Twitter First Approach, AI Assisted Content, LinkedIn Growth Strategy (Experiments Dec 2023 Week 2)

All right, everyone. Here's the new format. Let us know what you think about it. We're going to start with a couple of different sections and it starts with today. We are going to start off with experiments we are running right now. Neil and I, we're going to go back and forth and each episode is going to be about 15 minutes or so. And we're going to try to do the best here. So let's start with you, Neil. You got some good stuff that you've been sharing and you've been pushing a lot. So you pick one that you like first from your side.

Yeah. So the first thing that we've been doing, and it's been working out exceptionally well, is we figure out what to post on all our social networks purely based on Twitter. And here's what I mean by that. Whether it's Instagram, TikTok, Facebook, LinkedIn, if you post a lot per day and there's issues with— it doesn't actually— let's actually restart. If you post a lot per day on any social network, other than Twitter from what we see. And even if the content's amazing, the reach just starts getting crammed down. So if I have amazing posts that I post right now, and then an hour later I post another post that's super amazing, it actually can hurt the reach of the first one. On Twitter, on the flip side, if I post a tweet right now and I post a tweet hour later, another one hour later, it doesn't matter, right? The reach doesn't get crammed down. So what we're doing is we're posting a lot on Twitter and we're seeing what's hitting because a lot of it's duds. And then from there, we're then posting the best stuff on Instagram, LinkedIn, YouTube Shorts, TikTok, etc., Facebook. And the model's worked out. So I'm going to share some stats with you guys. And when I look at my Instagram alone, and I'm posting Reels, right? And on TikTok, I'm posting the text version. But once the text version does really well, then I go and record the video version. But when I look at Instagram alone, typically I was getting 25-ish thousand views. So if I look at a lot of my videos before I was doing this, 17,000 views, 22,000 views, 24, 37, 23, 26, 13,000, 13,000, 18,000, 15,000. All right, now let's look at some of my views of recent content that I've been posting based on this model. The latest video has 10,000 views, but it's only been up for 2 hours, so that's not bad. It should hit 30,000+ views. The day before, 37,400 views. Day before, 29,000 views. Day before, 144,000 views. Day before, 23,000 views. Day before, 25,000 views. Then 45,000 views. Then 28,000 views. Then 18,000 views. Then 45,000 views. Then 52,000 views. I stop at the 52 because that's when I started doing this. But as you can see, overall my average is higher. I haven't actually calculated it, but you can just tell looking at the numbers that I'm getting some, you know, that are in the, you know, 30,000, 40,000 range much more consistently, even 50,000, one at 144,000. The model's working out way better. We're noticing the same thing on LinkedIn. We're noticing a little bit better traction on YouTube. It's not getting as good as we want, but we'll work on fixing it. Facebook, not really seeing much of a difference, but Facebook's reach just sucks. Overall, the model's been working really well.

Yeah. And the way this— Alex Hormozi is a good example of a guy that has been doing this for a while where he just uses Twitter almost to get signal first, and then you just choose what to double down on. I'll give you an example here, and we're going to talk about this post, I believe, when we go to a what, Marketing 101 section. But Neil had a post which I didn't think was going to do well on the which one does better, Google versus, uh, Google Bard versus ChatGPT, right? And then that one got what, like 705,000 views or so? So that's a good example of what works. Yes.

And then I'll end up recording that, no joke, today in a few hours. And then within a week or so, we'll get it on all our social platforms. It also gives me ideas for my blog posts. I take this stuff that hits the best on Twitter And then I use it for a blog post. But here's the key. The first line is the hook, okay? That first line there, I end up using that as a title for my blog post, or I'll use it as the hook for my videos or the first slide. But what you can't do, and my team messed this up, I had a conversation with them today. I was like, look dude, we're not doing as well on YouTube Shorts. Who's changing the titles? They're like, oh, we're experimenting. I'm like, no, I didn't say that. I'm like, my instructions were clear. Copy exactly what I said on Twitter and use that. That is your hook, that is your title and nothing else. And, you know, we can dive into something else I'm doing, which is kind of experiment, or you can go next, Eric, on something you're doing. But we're speaking of hooks, we're tweaking the hooks. So a lot of time with Twitter, I'll post something with amazing data, like on backlinking or SEO or a trend I'm seeing on content marketing and how much text you need. To drive more traffic or less traffic. And a lot of times it does really poorly and you can tell in the first like 20, 30 minutes. And in case it's doing really poorly, but we know the data's good and people should look at it and it'd be valuable, we'll delete it and then redo the hook or the intro and try to jazz it up. And we've seen that when we do that, a lot of times we can get the post to hit again. Sometimes it still doesn't work. It's no joke, literally a 50/50 shot from the amount of times we've done it. And it's done quite well for us because a lot has to just do with the hook and how you grab them in, right? A hook could be the first line or it could be the title or the headline, whatever you want to call it. Got it.

