Welcome to this episode of ‘But Wait There’s Moore’ where we dive headfirst into boosting your marketing game with a little AI magic! In today’s video, we unravel the secrets of preparing a bulk upload CSV to schedule your social media captions across LinkedIn, Twitter, and Facebook using Claude AI and Agorapulse. Say goodbye to manual drudgery and hello to automation! We’ve got quirky tips, a touch of humor, and a ton of practical advice for you. Plus, don’t forget to check out more marketing insights on our Substack. Ready to get started? Then let’s jump right in!
00:00:00 Introduction to Claude Projects
00:00:34 Preparing Your CSV File for Bulk Upload
00:01:21 Using Agorapulse for Bulk Uploads
00:02:18 Setting Up Your CSV File in Excel
00:04:23 Interacting with Claude AI for Caption Generation
00:09:12 Reviewing and Correcting AI-Generated Captions
00:13:55 Finalizing and Uploading Your CSV File
00:17:42 Conclusion and Final Thoughts
Chapter 1: Introduction to Claude Projects
[00:00:00] We just did two posts that would go out on LinkedIn. You might have 20 posts, 200 posts that need to not only go out on LinkedIn, what if they also need to go out on Twitter and also need to go out on Facebook? You can use this method, though, where you’re grabbing these links and saying, let me at least give you a list of links, and based on where you grab those from, you may not have to do it where you do it like I did, and go copy them out of a URL bar.
[00:00:33] In this video. I’m going to show you how you can use Claude AI to prepare a bulk upload CSV file to schedule your social media captions in bulk, and it’ll work with your file.
Chapter 2: Preparing Your CSV File for Bulk Upload
[00: 00:46] It’ll actually prep a CSV file for you. It’ll drop the captions in there based on links you give it, and then it will turn out the file for you, so then you can go upload that into whatever your social media scheduler is. For today’s purposes and for this video’s purposes, I’m going to be using Agorapulse, which is the social media scheduler of my preference in choosing.
[00:01:07] Plus, I do have a partnership with them, uh, but this will work on any social media scheduling platform that provides you the option or that you have the option to do bulk upload scheduling with. So do check your social media platform. We are already in Agorapulse, and again, when you have the feature available to you to bulk upload, typically these platforms will provide you with a CSV sample file because if you think about it, it has its own way that it’s going to pull those things in in bulk that’s going to come from a CSV file.
Chapter 3: Using Agorapulse for Bulk Uploads
[00:01:40] But it needs to tell you how to format that because if you were just like, let me just create one willy-nilly and try to upload it. Well, if you’ve ever done database field mapping or anything like that, it’s gotta do that mapping, and it needs to know if we’re providing a link, which cell am I looking at? And you need to be able to provide that. I have downloaded that file. Now I’m gonna go ahead and open that.
[00:02:02] I always think it’s good to preview this because then at least you know what you’re talking about, because in a second we’re going to be having a conversation with Claude AI. And again, this will work with ChatGPT, or Perplexity, or Google Gemini, or anything like that.
Chapter 4: Setting Up Your CSV File in Excel
[00:02:18] So as you can see, I’ve opened this in my Excel document, and I’m gonna just spread those out. Here’s the post content. That’s gonna be the caption, the post link; that’s gonna be okay. Do I want it to include a link to, you know, a website address or something like that? If there’s a video URL, if you have a video or an image ready, and you want it to pull and use that image or video in the actual caption.
[00:02:44] I’m not gonna demonstrate that today. I typically don’t use it because what I’m gonna show you today is: if hey, I’m gonna use a website link that has a typical automatic link preview attached to it, that’s gonna show up in a social feed. Schedule date is here. Please notice, and for those of you who are living in the United States, that date might look weird to you.
[00:03:05] That’s actually pretty typical for European date styles is to be day, month, year, instead of month, day, year. So, what those dates are means it’s on January 29th, 2035. Obviously that’s nine years into the future. Now, uh, then you’ve got the scheduled time.
Please again, notice this is going by a 12-hour format where we distinguish if it’s AM or PM. I’m pointing these formats out to you because this is important. Lastly, Agorapulse has a column for post label.
[00:03:35] So we could label this if we wanted to and say, this is for campaign, this is for new published, or something like that. I’m gonna delete out these bottom here because I like to have a guide for myself for when I’m doing this. But remember, I’m not gonna be doing all this myself. We’re gonna actually walk through this together. I know for a fact that I wanna be able to share out some posts that are about my website. So I’m gonna go open my website actually. Great.
