Automate Your Social Media Posts with AI: A Step-by-Step Guide
Learn how to use Make.com and AI tools like Preplexity and OpenAI to automate your social media posts daily, saving time and boosting engagement.
File
This Social Media AI System Creates Unique Content Daily (100 Automated)
Added on 09/30/2024
Speakers
add Add new speaker

Speaker 1: After a long day of work, I'm sure the last thing you want to be doing is figuring out where you're going to post on Facebook and Twitter So wouldn't it be nice if you can train AI to write in your tone of voice and post for you on Automation every single day like clockwork This will help you build your brand as well as engage with your followers and this system is exactly what I'm going to show you How to set up today So the tool that we're going to be using is called make.com Make.com is an RPA or robotic processing automation tool that allows you to string along different types of software together in a flow on automation so that they all operate. Now let me show you how to set this up today. So once you get your Make.com account, you can just go up to the top here and create a new scenario. From here, you can see that it's actually very visual. This whole tool is no code. So that means you can set up this whole AI automation without writing one single line of code, which is incredible. We're gonna go here and press the plus button. The first thing we're gonna set up is a trigger. This will tell Make what is going to trigger, what is going to start this automation. And it's actually quite intuitive. You basically tell it, use this tool first, then use this tool, then use this tool, then use this tool. In a visual way, you can describe how you want this automation to operate. So the first thing that's going to trigger our automation is a Google Sheet. What we're going to do is put in the URL of a recent article about our industry or our niche in there, and then AI is going to go into work and then write all of our social media posts. I'm going to press the plus button here, and then I'm going to search for Google Sheets. And then after I select Google Sheets, I'm going to select Watch New Role. So this will trigger whenever I put in a new role, a new news article into my spreadsheet. I have already made my spreadsheet right over here, and as you can see, it looks like this. It just has a news link. And then I put one article in here for testing purposes. And this is a new article that just came out today about Elon Musk. And Elon Musk says that AI will take away all of our jobs. Great. So I have this article in here. And then I'm ready to turn this article into four social media posts. Back in Make, I'm going to select this spreadsheet. I'm going to click the blue button here. And then find my spreadsheet in my Google Drives folder. Then I have to select the sheet to use, which this spreadsheet only has one sheet. It's just sheet one. So I'm all good to go there. And lastly, in the limit, I'm gonna select one because I want it to write one article at a time. Now I'm gonna press OK. And then down at the bottom here, we're gonna change the scheduling setting. By default, it's going to run every 15 minutes, but I don't want a new social media post to be posted every 15 minutes to all of my accounts. So I'm going to change this interval to every day because I just want one post every day. And then at 8 AM is fine. And I'm going to press OK. All right, so just like that, we have our trigger set up. OK, now that we have our trigger set up, the next step is to have AI read that article. And we're going to use a tool called Preplexity to do that. Preplexity is a competitor to OpenAI and Cloud. But what it does better than these two other language models is that it can actually access the web and be able to return information about different articles to you in real time. So as you may know, OpenAI and Cloud actually have a cutoff date. And it only knows information generally from six months ago. So it doesn't have the latest information as of right now. I know that's going to be changing in the next couple of months. But as of right now, Preplexity is the only large language model that can access the web in real time. You can sign up for a free account in Preplexity, and you'll be taken to a dashboard that looks like this. And what you'll need to do is to get your API key and connect that into make.com. So once you're in Preflexity, you wanna go down to the bottom and click the gear icon. From here, you'll be able to see the top menu, which is Accounts, Profile, API, and Enterprise. You're gonna go ahead and click on your API key. This is where you can add in more additional funds. I had in $10 and this lasted me ages. And you're gonna go down here and copy your API key. Once you have copied your API key, you're going to go back to Make, and then press Add an Additional Module. From here, we're going to search for Preplexity, and then we're going to select Create a Chat Completion. If this is your first time accessing and using Preplexity in Make, you're going to press Add right up here, and then simply put in your API key, and then press Save, and this will work. You can leave the model at Llama3, and if you want to change to different models, that's fine as well, but Find Llama 3 works just fine. And then here, you will just put in what you want Prick Plexi to do in the content, and then give it a role. The role will always be user. I just leave the tokens and temperature blank for now. So tokens will determine how long the message will be. If you leave it to default, it's just gonna spread as long as it needs to, and then temperature tells it how creative it can be. I think the default settings for this will be fine. Here in the content, I'll just tell it to summarize this article, and then I'm gonna select the news link one. The reason I'm selecting news link in column A is because what this is going to fill out is the link that is already in there. Because I named the first column in the spreadsheet news link that is what I want PrePlexity to access. I'm gonna press okay here, save my work, really important, make this an auto save. So I've lost a lot of work over the years over that. And then I'm just going to press run once to see if we're getting the results that we want. It's important to continuously test your work so that you can find out the bugs earlier on. All right, so just like that, we have ran this successfully. And then let's just take a look. So from make, you can see the little pop-up bubbles at the top, and you can press the magnifying glass to dive deeper. And then if I go into the input, what we input it in to make. I'm going to expand on messages. So you can see, we told it to summarize this article. See how it auto-filled the value that was under column one? And then it's going to go to the next row the next time we run, and then the next row, and the next row, right? So we can see that is working perfectly. And then in the output, I'm going to expand on choices, and then message, and then content. And we can see that, yes, it did summarize the article here for us. That's perfect. Step one is working. So we have summarized the article. The next thing that we wanted to do is to write a Facebook, Instagram, LinkedIn, and X post for us. And the way that we're going to do this is to use a router. And the router will allow us to perform multiple actions at once. I'm going to create four different routers because we want four different posts to be created here. And then I'm going to use Cloud to write these posts. And the reason that I'm using Cloud is because currently I find that Cloud is just generating a bit better output than OpenAI is. So I'm going to click on Cloud. By the way, you can use OpenAI for this as well. I'll actually show you how to set up OpenAI in the next example so that you can choose which large language model that you want. I'm going to click on Create a Message. And then if this is your first time setting up Cloud, click Add, and then just put in your API key right over here. From your Cloud account, just go into Settings, and you'll be able to get your API key. OK, and under Models, I always select Cloud 3 Opus, which is the highest model that you can use. Under the row, I'm going to select User. And then under the content, I'm going to expand it. I'm going to put Text here. And then this is where I can put in my prompt, okay? I've already pre-written a prompt for this. Act as a social media manager and generate a Facebook post. The post should engage the audience with compelling introduction, provide essential details, and encourage interaction through likes, comments, and shares. And with clear call to action, include optional hashtags. Here's a summary of the article that I want you to write a Facebook post about. And this is where I actually enter in the result or the output from the previous step. So remember in make the previous step was that we went into preplexity and we generated the summary of the article. Now I can go in there and then I expanded on choices, then I expanded on message and I clicked on content. And now this part is going to pre-fill with the result from the previous step, okay? And I'm gonna press okay right here. And so just like that, we have our Facebook post written. Now, the next part is we're going to now set up Instagram. We're going to follow pretty much the same step, but in this case, I'll use OpenAI to write the Instagram post instead of Cloud just to show you how to do it using both methods. Okay, so I'm going to press the plus button here. I'm going to search for OpenAI, going to click on the OpenAI module, and I'm going to go down here and select the second one, which is create a completion. And this is where we can actually use OpenAI off of ChatGPT. And if this is your first time connecting to OpenAI, click add at the top and then put in your API key right over here. If you're trying to find your API key from your OpenAI account, simply go to platform.openai.com. And then once you log in, you can see that the last tab here is API key. Simply click on that and then create a new key, give it a name. I'm gonna call it test for now. Then it's going to give you this key. you simply copy this key right over here, go back to make, press add, and then put in your key right over here and press save. And then this is how you can use OpenAI off of ChatGPT. I'm gonna close that for now because I already have this connected. All right, so from our model, I'm gonna select GPT-4 because I want the most powerful model that I can use at the moment. And then once I select ChatGPT-4, under messages, this is where I can give it my prompt. Now for the role, again, I'm going to select user. And then for the content, I'm going to put in my prompt and I'm also going to attach the summary of the article that we just wrote. So this is my simple prompt here as a digital marketing specialist, create an Instagram post about the following artificial intelligence article, ensure the content is visually appealing and include an inspirational message, use emojis to enhance engagement, suggest relevant hashtags. Here's a summary of the article I want you to write. And then now, once I click here, I can see all of the outputs from the previous steps. We're going to go to preplexity, expand on choices, expand on message, and then click on content because this is where the API is storing all of the previous outputs that we just have generated from the article, right? Okay, so just like that, we finished our Instagram. And at the bottom, we need to put in a number of tokens. So for this, I'm gonna put in, let's say 2,000 maximum. All right, by the way, currently, The maximum you can put in, I think, is 4,096. OK, now that we have our Instagram set up, let's now do the same for X. OK, so now we're just repeating the process for Twitter, where X. So here, I set up the OpenAI module just as we have previously. And I put in my prompt to tell it to write a Twitter post. And again, I'm going to go under Choices, under Message, and then under content to populate the summary of the article I just chosen. And I'm gonna press OK here at the bottom. Lastly, I'm gonna set up my LinkedIn. And for this one, I'm gonna do Claude again. So I'm gonna press the plus button, I'm gonna search for Claude, and then create a message. From here, I'm going to select the model I'm gonna use, which is Claude3Opus, the most powerful model available so far. Give it a number of tokens. For the role, I'm going to select User. For the content type, I'm going to select Text. And then under Text, this is where I'm going to put in my prompt again. All right, I have put in my LinkedIn prompt right here. All right, great, so that's my prompt for LinkedIn. I'm going to press OK. And the last thing before we connect it to the actual platforms is for Instagram, I actually wanted to generate an image as well. So from the Instagram router here, I'm going to press another plus symbol, which allows me to add an additional module here. And then I'm going to use OpenAI to generate an image. You can also use stable diffusion as well. But for this example, I'm going to use OpenAI. And then I'm going to select Generate an Image from this list. This will allow me to generate a image using DALI. And I'm going to use DALI 3. And then this is where I can put in my prompt. All right, I have put in my prompt for this image. I'm gonna press okay. So just like that, we have our whole system set up. Let's go ahead and save it and run it and then see if everything's working properly. If it is, then all we need to do is connect it to all of our social media platforms and then we're good to go. All right, our testing has finished. So let's take a look at the results here. At the top for our Facebook output, I'm just gonna click on the little one and the little bubble that comes out to see what the results are. Under outputs, I'm gonna look at content. one, and then the text. And then here is where I can read what the AI has written for our Facebook post. And if you don't like what it has written, then modify your prompt until it sounds like. You can modify the length, you can tell it to include more emojis, you can tell it to include more hashtags, you can tell it to shorten it, you can give it maximum length. So you can really play around with the prompt until you get it to exactly what you want. And this looks fine to me. Let's take a look at the Instagram one. Under Outputs, I'm going to press on Choices and then look inside the message and the message content. And I can see, whoa, this is a bit too long. So I can modify my prompt to actually tell it to shorten it. Let's take a look at the image that Dali has generated here. Again, under Outputs, I'm going to look at the data and it's giving me this URL. So let's go ahead and take a look at it. This is the image that it generated for this article. Not bad. I probably will prompt it to not have words on it, but generally not bad. What you would do during this step is to fix your prompt until you get the output to be exactly what you need it to be and you can check all of your outputs here. Once you're happy with those prompts, what you would do is to add another module and this is where you would actually start connecting it to your account. So for the first one, I'm gonna connect it to Facebook. I'm going to search for Facebook. You can either connect it to your Facebook page for your particular account. I needed to connect it to my Facebook page for engagement for my business. So I'm gonna press Facebook page and then press create a post. From here I had to create that connection to my account and press save. Facebook is going to pop up the authorization and once I have authorized my Facebook connection, I can select the specific page that I want it to be posted to. Once I have selected the page, Here, I can put in the exact message that I wanted to post, and this would be the output from the previous step. So as we can see all of the results from the previous step now from Entropic or Clog, and then under Content, I'm gonna select the text that it just generated, and then press OK. Great, and just like that, I'll have my Facebook post posting at 8.23 a.m. every single day. Now, I'm gonna repeat the process for Instagram. Under the Search app, I'm going to look for Instagram for business, and then I'm going to create a photo post. I'm going to have to add an account here, connect it to my account. After I connect it to my account, I need to select the photo URL, which was generated from the previous step. So I'm going to expand on the data of the previous step, and then get that URL of the photo that was just created by Dali3. Next, I need to put in the caption, and the caption was generated two steps before, right? And then we're gonna expand on choices under the OpenAI module again, expand on message, and then select the content that we have created for the Instagram post, and then press OK. Great, remember to periodically save your work. Next thing you wanna do is look for Twitter or X, and then we're gonna select create a post again. Again, connect to your account. And lastly, we just want to connect to LinkedIn. Again, same process, press the Add button, and then now look for LinkedIn, select create user text post. Perfect. From here, connect to your LinkedIn account, press save, put in your username and password to authorize this connection, and then you're all good to go. Lastly, you just need to turn on this automation for everything to run. Before we go, if you don't want it to post automatically to start with, if you want to start reviewing the work, you could actually disconnect this by right-clicking and press on link, have the post sent to you on an email or a Google Doc, and you'll review it for the first couple of days or weeks until you get the prompt down, until you feel comfortable having the AI system run on automation. You can do so by adding the plus button, and then from here, if you wanted to send it to you on your email, just click Gmail, and then you can select Send an Email. You can just send this email to yourself, so put in your email address right here, and simply put in the result from Entropic or OpenAI in the content section, then press OK. Just like that, the AI will send you an email with the content of all your posts, so you can review it and then maybe modify the prompt until you really get that prompt down and you feel comfortable doing all this on automation. Lastly, to have your social media AI automation running every day, make sure in your spreadsheet you have enough news links. The way that we have set it up, remember, is that every single day at 8.23 a.m., what this automation will do is that it will be triggered and it will take the next news link URL that's on your spreadsheet, and Preplexity is going to read and summarize that, and then OpenAI and Cloud is going to turn this URL into Facebook, Instagram, as well as Twitter and LinkedIn posts, okay? So just make sure you have enough URL on here, and then your automation will be all set to go. I hope you guys find this as fascinating and as cool as I do. It's absolutely mind-blowing how much you could do with AI and automations. And just like that, within 10, 20 minutes, you have a whole system set up for your social media that will be able to post for you every single day. If you liked and enjoyed this video, please make sure you like and subscribe to the channel below for more content just like this. Also, if you're looking to learn more about how to use AI and automation, sign up for my free course below as well, and I'll see you soon.

ai AI Insights
Summary

Generate a brief summary highlighting the main points of the transcript.

Generate
Title

Generate a concise and relevant title for the transcript based on the main themes and content discussed.

Generate
Keywords

Identify and highlight the key words or phrases most relevant to the content of the transcript.

Generate
Enter your query
Sentiments

Analyze the emotional tone of the transcript to determine whether the sentiment is positive, negative, or neutral.

Generate
Quizzes

Create interactive quizzes based on the content of the transcript to test comprehension or engage users.

Generate
{{ secondsToHumanTime(time) }}
Back
Forward
{{ Math.round(speed * 100) / 100 }}x
{{ secondsToHumanTime(duration) }}
close
New speaker
Add speaker
close
Edit speaker
Save changes
close
Share Transcript