All right. So that was about 6 minutes over there on that one. So we got about a good 7, 8 minutes, 9 minutes to go. So we're going to talk about this more specifically business coaches in a future episode, but I I want to talk about this in the context of almost marketing coaches, right? And here's what I mean by that. So for example, I joined this little newsletter group and there's someone in there that is helping coach our current copywriter at the moment. And basically we're paying, I think it's a couple hundred dollars for an hour of time, right? And this is someone that's been there, done that. And she really knows how to grow email lists. The list that she manages, they do I think they do what, $20 million a year or something like that. And she's their director of growth. And I was like, okay, well, she really seemed to know what she was talking about when I joined one of those calls. And I was like, you know what? I'm going to reach out to her to see if she can coach one of my guys. And he actually sent me the stats. So, Mike from my team, here's what he said. So, he said— I was like, give me the results, right? And he's like, sure, here's the full picture. So, before, our email CPA, so when we're buying ads to get more people to our email list, It was about $11, which is pretty expensive for an email. And just a week after working with her, she told him to make some tweaks. And from that 1-hour call, the CPA decreased to $5.31. So that's a 52% decrease, right? Now, the conversion rate went up 21%. So what was it? It was 42%. It went up to 51%, right? And the click-through rate also went up very slightly from 2.14% to 2.2%. And the cost per click went down 32% from $4 to $2.72. So we're not talking about business coaches in this context. We're talking about someone that we're also doing this for, for someone that has led creative at large agencies before, and they're coaching up someone from my team who's building that out right now. So that is a good use of time. And I think the return on that's exponential, and they're down to get paid a couple hundred dollars for their time.

So you and I are going to disagree on this, but I think it's a bad use of time and money. If it was me, I just would've went to them and say, who's similar to you or that you believe could do a good job that I can hire? And if they want a recruiting fee, I would pay it. But I'm a big believer instead of having to train up, if you want to build a big business, you're better off just hiring someone who already knows how to do it and just have them go and execute versus having your guys get trained up. And there's no guarantee that they're going to do as good a job, even if the stats are improving.

Yeah.

It's still not, it may not be as good as that person just doing it or someone that they know is amazing at it.

Yeah. So, I agree if you have resources. I disagree if you don't have resources.

You're totally right. If you don't have the money, you don't have a choice. But if you have the resources or venture funding or anything, you're just better off.

Correct. Yeah. I think it's very nuanced at the end of the day, right? So, I think coaches in those contexts have always worked really well. And then, here's the other thing. To Neil's point, if they're not able to figure it out on your team, they probably shouldn't be on your team, right? So that is like a proxy for trying to vet them out. Because the coach, in some cases, I'll go to the coach sometimes and be like, hey, how are they doing, right? And in some cases in the past, like, hey, I'm not sure they're going to figure it out. And then it's like, okay, we need to have a real conversation here. But that is something that works out well. Let's go to your next one. What do you got?

So next one that we have is, What is it called? Posting videos on LinkedIn and also posting text on LinkedIn. So here's what I mean by that. We've been doing some A/B tests in which we've been posting videos on LinkedIn, text on LinkedIn, and videos in combination with text on my profile, quite a few other profiles. And then we're deleting the ones that aren't doing as well, modifying the ones that are doing mediocre. What we're finding, funny enough, is text-based content is doing the best on LinkedIn. Like the video is the exact same message as a text, like word for word, no difference, right? The versions with text, as I mentioned, is just the same as a video, it's just text format. The ones that are a combination is text first because LinkedIn puts a video at the bottom. And then the video at the bottom is literally me talking or someone else talking depending on whose page it is, verbatim saying the same exact thing word for word as the text. And what we're finding typically is text is performing better than text with video or just video. On my profile page, we decided to still just push video with text. The reason being is we found video with text doing better than just video. And the reason we push that is we feel that video is better for branding and for people to resonate and connect with you on a deeper level than just reading So that's why we choose to do it because our business is a service-based business. So relationships matter. And there's no proven data that I have on my end that adding video with text is going to generate me more revenue.

But so to confirm, Neil, your LinkedIn right now, you're saying, because I saw your post, you do have a lot of video, but you're trying to build a relationship with the video. Yes.

But if you also look above the video, the text word for word is repeating what the video says. Yes.

Yeah, got it. Cool.

Because we found that to do better than just the video, even though we found the text to do better with text and video, we feel the video plus text helps build a deeper relationship with people.

Yep. Makes a lot of sense. I want to share screen real quick before I move to mine. So let's do this real quick. So this is Neil's, those of you that are watching on YouTube and please subscribe to our YouTube. So you can see his post over here on ChatGPT or Bard, 709,000 views on it. But a lot of these other ones, 25K, which is 22K over here, which is great, 37K, 15K or so. And then he'll have some ones that take off, like 700K is a lot, right? But then the other one he posted, which we're going to talk about in a bit as well, is this 179K over here. So there's a lot to glean from here. Now, the next thing I wanted to talk about was this concept of AI-assisted content. So I wanted to I wanted to— here, I'm going to move over to this tab over here. Can you see this?

Yeah.

Yep. So AI-assisted content, this test has been going pretty well for us. Basically what we're doing is we're taking podcast content, for example, and then we're loading it into a Bard or a ChatGPT, for example. We're saying, hey, take the transcript that we have from Riverside and make it into a blog post. Now also what we're doing is because we only have so much content that we're putting out, We're also thinking about, okay, well, there's a lot of news sites out there. Let's say like Search Engine Land, for example, right? They might have an XML sitemap. And so what we can do is we can scan their XML sitemap and then find the newest topics that they're talking about. But then also later on, all the content that I've put out in the past, all videos, everything, right? And it's like, okay. And also what I've talked about recently on the podcast, and it's like, based on how— based on our voice, based on our insights, What is a blog post that we can put out? So maybe the AI can help with 40 to 50% of that, and then you can have a human in the loop. But we're going to double down on this because the graph is starting to compound and we should push it even harder. So yeah.

Have you seen any conversions from this? Because it's a decent amount of traffic, 17,500 clicks of visitors.

Yep.

And this is to the Single Grain website. What have you guys seen to traffic from, I mean, leads to it?

Right.

Because you guys collect revenue.

So I would say the conversions on this one's about anywhere from quarter of a percent to half a percent.

Okay. So you are seeing leads?

Correct.

And then from the leads, I'm assuming it's too early based on the graph.

Yeah.

To revenue because of the sales cycle.

Correct. Too early right now, but we are able to back into it because we can see when the lead fills out the form, their first touch page and also the last Touch where they filled it out and then we're able to track how they're behaving on the site.

And then have you noticed any specific type of AI-related content that you're creating works better than others?

Yeah, you know what I'll say? We got a cancer research center that's publicly traded. They reached out and it was like an AI-related term. I forgot what it was, but we had to refer them to somebody else.

Got it. So when you're using, I get you're using AI to help with this, but when you create the content around AI, You're seeing better leads. Could be—

Correct. And then we're also optimizing these pages at scale too, right? So it's almost like we're doing this and then we're programmatic CRO at the same time because we have these templates to work off of.

Cool.

Yep. So anyway, I think we're at time here. So let us know what you think about this format. You can tweet at both— sorry, post on X to Neil and myself. Let us know what you think about this because we're going to continually iterate on it. We don't think it's perfect. by any means, but we think this is the way for you, for you to maximize more value. And also, please don't forget to rate, review, and subscribe. Helps us grow. And share this with your friends too. That also helps us grow. So we'll see you tomorrow. Thank you.