s bite, the video could get pushed to more 18-35s. Some thumbnails accidentally skew young and "win" by pulling an audience we don't actually want.
Fatigue. "Golden Retriever meets…" did numbers for two years. Now it's tired. A 2025 result tells you nothing about a 2026 upload.
Survivorship. When that format lost a test we reran without it straight away, so the log barely records the losses. On paper it looks like it never loses. It did.
That's why the patterns keep falling apart. You're not looking at a clean experiment, you're looking at packaging plus content plus algorithm plus timing plus who happened to click first. And these are only some of the reasons, there’s a whole lot more…
Except one thing. A visibly scared dog in the thumbnail wins, by a lot, every time it's been tested properly.
Which makes sense, and is a bit annoying, because we already knew that.
So is A/B testing pointless? No. But it's not a rulebook, because what wins on one video can flop on the next. It's for testing different concepts against each other, and it only really works if you're clear on which audience you're trying to attract in the first place.
If the video starts getting pulled to a wider audience, rerun the test, because that new crowd might prefer a completely different concept to your core viewers.
And write a hypothesis before publishing. Which packaging concept do you think wins, what demographic will it pull, and why. If you don't, all you get is a number. If you do, every test teaches you something about who your audience actually is, whether you were right or not.
Which brings me to the bit I actually want people to take from this. I see a lot of creators using AI to try and help grow their YouTube channel. "Here's my thumbnail, why isn't it working." "Analyse my last 10 videos and tell me what they didn’t go viral.”
For that to actually work it would need the context of everything. Not some of it. All of your channel data, demographics included. When the video went up. What was trending in the niche that week. What was happening on your channel that week. Did a large competitor post the same day as you? What your previous video was. Whether this one was an escalation of a video that did well, or a twist on one that did well, or a twist on someone else's video that did well. Then on top of all that (still not everything it needs to know btw) it would need the years of subconscious packaging judgement that you can't write down.
And here's the thing, even if you gave it all of that, it would still send you down the wrong road sometimes. And you would not know, unless you already knew enough to challenge it.
That's exactly what happened today. It confidently found patterns in my tests. They were wrong. I only knew that because I've lived this channel for years.
So asking Claude or ChatGPT to "analyse this and tell me why it's not working" is properly dangerous. It will give you a confident, well-written answer. It might be right. But without the context, it can also send you the complete wrong way and sound just as sure doing it.
One caveat, I've met exactly one person who's genuinely got AI helping ideate and analyze viral videos. He told me it took him seven years to build, it runs on a whole stack of different agents, and the only reason it works is he already knew exactly what he was doing and basically trained the thing on his own brain. So is that even AI doing it? I guess. But the database is him. That's the bit people skip.
Anyway, Claude wrote this. It also spent the afternoon confidently spotting patterns in my data, then I explained five reasons the patterns were shit and it went "yeah, fair" and rewrote its own conclusions (which had somewhat useful insight, but not much).
Also, I could've packaged this whole tweet as "I built an insane AI workflow that analyses every A/B test and here's how to win packaging on YouTube" and it would've done numbers. Instead you've read all of this to find out the process was basically pointless. Apologies.
2
8
952
YouTube RPM will now drop for everything right? RPM is driven by views, and views just got bigger with YouTube’s new update. Actual earnings won’t change, but I guess we now need to re consider what a high RPM is. Correct me if I’m wrong, but I don’t believe YouTube is planning on changing RPM to engaged views
1
3
578
This has to be one of the most insane comments to view ratio, we did a video explaining why we’re leaving the UK to move to the US. The comment section was WILD
2
5
437
JJ Knight retweeted
I am picking 10 of you and teaching you to do my job as a YouTube strategist.
I am ready to help great YouTube strategy talent do better business for themselves and creators in this space.
Apply now. Invites go out mid-September.
youtubestrategistschool.com/
YouTube updating their YPP is interesting… on one hand, if I was starting out, this update wouldn’t affect me. I originally started YouTube to rescue dogs and turn this into a full time business, mostly to be able to rescue more dogs but also to be free myself. And the way I’d see it, is hitting 8k watch hours in a month isn’t even close to the minimum I’d need to hit to earn a stable income (and you have 365 days to do it). So if you’re on YouTube to start a business, focusing on ad revenue at the start, needing to get 8k watch hours to get monetized shouldn’t scare you, it’s way below the minimum you’d need to have a viable business anyway.
But on the other hand, I know a lot of people that are really good at what they do on YouTube, and they know how to create high quality channels and monetize them. These people will still be able to get monetized, however, they could lose out on revenue if they hit a 1m+ viewed video before getting monetized. But, I would say, these people usually look at the long run, and yes you may lose out on $5k+, but ultimately, in 3 years time, with a growing channel earning tens or hundreds of thousands in ad revenue, does it really matter? Probably not
I’m looking at this through a very biased lens, as I’m currently monetized. But, when I was first trying to get monetized, hitting 4k watch hours never bothered me. Did it take me longer than I originally thought to get monetized? Yes. But once I was monetized, it took me a few months to figure out how to really earn before I actually earned good money. And I always thought to myself, hitting 4k watch hours in a month is worth nothing ad revenue wise, hitting 8k watch hours in a month is also worth nothing ad revenue wise. Think longer term, and what you want out of uploading videos onto YouTube.
Oh and when it comes to shorts, if your view floor isn’t 40m views a month, let’s be honest, you don’t have a sustainable business model anyway
JJ Knight retweeted
🤩🥳WE WON A SHORTY AWARD! 🤩🥳
When we started working together about 1.5-2yrs ago, Golden Kobe Family was a small, ambitious team of two that wanted to conquer YouTube through great dog rescue content.
Today, we are building a multi-platform brand, and millions of people decide to watch our content every single month.
So proud of our little team and how fast we’ve grown! Hours upon hours of calls, two times a week, to get where we are today.
I’m pretty sure I’m the first freelance YouTube strategist officially credited on one of these prestigious awards, that’s crazy! :)
But I’m mostly proud of the hard work JJ and Ella have put in.
All my direction, tips, coaching, being a friend, whatever, would be for nothing if they don’t act on it. And boy do they.
I will never stop being a YouTube strategist in some capacity, but I’m very excited to spend even more time on this brand and bring it into every living room in different ways starting 2027.
WHOS DOING THIS?! 🥰
I feel sorry for this sweet old man that’s been paid to read this script, he has no idea he’s the face of an AI slop channel wrongfully trying to get their monetization back…
It’s creators like these that mess it up for the rest of us, putting out false accusations and false statements. It means more rigorous systems will likely have to be put into place even when you have a legitimate appeal, because of people like this.
No sympathy whatsoever, and shame on you for profiting on the misfortune of animals all for self gain
This Post is from an account that no longer exists. Learn more
4
811


