[00:04:01] I’m gonna go to here, and I’m actually just gonna grab that link ’cause I know I would like there to be a social post about this blog post. I’m gonna go back to my Excel sheet and remember where we said post link, and I’m gonna drop that right here. Well, then I don’t have a caption yet, on purpose, not gonna put one there yet. Let’s go ahead and do a scheduled date. Now remember, you don’t necessarily have to do this.
Chapter 5: Interacting with Claude AI for Caption Generation
[00:04:23] These are things you can be telling Claude or your AI chat program of choice to do. I am gonna say how to video like I already did. I’m gonna pull that down. Okay, so now I’ve got my pattern here, and I’m gonna go ahead and take out, I’m gonna delete this row. Let’s grab another one, just so we can demonstrate and see what all is happening here. So, I’m gonna go back, and maybe I would like you to go give people information about the MenopauseIt podcast. I’m gonna put that here.
[00:04:50] Let’s just see what it does, right? Because maybe I don’t wanna just do blogs. Okay. So, I’m gonna do this, and I’m gonna rename this file just ’cause I like to keep things straight. I’m gonna call this 0102-bulk-import-how-to-video. Okay, ’cause that’s what we’re doing right now. I’m gonna keep it as a CSV file. I’m just gonna save it in Downloads right now.
[00:05:12] Obviously with your Finder, and again, I’m doing this on Mac, but this would also work with Windows. You can save it wherever you need to do, as long as you know where it is. Okay, I’m gonna leave that blank, and I’m leaving on purpose, the scheduled date and scheduled time blank in these two columns. Because we’re gonna now switch over to Claude and we’re gonna have it work with us a bit. Now, I have a project show already, and please do check out my other video about how to create a project in Claude, which I believe is for a pro account, which is what I have.
[00:05:40] If you do not have pro, you’re just gonna be doing this via chat. I’m gonna type in prompts, but I’m not gonna make you sit through that. I’ll basically skip through so you can see what prompt I entered. All right, so I’m telling it that’s what we’re gonna be doing. Now, Claude’s gonna be like, “Sweet, I cannot wait.” See, this is what is cracks me up about AI. Just wants to tell you what are we doing here? Now it’s asking some questions. I really recommend that you let it ask you these questions and answer them, because this is what I do enjoy about AI.
[00:06:09] A lot of times, because I, you know, it’s been in use for a while and people are using it for this purpose. It kind of knows, based on that worldwide history, saying, ah, I need a little more information because I know there’s not just one social media app out there right now. You may just have one social media app you’re writing for or posting to, but right now it’s asking, well, I need those specifics. So it’s saying, go ahead and upload the CSV file, which we’ll do in a minute, and let me know which platforms we’re writing for. I’m going to go ahead and upload that file. Add file. Remember, I saved it in my downloads, and I saved it with that date on it. So, I’m gonna hit open. Okay, my file is there.
[00:06:46] All right, but now I’m gonna reply to the rest of its prompt. Okay. Now I’ve just typed in my answers to all these prompts. I saved you that long slog of doing that. But let’s quickly go through. So, I did answer its questions. I said that these are going to be for LinkedIn. And then in question number two, I answered it because it asked what tone and voice, and I’m giving it my tone of voice. This will be whatever you and your brand’s tone of voice needs to be.
[00:07:13] And if you’ve already uploaded that into a project, a Claude project, or in the same chat, you’ve told it what your tone of voice is, you may not need to do this. Okay? You’ve already established what that’s going to be, and it may already know that, but it’s asking me this here. And that’s a good thing to ask. And if I had given it that tone of voice, I could say, “Hey, refer back to this project. Refer back to earlier notes about it.” What’s the goal for these captions? Now this is specific to me. You can do whatever you want. This is a little bit of social media: how to bake it in.
[00:07:43] Remember, we are dealing with shortening, shortening, shortening, shortening attention spans. I know my biggest hurdle is to have somebody actually pause and look at the social post. So, that’s what I told it. I’m like, especially when things get truncated, LinkedIn will do that. Usually after the first full sentence, you get the dot, dot, dot read more. I want to make sure that I mitigate against that by making those, that first statement.
[00:08:08] Something that’s gonna make them screech to a stop in their tracks and want to click more and actually read the post. Then, hopefully, click on my link. So, I told it that. I personally am not using hashtags really anywhere. If you are, please tell it what to do here. But you can also ask it, saying, well, based on what the link is gonna say, if you want to recommend some based on the content of the links in the CSV, please go ahead.
[00:08:38] All right, y’all, my CSV is attached and uploaded. I have the answers to the prompts and I’m gonna go ahead and hit go. Quick note: if you connect Claude or whatever your AI tool is and you connect it to have access to your files, and I do have that set up, you can actually give it the link in your Documents folder on your hard drive, you can, or even, it’ll take it from your Google Drive. If you have that connected through Windows or your Finder, it’ll actually go open the file for you and save it to the same file without creating something new. But you can always do it this way. So, it’s looking at the CSV structure – lemme go read each page so I can craft captions.
Chapter 6: Reviewing and Correcting AI-Generated Captions
[00:09:18] It failed to fetch it. All right. I am glad that it did this. Let’s, first of all, we’re gonna stick with the left side of this, because it did open the CSV file and give us a little hint at what it looks like on the right. It’s wrong. Let’s pause here for a second though, and read over here so it synthesized. I got the content Fear versus FOMO, grabbing the second link. Now, you are gonna run into that sometimes where it’s like, I can’t grab that link.
[00:09:41] That is gonna be based on APIs and like the tokens that it’s using to access things on the internet. But again, it’s using context on the internet, metadata, things like that. All right, so it’s got everything we need. Here’s what we’re working with. Fear versus FOMO article is the first link, and then a menopause podcast page. So, let me draft those scroll-stopping captions. Okay, so now it’s giving me a summary of what it did.
[00:10:04] Again, we’re focusing on what it’s telling me it did versus what it actually did, because like I said, it’s gonna mess up. All right, it’s written the captions based on what it could find about the context about each of those and the content in each of those links. Now, notice that it did pull in some expertise about how LinkedIn is working. Lately, the algorithm is not really digging where links are sitting in the caption and preview, but rather to appear in the comments.
[00:10:34] You can absolutely do it. Test, try to see what works. We can totally do it for this. Alright, so hashtag recommendations. It gave us those. All right, now let’s go look at what it did give us. It’s giving us a preview of the CSV file, and I want you to look at what we got. Do you notice anything amiss? I do. Basically, what it did was it gave me a caption. Now let’s read the caption. This is across several cells. Not gonna work. Okay, now we’ve even gone to the next column.
[00:11:05] Now, also notice when I gave it my prompts, I didn’t tell it how things should be formatted. It kind of figured it out. Now, I could have included saying, Hey, just so you know, the links I’m giving you are in column B. I need the captions to be in column A corresponding to the same row that their link is. That is what I’m about to do now, ’cause I’m gonna tell it to correct itself, but that’s not all. I also need to correct it a bit because I technically didn’t drive across the country for that. So I’m gonna tell it what it messed up. All right.
[00:11:38] So, I told it what it did wrong. I’m like, you gotta notice that you kind of fucked up here. Now, we’re gonna see what it says. So, it’s now thinking about consolidating fragmented captioning, which is exactly what happened. Look, it’s telling me I’m right. (I know!) Line breaks and commas in my captions wreak havoc on the CSV structure. Let me fix that. So again, notice I don’t have to fix it. Could I have gone and fixed that myself? Absolutely. But I can have AI fix it for me and say, Nope. Take another stab at it and see if you can make this work better. Now it gets into things that I don’t know about called Python.
[00:12:10] Okay, let’s actually see if it did it correctly. So, the artifact didn’t quite update, and the artifact basically is when it’s like, “Hey, we’re having a chat over here.” Artifact is when it’s showing you kind of the results, the output, the outcome, the file. Anything that it’s generating, saying, here’s something you can go back to and grab. This is the result of what we are. It’s a product basically. But notice if I’m gonna go in here and I am gonna go ahead and spread these guys out again, ’cause I do like that. You’ve been a bit; now let me do a wrap text on these so I can actually see. Okay, there it is. And it’s got the links over here and everything like that.
[00:12:47] Now, I did say I was gonna tell it to try to rewrite these, but now that it’s kind of got it situated and it’s no longer formatted incorrectly. Let’s go ahead and tackle the caption portion. And again, you may get, you may luck out on your first try with this. As you continue doing this with it and it gets to know your voice and know what you approve of, it’s going to do a better job. The key here is to tell it if you want anything changed. Also, I have just opened this file. I can make edits in here based on what I think it should say.
[00:13:19] I highly recommend that you give that feedback to the chat, to the AI chat, because then it’s gonna say, ah, I see what you changed. It can even ask you why you changed it, and it’s great to give it that context, but then now it’s going to remember that and apply those lessons to the future captions that it drafts for you. