How to Use Claude To Grow From 0 to 100K Followers (Full Course) transcript

In this video, I'm going to explain how to use Claude to grow faster on any social media platform. And this will cover the entire content workflow from strategy and topics to formats, hooks, scripts, [music] edits, and systems. If you wanted to learn how to use AI to make better content faster that isn't slop, this is going to be a master class for you. By the way, if you don't know me, my name is Callaway. I have millions of followers.

I've done billions of views. And content is all I do all day long. All right. Now, here's the workflow we're going to be running through in this video. I call it the content growth system. First, we start with strategy and positioning. There are five cloud skills that I think anyone making content should run in succession. So, we're going to walk through that whole thing. And then from there, we're going to get into the actual workflow to make a video itself.

You go from topic to research to format to hook, script, record, edit, that whole flow. We'll talk about the AI placement within that flow and every skill you want to run. And then lastly, after you go through that workflow is the learning loop to actually take the winners, learn from them, and then reinforce that back in. We call that the content coach. So, we'll go through that at the very end. Now, to run this whole flow, all we need are two things.

We need the Claude desktop app, and we need the Sandcastles.ai MCP, the social media plug-in installed for Claude. I'm not going to go through the full installation process for how to set this up. I've got videos linked below. It takes like 3 minutes. It's very easy to do, but very simply, the reason we need this is because Sandcastles is the data layer for social media. It allows us to deep analyze any video from our channel, our competitors, map the niche, get the transcripts, and that really is the engine that allows us to unlock these hidden insights for how we should twist the strategy.

MCP is just a fancy word for plugin. It's the bridge that allows you to get the social media data into Claude. If you don't have the social media data in Claude, that's where you get AI slop. Claude is just guessing. It has no context on you. It doesn't know what videos are working. Doesn't know how the competitors are posting. And so it just generates absolute trash. But once you have the social media data piped in from your channel and your competitors and it is deep analyzed to give to the clawed brain, that's where the magic happens.

This combination of Claude and Sandcastles, the data loop is really why I've been able to grow so much faster than most people and why the business owners that I coach directly are crushing in their niche. They're all running this learning loop. So at this point in the video, I'm going to assume you have the Claw desktop app and you've installed the Sandcastles.ai MCP plugin. By the way, for everything we do, we're going to run with Fable 5.

That is the most expensive but the best model in Claude. You can probably run with Opus 5 as well. I just want the best insights we can get. Now, let's start with the content strategy and positioning piece. Like I mentioned, there are five critical Claude skills I recommend every single person run. I will link them below. You can run them in succession. You've got the niche interviewer, you've got the unique positioning builder, you've got the bullseye builder, you have the topic scanner, and the format finder.

Now, for this part of the video, I'm going to pretend like I'm a plastic surgeon. I'm not going to be the social media growth guy. I'm not going to be the tech guy. I'm a plastic surgeon in LA that's trying to use content to drive leads and clients for my practice. Call me Dr. Callaway. And you can tell I'm obviously not a real plastic surgeon. I would have fixed this nose a long time ago. Now, this process is what I actually run on both sides of my content world.

But I want to prove that anybody, even if you don't know what you're doing with content, can run this workflow and get insane insights. So that's what we're going to do, Dr. Callaway, for this. Now, to start this process off, all you have to do is go into Claude, go to new chat, make sure Fable 5 is selected, you paste in this prompt, and I'll explain this in a second, and you drag in these six skills. So, I'll have these linked.

You guys will be able to do this. We'll drag them in. So, we have those six skills in here. So, we have the growth orchestrator, kind of like the growth system, the master skill, and then we have the first five that talk about the strategy and positioning like I mentioned. Now, the prompt very simply is I want to run the content growth system. I am a insert your avatar here. So, I'm a plastic surgeon in LA, Dr. Callaway, that wants to use content to drive leads and clients for my practice.

You have this. I would recommend running this in co-work. You've got Fable 5 selected, and we press go. So, now that is kicking off, and I'm going to step you through each piece of this flow. All right, first up is the niche interviewer skill. This is very critical for everyone to start with. Honestly, if you make content, if you're a marketer, creator, business owner, founder, and you've never been interviewed by Claude to kind of contextualize your positioning for content, I recommend doing this. the skills obviously linked below for free.

This will be five to six questions that'll interview you running through a framework that I know works for positioning and you can kind of answer these questions, help set up the framework for this whole system. So, what this will do when you're done with this super quick 5-minute interview, it'll give you the answer to who is your ideal avatar, what is your target authority statement, what are the pain points of that avatar, what's their dream outcome, and ultimately what is your potential competitive differentiation?

You giving those answers will help this engine a lot, right? So, I'm going to go through what that looks like for me, Dr. Callaway, as the plastic surgeon. So, just at a high level, this is kind of what I wanted to do. We serve upscale clients, typically women between 45 and 55. These are people who have spent money on facial filler or Botox or other smaller procedures, never done a full plastic surgery run. They're very concerned about the procedure not working super well.

They obviously want it to look great, and they're concerned about their aging. They have kind of upper class or high netw worth income, but they're not A-list celebrities. So, just roughly like making that up, that was kind of the positioning that I wanted to represent Dr. Callaway with. Now, I actually used Claude to kind of give me some answers for what these questions would be cuz I wanted to be extremely thorough.

I want you to replicate the level of detail that we have here. So, you can see the very first question is, what is the context behind the business? What do you actually sell? Who pays you? What procedures make up most of your revenue? Right? So, you can see I have that right here. board certified plastic surgeon, solo practice on Bedford Drive in LA. I do about 150 surgical cases per year. Facial only. That's very important.

We don't do boob jobs. We don't do BBLs. We only do faces. Deep plane facelift and neck is the core of it. That's about 60% of the surgeries, etc. Like you can see how how far this is, right? Then we keep going. Question number two is the chair test. So if you could put one core customer, one core viewer in a chair in front of you, what specifically would be the demographic, psychographics, their beliefs? And this is really important in in forming the psychology that we'll use later around topics and hooks and scripts.

So you can see here we go. Our core avatar is 52. She lives in Brentwood. Could be Calabasis. She's married. Husband and entertainment. She spent 80k to 200k on Botox filler different therapies. She's starting to realize it's not working. She wants to stay younger obviously, right? This is a classic kind of like mo for someone who's interested in plastic surgery. So that's question number two. Question number three is her dream outcome.

Question number four would be her pain points. And you can see how detailed this is. This is critical to have this much depth. Again, the more you give, the better the targeting will be. And I'll show you how that ripples downstream. So, you really want to spend like 5 10 minutes answering these. Then the last couple questions are just what is your target authority statement. If you could summarize like me as a plastic surgeon, how do I want to come across in the content one sentence?

It would just be this. I do facial plastic surgery explained honestly for women who've spent years on injectables and are quietly wondering if it's time for something real. As I do this, I'm like, why did I pick plastic surgery as a use case? I have no idea. You can see how crooked my face is. This is going to be fun, though. The last question is, what is your unfair advantage? Me as a plastic surgeon or you as whatever your expertise are.

What is your unfair advantage here? It's one, I only do faces. Two, I have a certain type of practice in my consults, etc. Right? So, obviously, I went through this and I answered these questions, and I recommend you do the same thing in the level of depth that I have here. Now, what this skill will do before it runs anything else, we're still on the first skill, is it will take a stab at building out what your competitive position could look like in the niche.

And this is before it goes out and finds competitors, analyzes videos. This is before it hits the sand castle's data. It's just guessing based on the level of depth that you provide. And so, there are these seven attributes I'm going to explain in a second for how you can carve a unique lane in any niche. So, it kind of does an initial preliminary assessment here. And that's kind of the wrap-up for skill number one, the niche interviewer.

All right. Now, second, we have the unique positioning builder. And this will help you craft what I call your unique positioning angle, which is very important. This is how you craft a lane, especially if you're entering a crowded category where there's tons of people who have been crushing with content and you're just starting out. Now, I've talked about this in a couple other videos, which I will link below if you want a deeper dive, but there are basically seven core attributes that make up differentiation in a positioning strategy, right?

So the first one would be topic selection. So this is where you talk about topics that are not typically covered in the niche. So if you come up with some topics and you become the first one to talk about it, you can carve a lane and have white space in that niche. The second one would be substance depth. So you're talking about common topics, but you do it in a way or a level of depth or with certain granularity and tactics that most people aren't doing.

The third one would be unique stories or proof. So you've worked with certain types of clients or you've just seen more cases or you apply different frameworks that most people can't say they've done without lying. So you have an edge there if that applies to you. The fourth bucket would be avatar specificity. This is very common to carve positioning early. You are serving an underserved slice in the market. So we might say plastic surgery content is all serving women who are 25 to 40 trying to do nose jobs, but it's really leaving the upper end of the market out.

For example, if you started serving older women as an avatar, that could be some place that you could carve a lane. The fifth category would be delivery style. So, this is purely just how you deliver in the content. Kind of the way you do it, but also how you articulate your voice, the rhythm, the tone, things like this. The sixth one would be storytelling formats. This is classic content strategy. You're using formats that other people aren't doing.

If you're the first one to kind of do listicles, breaking down the different types of surgery, but you do it in a cool way, that could be white space that you fill. And then the last one would be the visual format. So you're actually writing on a transparent overlay that's different than the classic like hype edits or different styles people do. That could be a point of differentiation as well. So your goal really in this stage and you don't have to solve it in one day is to figure out where in those buckets of seven is there a white space in the niche that your unique background and expertise can help you fill.

And you can kind of take multiple of those and multiply them together. And that's how you craft your own lane to grow faster. Now, you don't have to have your own lane from day one, but the more you can carve a differentiated niche, the faster you will grow if the content is good. And so, what we're doing with this skill is we're going out and canvasing the core competitors that map closest to the topics you'd cover and the avatar you're serving.

And we try to map where do they sit in those seven areas and how can we carve our own lane. So, I want to show you that right now. This first thing on the screen was just Claude's attempt at doing this without actually doing the competitor research. But now, we're going to do the competitor research. So the first thing it will do is it will build out a competitor list. We call this a watch list on Sandcastles. Now the magic is it's actually able to do this automatically if you have the Sandcastle's MCP installed.

So like without you doing any work, it'll say, "Hey, these eight accounts are the closest plastic surgeons to what we think you would resonate with." And then serving the same avatar as you talking about similar things. You can see these are all facial plastic surgeons or they talk about facial rejuvenation like very closely tied to your avatar. It also recommends tier B people on the list who are plastic surgeons but maybe don't do facial stuff or they're talking about celebrities in a way that wouldn't appeal to driving leads for your business.

So what Claude will do is it will recommend, hey, I think we should add all of the top eight plus number 9, 10, and 12 into your list and we use that list to build the competitive set. So if you don't have any clue what your niche looks like, all you have to do is connect the sand castles. MCP and run this and it will build the list for you. It'll even create the watch list for you in Sandcastles, which I'll show you right now.

So now I'm actually in the Sandcastle's web app. I've got the plastic surgery workspace that I created. And this channel list was automatically populated for me by Claude, which is pretty cool. The MCP can talk back and forth to Sand Castles. Now, I wanted to show this because let's just say Claude pulls an initial list of 8 to 12 accounts, but you actually know there's other plastic surgeons, like maybe there's some of your colleagues in the space, and it didn't pull those for whatever reason.

There's a lot of ways that you can build this list out more directly in Sandcastle. So, if you just come into describe, you can describe plastic surgery and we'll say facial faceelifts and filler we'll say. So, you can type some keywords. We can say we only want them on Instagram and then we want them ideally between micro, small, medium. And we can search and this will pull up all of the possible accounts that we have on Sand Castles that fit this.

Now, Sand Castles has hundreds of thousands of the top creators, but we don't have billions, right, that are on Instagram and Tik Tok. So, I'm going to show you how you can find others in a second, but one way of doing this would be to search for things like this, and then based on who comes up, you can see is there any white space for different people that are missing. Now, this should be pretty good because this is already pulled based on the optimization of all the answers in your niche interviewer.

But just in case there are other people like Dr. Larry Fan, you can add him in here. Now, a second way to do this is let's say chestnut.md is like a perfect fit for us and he's already in the list. All we have to do is hit these double arrows and San Castles will auto resort our entire database closest to chestnut.md. So then he becomes the target instead of random keywords that you write. So this is a great way to find other people as well.

Now let's say you're on Instagram and you're like there are accounts I'm not even seeing come up here. That's totally fine and expected. Like I said we have 100 thousand accounts or a couple hundred thousand. There are millions or billions of accounts on Instagram and Tik Tok. So, all you have to do is go to Instagram and Tik Tok, pull the channel URL, just copy their URL, instagram.com/ whatever their username is, come into ad URL, and you can actually paste it directly in here, and you can add them to the watch list.

If they're not in Sandcastles, we will import them in over 15 minutes, and you'll be good to go. So, what I like to do if you really aren't sure, if you know Chestnut MD is great, you just go to Chestnut MD and you open it in a new tab. And then you could also go to like who he's following. So chances are if he cares about this plastic surgery facial reconstruction niche, he's probably following a lot of people in that niche.

You can actually look at who he's following. These will be sorted by like who I'm following as well. So a lot of these are not plastic surgeons. But if you go a little bit further down, you will see there's going to be a lot of people in this space that are colleagues of his. And so you can actually copy those out. Like let's just say botch TV. I could just press the botch TV. I could copy this URL. I could go to Sandcastles.

I could put this in here. And I could press parse. This then shows look botch we don't have in the network. All you have to do is press plus and it will add it to the list for you. After 15, 20 minutes, it'll be in there. So, the point is there's a lot of different ways that you can build out this list. But all you have to do is get this list filled out with some competitors in your niche and you'll see the rest of the engine works super well.

So, let's shift back to Claude. Now, we're here with the initial list. I think it's pretty magical that you can just answer a couple questions about your niche and then Claude will not only build out the channel list for you, but also build it in Sandcastles and set up that framework for the rest of this. It's pretty amazing that it can do this automatically. And the reason why is because of that sand castles.ai MCP.

So this is kind of the first part of the skill is it just actually building this foundation of the channels. So you can see I took its recommendation. I added a couple others that we had in there and we're good to go. Now the next thing that it does is very cool. It actually builds out a visualization of the positioning attributes. So if we go down and we click on this visualization, it's called the positioning wheel.

So what this did is it took those seven categories that I mentioned and it mapped all of the individual accounts from our list with respect to those categories. And so it's trying to show us where is their whites space, right? So if we go to let's say topic selection, which was the first one, it says topic selection is pretty crowded. This is the full menu of different topics that people talk about. This is what we're currently seeing in the niche with respect to who's covering what topic.

And here is where there's white space, right? So you can go around the wheel and all of these different positioning attributes. You can start understanding where is their whites space, what is crowded, what is not, what might be some ideas. Now you don't have to only follow this, right? This is just an initial swag at this like what does the competitive set look like. You can then start conversing with this if you want.

You can like go off the path of this workflow and just start having a conversation, really asking questions and digging into this data. The only reason it's able to do this is because it went out and deep analyzed a couple of the top performers from every single one of these accounts for you and created this mapping. So, I find this to be an extremely helpful strategy exercise when it comes to what is my counter position.

How am I going to position myself for a new lane in the niche, right? It also gives you like synthesis here for what that is. So, that's skill number two, actually building that competitor list and then building out this positioning wheel to understand, okay, for this person, Dr. Dr. Kevin, we click on it. Dr. Kevin has kind of differentiation in these four sectors. So now we can go open Dr. Kevin's channel, look at a couple of the videos and be like, okay, he does this format.

He covers these topics. Can we be different than Dr. Kevin? Yeah, we could be different here. There's potential white space on this visual format, etc. So this isn't like an immediately tactical thing that you're going to go apply right away, but having it in your head for counter positioning, I find to be very helpful. All right. Now, the third strategy skill you're going to run, I think this is the most important one, is called the bullseye builder.

And what this is doing, as you can see in the visual, is it's building out five levels of breadth with respect to the topics that you cover. A very common question I get asked, especially from business owners that are trying to make content to drive leads is how broad or narrow should I go with my content? What should my mix be? Some coaches are like, I should go super broad and just go really viral and I'll get tons of views and the followers will come in.

But I never see sales when I do that. There's a good reason for that is because when you go too viral, it's actually bringing in the wrong people. However, on the flip side, if you go too narrow, every video gets a couple thousand views, you're really not trickling in sales because it's not big enough to have kind of a spike of credibility and drive enough leads for it to move the needle. So, you're not sure how you should play this.

The bullseye builder is the best exercise to understand exactly what topic altitude you should be playing at. And I want to explain exactly how this works. So the first thing this skill does is it maps out in five levels what the topic breadth would look like. At the very center of the bullseye is the most narrow, right? So this would be like Elizabeth, affluent LA woman, 48 to 58. She has spent six figures on injectables and devices.

She's researching a facelift. She's terrified of it looking too done. That is like bullseye bullseye. Very targeted for what we want. I would argue that's too narrow. If you just made content only for this person, it would convert at an extremely high rate, but you'd only get a few thousand views because it's way too narrow. If we go out one more ring from there, we've got filler fatigued women 45 to 65. This is any woman in this age range who has been on filler, Botox or devices for years and are just curious about any type of plastic surgery.

One ring from there would be any woman considering facial surgery. They haven't done filler. They haven't done Botox. They're just very lightly curious about the space. The blue number four would be any woman 35 plus who is aging face curious. They notice their face aging. They're open to anything, not just surgery, not just filler, not just pharmaceuticals. They're open to hearing about anything. And then number five would be all beauty and skincare.

So we're talking about any makeup brand, any skincare treatment, right? And so you can quickly understand as a plastic surgeon, Dr. Callaway, trying to drive people to my practice, if I made videos about any type of skincare routine, that would be way too broad. It might go super viral, but like people will not come and pay me $100,000 for a facelift because of that. And so I actually really want to stay away from ring number five.

I'll explain what you want to do in a second, but this exercise is extremely helpful and it generates automatically as long as you answer the niche interviewer stuff. That's why this system is so cool. You answer those questions and it kind of like trickles down, right? So, this is what it started doing is it built out an initial swag of what those five levels would be. And then it asked me to gut check like is the fifth actually too broad?

Is the first the correct bullseye? Like you want to dial this in, right? So, we run this and it's correct. Now, I think this next part is actually the coolest piece of this whole flow. So the content mix that I recommend all business owners run is called the 321 content strategy. You have three buckets. Assume you make seven videos in a week. Seven videos in 7 days. Of those seven, you have three buckets of two and then one random one which I call the chaos agent.

Of your three buckets, one of them is more broadly positioned to get more virality and followers but still acrew trust directionally in the right space. Two of those buckets, four of the seven videos are more narrowly positioned. If someone who is the right fit watched that video, they would come down the funnel. And then the last one is a chaos agent meant for random testing. So I want to explain what that means. The broad bucket should be in this light blue, the fourth ring.

So it's broad enough where you could get a couple hundred thousand views. You could drive several thousand followers on one video. You could drive some conversion, but it's really meant to like stepch change your traffic to bring a lot of new eyes onto the page. The narrower buckets are meant to be either orange or yellow. So these are narrow enough like if you made a video about filler fatigued women 45 to 65 something about why filler is fatiguing the face or there's something wrong with filler those people are only one step away from a facelift plastic surgery.

So they are very on target and they will probably convert at a high level. The chaos agent can be anything from red to light blue. You still don't want to go broad dark blue because if you do that and you go viral on accident, you could mess up the page. You'll bring in a lot of people that are not fits and the algorithm will get confused. So the chaos can be anything within those four. Now what's cool is this skill based on all the research it did on your competitors and the niche interview.

It'll start recommending topic buckets that it thinks works well for you. So this first one will be the narrow. So this is in the yellow and orange. So it says we could do stuff about it's quily naming it the filler reckoning, but it's like what years of injectables did to your face like as a bucket of stuff. You could also do why it failed. So, anonymized revision autopsies about how facelifts and nose jobs went wrong.

It also gives a couple others like it recommends those two we should do for our two buckets initially. But it could be how to vet a surgeon. It could be talking about the consult room. You have somebody in a consult, what do they ask? Why are they kind of blocked? Why won't they buy? And then some other ones, right? So, these are the narrow buckets. These would fit yellow and orange. Your job is to pick two of these buckets.

Now, if you don't like any of these, you could say, "Give me more." Or, "I want to brainstorm on this. What about this?" But it's just giving you an initial swag of what might work. Now we go down and there's also some on the broad. So one of these broad in the blue, how faces actually age. So the anatomy of a face, drawing it on a mirror potentially and kind of talking about how facial age works. That's like a broader concept, but it's still directionally applying to somebody who may want plastic surgery on their face if it's the right avatar slice.

Again, the surgical read, doctor verse influencer, there's all these buckets that it proactively comes up with. And this is really really targeted because it has all of the data from your competitive niche piped in. It also gives you a recommendation on these chaos ideas. I find these to be very interesting. So it says the $200,000 face. What if you walked through an anonymized composite of what Elizabeth's 15 years of injectables actually cost item by item and what it brought her, right?

So this is talking about the money angle. That's like a random idea. Or the mirror lecture. You go up to the mirror. You have a picture of her face as a plastic surgeon and you show in long form, no cuts, exactly what you would do and like explain your thoughts. Now, these are obviously random one-off ideas that are within the red to the light blue, but they're just like one-offs. They're not they're not in the buckets.

And so, what I think is really cool is that it actually proactively gives us some suggestions and then we can brainstorm from here. Now, what it does is it asks you to pick 3 2 1 just to kind of fill out the topic work moving forward. If you had to pick three buckets, one broad, two narrow, and then one chaos agent, which of these would you pick? So, it gave me a recommendation and I ended up going with that recommendation just to continue the demo.

And then what it did here, which I think is really cool, is it layered on the buckets within the rings. So, this red bucket is one of our narrow ones. The orange is one of our narrow ones. And then this blue is one of the broader ones. And if you go down, this is my seven days of content right here. I have my red and my orange. I've got my blue broad around how faces actually age. And then I have my chaos agent around the mirror lecture.

Now, down here are all the other ones. So, if I just want to click some of these other ones, I can switch and these will swap out, which is pretty cool. So, I'm going to keep exactly what we had as we go. But, this is like a really helpful batch one plan. Then, it says, "Hey, here's what it might look like after a couple batches of seven, right?" Cuz the whole goal is like, you post seven videos, you see what happened.

If some of them worked, you double down, and so on and so on. So, it's even projecting out, hey, after two to three batches, here's how you might want to rebalance, you'll drop this bucket and replace it with this. So, I find this to be an extremely helpful visual. It also gives you a sourcing map that kind of is a more summarized view of these different levels. So, between those two, I find the bullseye builder to be extremely helpful in articulating what my content mix should look like, how broad or how narrow should I go.

All right. Now, step number four is the topic scanner. And this one is very important. This will give you what you're actually looking for, which is a list of topics that are validated within your strategy that you can go make. The most valuable data you could use are the winning topics from other people in the niche, but then remix based on your own angle or substance. And the reason this is important is because certain niches have certain topic buckets that tend to do better than others.

And it just goes back to the psychology of what the average viewer thinks is important. When it comes to plastic surgery facelifts, maybe they think filler is a very hot button topic that's important to them, but they think techniques for sleep aren't. So, you wouldn't want to talk about sleep if the data wasn't proving that the sleep category worked. You'd want a bias towards filler. Now, to figure this out, we're going to run the topic scanner skill, which will take the top 50 videos across all of our competitors and rank them by performance against the buckets and the avatar we're going for.

Now, what's really cool here is because we've done the niche interview and we've done the channels and we now have the buckets, it's able to go through and scan a ton of videos, but rule out all the ones that aren't relevant. So, like if Dr. Mendez is in his Ferrari going up the coast on a vlog weekend with his wife. That's not relevant for me to drive customers into my practice. So, it just screens that out. It doesn't charge me a credit to run it and it doesn't waste compute to do it.

So, you can see this ran through 400 videos on the watch list from the last 180 days. It resorted them and screened only the top 130. 73 of them qualified. 57 of them were excluded because they were lifestyle vlogs. They were about a different part of the body or different plastic surgery procedures we didn't do, etc. And so what we have left here is the top 50 videos that apply to us. And so we run these and this takes 20 30 minutes to run because it's going to sand castles and deep analyzing.

So just let this run. And what it gives you is a couple really helpful things. The first thing is this topic scanner dock. So not only do you have a performance graph on the topics like rhinoplasty, how faces are aging, etc., but you also have performance on the formats, which we'll get to in a second. Then if you switch to the bullseye placement, this shows you for every topic bucket within the top 50, where does it fall on the rings.

And so you can see some of these fall in the dark blue. We want to stay away from those. Other ones are kind of in the mix in the range that we want. So this is cool just in terms of visual. But what you really are after is this top 50 document. And this top 50 document really is the gold that you are wanting, right? So this is the top 50 videos from your competitors already screened with the information around the topic bucket and then the link to the individual video and the deep link into sand castles and it's all right here for you.

So what you can then do is take this and go one by one and be like is this video applicable to my customer? Do I have something interesting to say? If so, you can then react and start making it in the actual video workflow that I'm going to go through. If not, you just skip it. And so as you scroll down, you can see some of these have been ruled out. Like if we go through this one right here is off niche, but most of the ones that would be rolled out have already been pre-screened.

So this is like a very pure working set of outliers for you and it's organized by view and outlier score. You can even see which topic bucket each one fits into. So if we're like, oh, we want to do the filler reckoning talking about filler, there's a star here on number eight, on number 10, on number 15, 16, like these are the buckets that we green lit. And so again, this is like a proactive ideiation agent working for you.

This doc is extremely helpful. So, that is the fourth step, which is the topic scanner. There's one more skill in this strategy set, and then we're going to go into the actual video workflow. All right, the fifth and final skill in this strategy set is called the format finder. And I find this to be another very helpful visualization that I can actually use to decide which formats to make, where there might be white space, and where it's too crowded.

Now, when I say format, formats are a little overblown as a term. It's just a combination of how you structure the information and how you visualize it. Those together make a format. People are always like, "Oh, I need to find all the hidden formats." Formats are a helpful way to structure your information, but I find it can be a bit overwhelming to try to hunt for like the optimal format. Typically, there's a couple that you like making and you just run that back and forth.

Now, to help you with formats, we actually built the collections tab in Sandcastle. And so, if you go to collections and you hit this little tag and go to formats, all of the core formats that you're going to see are going to be what typically work for most people trying to acrewue expertise. And so, you can click into any of these like A versus B comparison. You can see a bunch of examples in different niches of this and you can click in, you can see the transcript, you can see a bunch of the analysis, you can actually watch the video.

So, we find collections to be like the best go-to resource if you want inspiration examples across different formats. Now, in addition to collections, what's very helpful is to understand the formats that are working for your specific niche. And so, what you're looking at here is the format finder matrix. And this is pretty cool. So Claude is able to take all of those videos that were analyzed from your competitors and plot them across this matrix.

So you've got on the vertical axis, these are storytelling structures. So things like case study, Q&A, problem solution, ways of kind of organizing the information. You would see all of those in the collections and sand castles. And then across the top you have the visual execution, talking head, podcast clip, the surgeon and the patient on the screen together, marking on the face, right? And so what you're seeing is where the outliers actually live and more importantly where the white space is.

So you can see this visual 10 option mirror drawing. It's suggesting that's wide open. No one's done it, but it would be great given the detailed information that you shared in the niche interview. So for example, we said we're really good at going on the mirror and kind of drawing where we would do the adjustments on the facelift. So why don't we turn that into content? That's an amazing visual format that is untapped.

And so this again as a resource for which formats are working, what's not working, really helpful to kind of organize your thoughts for what you might want to try. Now, to be clear, when it comes to formats, typically you're going to pick like two or three every single round of seven, and you're going to experiment. And then most people will get comfortable with one or two that they like making over and over. So, in this first phase, if you're just starting out with content, you want to be able to wildly experiment.

So, I like picking ones that have either worked really well or find the white space where you think you have a unique advantage. If you're really good at drawing, maybe doing the mirror kind of overlay might work well for you where it's too hard for others to do. And so, this is the final step in the strategy flow. Basically, as a summary, what have we done? We have done the niche interview to actually give context on who we are and the avatar we're trying to reach.

We've done the unique positioning builder to understand what the competitive niche looks like and where there might be white space. We've done the bullseye builder to figure out what is our content mix. What's not too broad but not too narrow that would be good at driving customers but also followers for us. Then we did the topic scanner to actually look at the top 400 and ultimately top 50 videos that we deep analyzed that would fit really well within our topic buckets.

And finally we went through the format finder to understand what formats are working best in our niche. If you go through these five steps, you get Claude, you integrate Sand Castle's the MCP, and you run this five skill system, you will have so much more clarity on exactly how to set up what your topic mix should be and how to actually go forward to the next piece, which is making these videos. By the way, guys, if you're liking how I'm breaking this down, all of this is available in Short Form Academy.

That's my program where I work one-on-one with business owners and I really help them rev up, solve their content problems, build a content system, and then they graduate out ripping. This is exactly what I run. There are dozens and dozens of business owners in there that have had huge success. So, if you're looking to work with me one-on-one to kind of install this system and have a much more rigorous approach to winning on social media.

There's a link below if you want to learn more. All right. Next, we're getting into the full tactical topic workflow and I'm going to take my hat off as the plastic surgeon. Dr. Callaway is retiring and I'm going to go back to my social media growth side. And it's just cuz I want to actually show you the process of making a tactical video. And I don't know anything about plastic surgery. So, amazing for the strategy if I'm trying to go from scratch.

But if you're an actual plastic surgeon, you'll be able to give real context on these videos. So, we're shifting back to the social media growth side. I've gone ahead and I've run the topic top 50 exercise for my stuff, my niche, my competitors, my channel, and I've built out that same flow here. And so, you can see I've got that topic mapping here. So, here are my 50 videos. It just did it in a different format, but you can see like all the ones with the checks is the exact same thing.

Here's the bucket they fit into. For me, on my side, I do AI content workflows, hooks, and storytelling. So, it's calling out where the buckets map. It's got all the outliers. It's got the original video link and the analysis link directly to Sand Castles. Everything is the exact same. So, I ran this process to pick it up on my side. And what I ultimately want is the top 50 doc, which is right here. And so, now we're starting the process of how do we actually make a video end to end.

So, once you have this strategy piece, all you have to do is start a new chat in Claude and you're going to paste in the other three skills. And these skills will help you do research on the topic, help you write hooks and scripts on the topic, and ultimately build a finished script for you, which is what you're really after. You want to understand what topics do I make and then how do I take a topic and get a script that actually works as soon as possible.

So that's this part of the workflow that we're going to go through. So what I do is I look at this top 50 list and I literally go down the list and I say do I have something interesting to say for this topic, something different than the outlier and would it be relevant and valuable for the person I'm trying to attract? Right? So I go down number one AI presentation generation build professional presentations in seconds with AI.

That's not what I'm interested in. That's not going to drive people to SFA or Sandcastle. So I'm going to skip number two and number three were off topic. So you can see it strike it out. Number four, [music] Instagram profile optimization. Use these settings. Too transactional. I can't say anything different. I don't like it. Next. Viral video psychology. Use psychological pattern breaks to skyrocket video views. This could work, but he's probably using examples.

Can I find something unique to say? I'm not sure. AI design iteration with Canva. Not a fit. Viral content psychology. Again, using pattern interrupts. I would watch this one and potentially be a fit. This one's too broad. This one's not a fit. Ultimately, I get to number 11, which is video scripting architecture. Deconstruct the ideal scripting structure for short form content. Now, that's something that I have a take on.

What is the ultimate optimal storytelling structure for a script in short form? So, I see that and I watch the original video from him and I say, "Okay, do I have something different to say other than what he said? Do I have something better to say? Is what I have to say more interesting, more compelling, more succinct? Do I have some edge? If yes, I'm going to make that video." And so, that's what I'm going to do in this workflow.

Now, you might say, "Wa, you just skipped through the top 10." Well, the reality is I only need one to make today and I need seven to make per week. So, if this is a top 50, I need seven from here and I can run this next week with all fresh videos and go again. So, you're allowed to pass on a bunch of stuff that's not a fit. You don't have to pick the highest ranked stuff on the top. So, I'm going to go to this chat and you can see this is what I did.

All right, let's take number 11 from Real Sky Tan. This is what I believe to be true. And so, you just brain dump. I recommend using Whisper Flow or you can use this microphone and just brain dump the context that you think to be best about this idea. What do you know that most people don't know? What do you believe to be true that most people don't believe to be true? You're going to brain dump that. Now, from here, we've actually built in a research flow.

So, if you want it to do additional research for you to add on to your take, it'll do that. And I'll explain how it does that. If you just want your context, you don't want any research, you just say no to the research. Right now, I went ahead and ran additional research here. And the reason why is because this research process is actually looking at other social media videos. It's not just a random 360 clawed browser research.

That's not what it's doing. What it's doing is it's looking at all of my previous winners that have been deep analyzed. All of this top 50 that has been previously deep analyzed. All of the videos I've deep analyzed ever in sand castles. And then it's searching the global database of sand castles. And it's trying to find other videos that have talked about this thing perfect scripting structure. It takes all those videos and feeds into a research report for you.

Now, you can choose to incorporate that stuff or not, but being able to run research in this way, essentially a super scan across all of social media is very, very helpful. So, I went ahead and ran that and I'll show you what it actually does. It ends up producing a document that looks like this. So, this is the substance sheet. So this is like what are the core claims that I said initially in the poll and then it goes down and it takes all of the research that it took from other places and combines that in.

Right? So what's cool about this is what people ultimately want to do is they want to take a topic, they want to brain dump on that topic and they want the script to end up magically there. So I'm just showing you the in between step of the research. What we actually do is give you this output. So this solves three things for you. It solves the spoken hooks, it solves the text hooks, and it solves the scripts. Now, before I go through this, I just want to call out oneshotting writing with Claude is the final frontier.

I do not believe Claude is able to write as good as a human today. So, the goal really with this is getting 80 to 90% of the way there and then you manually tweaking it. You tweak it, you feed in the tweaked version, it hones itself, you do that 20 or 30 times, and then it starts to get close to the level that you want. On day one, this will not be able to oneshot scripts that are better than what you could write. It just can't.

And so you have to understand that that's the case. Now what we've done here, which I'll show you, is we've built a datadriven engine for spoken hooks, text hooks, and scripts. And I'll explain what that looks like in a second. Let me just show you the output that you actually get from this topic. So the first thing it does is it gives you three hook options. These are the text hook options right here, kind of rendered on the phone.

These are the spoken hook options, these three here. And then these are the two script options. You've got some with your context only, and then you've got some with research. Now again, you may look at these and be like, "This script is not what I would say." You have to dial this in over time. This is the first generation from this machine. But the point is, if you have a machine that you can train to get better, it will get better.

And this is a datadriven machine off your own videos and the competitor videos. So what you see here is you've got the text hooks, two options right here. So the script architecture that fires dopamine twice, the 2x two script, two dopamine hits per video, how to script the real viewers can't scroll away from script structure that hypnotizes the scroll. Some of these are good, some of these are bad. I think the point is the power words are pretty compelling.

So, hypnotizes the scroll is something I might actually want to use. Your hook is fine. Your lockin zone is missing. So, like the frame lockin zone, it pulled that out. That's something I might want to use. And then we go to the actual script itself. I'll turn on the research. This first one, hook, lockin zone, block one, rehook. This is modeling after the outlier. I don't actually like this frame. Here it goes. The reason people watch your hook and still scroll is because you have no lock in zone.

That's actually not bad from a hook perspective to create curiosity and kind of like create a contrast against what people believe. So as you go down, this is not bad. And the important part is that it will get way better the more videos you analyze and the more you train this. So my goal is to build you a framework that you can feed in better scripts to and iteratively improve it over time. It's just never going to be able to oneshot the way you want.

That's just not where AI is. It's important to accept that. Now, the last thing I wanted to show you is the actual underlying engines that are driving these recommendations, right? So, this is just one output for the topic that we decided to choose. I brain dumped a little bit, but not even that much context. And this is actually pretty good. This is a good starting point that I could iterate off of. Now, if you want to see the engine, we'll scroll down to the bottom, and these are the engines that it built.

So, the first thing we'll go through is the spoken hooks engine. So, I think the the HTML visual version is better. So, what it ended up doing is it took all the videos that we've deep analyzed. It took our videos and it said, "What are the best performing spoken hooks by type, but also by psychology?" And it built this power word or power phrase cloud, which I think is pretty cool. These are all the words that seem to be popping the most in hooks that work the best.

Getting addicted to your content, hypnotizing, cracked out brains, physically cannot scroll away, the opposite of slop. Like some of these obviously insane update, you're not going to use that. But what you want is just to directionally understand the words, the psychology that's actually driving people to hook. And so this is the underlying engine. Now you can set this up recurring so that it automatically adds new videos in.

Every time that you write and iterate a video, it learns. This is the framework and foundation you'll start from. We did the same thing on text hooks. So you have the text hook formats and then you also have the text hook frames. And then lastly on scripts. So these are script templates. Now again, I'll say this for the third time. This is not exactly where I'd want it to be on day one. And that's because there isn't enough reinforcement learning on the writing.

I want this to be tighter. I want this to be better. and faster and me to be able to write something that sounds exactly like me. The only way to do that is to give it versions of your writing that sound like you. So, if you want to run this workflow where you're like, eh, I don't even want to bother with the AI scripting. I just want to do the strategy. I want to get the topics, the top 50. I want to get that format fit.

I want to be able to ramble on my topics. It give me a deep research report across all social media. And then I'm going to write my hooks and my scripts on my own. That's totally fine. I'll show you how I actually write my hooks and scripts now. And I don't use this last mile skill. I do everything up to being able to brain dump and then I write it from scratch. So this is what I actually do to write my hooks and scripts.

Right? So I pull up the topic and I get that research report and I start thinking about like what could the interesting power phrases be. So I do two things. One, I go to the videos tab. I pull up my channel which is right here. I then sort by outlier score. I have all of them deep analyzed. I start with the first one that's relevant to the topic bucket that I'm making. So if I'm making an AI workflow, I will click this first one which is about AI workflow and I will read the transcript.

This is how you turn Claude into a social media machine, etc., etc. Now I'm doing two things. One, if I turn on this show sections piece, I can see how the structure, the different segments or sections within the script worked and I can try to mimic that if I want or I can just look at how I'm flowing the facts through. Now, I've done this thousands of times, so like it comes second nature to me. But really, studying individual transcripts of videos you like that are mapped to a similar topic is the fastest way to get better at scripting.

Obviously, I have a blank page. The first thing I need to do is hooks. And so, I'm like, what are the power words that I can use to hook someone in this case? So, I go to the hooks tab. I type in my channel, Callaway Marketing. I sort the hooks by outlier score. These are all my hooks. And I turn off this so I can see the actual words. And I literally just scan down. This is how you turn Claude into a social media machine.

Here's how to plan your next 30 days of content in 30 minutes. Scientifically impossible to skip. You're using the wrong story arc. Addicted to your storytelling. Like, I'm trying to look for these psychologydriven power phrases. And then I'm going to ask myself, well, what actually applies to the topic I'm talking about? So, in this case, it's a script writing structure. So, it's like, here's how to write so good it's basically unskippable.

So, I could kind of adjust this like here's how to build a script so that nobody scrolls by or here's how to write a script so good it is basically impossible to skip or it is basically Impossible to ignore. I'm just like playing with these phrases and I'm writing them down. I'm reading it. I'm like, eh, too corny. Delete. H, too long-winded. Delete. H, it's the sync, but I'm not sure what it means. Delete. That is really the process of script writing.

Nobody's using AI to oneshot these scripts. I wanted to show you the skills to build a hook engine and a script machine so you can improve it over time with your vernacular in your language. It's just not there today. So, this is kind of what I do on the hook front. And then on the script front, like I said, I literally just pull up the individual videos. I look at the transcripts and I just start writing. If you want to, you can copy this.

Just literally press copy, put that on your page. I will never word for word take what someone's saying. But I can look at the rhythm and the flow and the way the sentences are laid out and how the points are sequentially moving and I can say, okay, they started with this block talking about the how. So what would the how be for me? Okay, I'll write the how, write a couple sentences and then move from there. That's how I actually approach script writing.

It is not an exact science today. So, I wanted to show you like I use the Sand Castle's web app from hooks and scripts, but I use the workflow that we just went through for Claude for strategy and to get the topics, to get the formats, and then I'm really liking this brainstorm function where I can open a new chat, brain dump, and then it researches across social media. I actually think that's a huge hack. So, that's why I wanted to show you on the hooks and scripts.

Obviously, it's not super super ideal cuz you kind of have to iterate through it. I want to give you the easy shortcut answer for everything, but that's kind of where we're at on hooks and scripts if we're being honest. on recording and editing. This is really easy. There are no clawed skills or claude workflows for recording and editing. You don't really need them. Recording, all you have to do is model after an existing video that's in the format you want to make.

So, if you're trying to make patient and doctor, we'll find patient and doctor videos that have crushed. Look at how they're shooting it. What are the angles? What is the distance from the camera? How is the lighting? Who is on what side? Where is the text hook relative to their position? Just model off of that recording. What is the cadence? What is the rhythm? So that's how you would do recording. And then editing is similar.

You want to take the model and then try to replicate the model timing, [music] pacing, sequencing, and camera angle as best you can. There are no great oneshot skills for this today. People can talk about how there are reotion and like different auto editing templates. They're all pretty trash. So once you have a script, I recommend recording it as close to the model as you can, editing it as close as you can. Now, if you're a business owner, you're probably not going to be doing the editing.

If you are, just do it in reels or Cap Cut. Very easy. If you're not and you want to hire the guys that I work with, I've got a link below. I'll introduce you and you can get that set up. Most business owners should try to outsource the editing as soon as possible. And so that's the best way to kind of automate this. All right. Now, the last piece in this flow after you start making videos and posting them is actually very important.

And this is the iterative learning loop or the system that I referred to at the beginning. Basically, once you make seven videos and you post them, you want to look at what's working and you want to do more of what's working. Well, it's kind of hard to understand why certain things are working and not if you don't know what you're looking at. And so the best way to do this is to use the Sandcastle's dashboard. So what we have in the dashboard at the top half are your aggregate metrics, which is cool to see across all of your channels.

But the real power is this bottom half. So this is the content strategy audit. If you go to the reports tab, you can actually run this audit on any channel. It doesn't have to just be yours. But if you're running it on your own, this will render in the bottom of the dashboard and you can see across topics, hooks, formats, and script writing tactics what is working best for you in the last however many videos you run it for.

You can go to reports and if you want to edit this, you could say, "Okay, I want to run this every 30 days or seven days, every day. I want to do it my last 30 videos, 50 videos." You can configure this however you want, but once you do, it's going to show up here and give you a sense for what actually is working. So, what I like to do is I look at topics and I say, "All right, what topics are crushing?" Well, this was kind of like a one-off roundup resource library, but I've got like AI workflows, hookw writing, personal brand growth is seeming to pop, and storytelling is seeming to drop.

I thought storytelling was working, but the data is showing otherwise, so maybe I should do more personal brand growth content. This dashboard is very helpful to kind of reinforce the specific things that are working and not working. Now, this is kind of just the initial swag. What's really cool about this is you can actually bring this into Claude and talk to it. And so, the way to do this, you can go to reports. You can go to any of your reports.

Like this would be my Instagram one. I can click into it. And if I just click this button, continue with MCP, and I copy this in, it'll actually pull it into Claude. I want to show you what that looks like now. All right. All right. So, here we have the content coach or the channel coach skill. And really what this is doing is it's pulling in the dashboard information and it's giving me more specifics on exactly what to do.

I've had it visualize some of this information in a different way here. So, I can see my follower growth over time, the topic buckets. A lot of this is the exact same from the render, but it's giving me more information. The big thing here is that I can communicate with this. I can have a conversation with my data and be like, "Hey, I was thinking about doing this other topic bucket next week. Do you think I should based on the data or not?" or here it's saying this is what I would bias for for the next seven videos and it's showing me do this more, do this less.

It's very dynamic and adaptive as if you had me as a content coach coaching you every seven days. So this is what you really want to be running once you have the system up and going. It's kind of three buckets, right? It's the initial strategy that you do once every 30 days. It's the topic workflow that you do every video. And then it's the content coach that you do every 7 days. That is kind of the sequence that I recommend running.

And all of this is kind of the clawed version of it. All right, guys. That is all I've got for this video. I know it was a long one. I tried to make this the master class when it comes to clawed workflows for social media. If you're a business owner, you're a marketer, you're somebody that makes content trying to drive leads to grow a business. Hopefully, you can tell this will be the most valuable channel you could ever watch for social media growth, I'm trying my absolute best to customize the videos for you.

If you want to work with me one-on-one to improve your content strategy, get better at making content, but also get my eyes on every individual video, I'll review them actively and calls for you, that's what SFA is. I've got the program linked below if you want. And all of these skills with a bunch more like it are included in SFA for free. Like I mentioned, this is the exact workflow that I'm actually running on my own channels.

My new marketing channel, Callaway Marketing on Instagram, went from 0 to 75,000 followers in like 4 months. But most importantly, it drove 30,000 email subs and over a 100,000 in attributable revenue. All because I'm running this flow, using the content coach to figure out what to double down on, and then continuing to do that every single batch. That's what you should be doing if you're a business owner, trying to use the data to get better at social media.

As always, I'm trying my absolute best to push the boundaries for what content is out there. I'm trying to make every video feel like a completely free course for you. If you have any questions, please leave a comment. And I have a bunch of free stuff, all the skills in the description. So, please check those out. Let me know what you think. And we will see you guys on the next one.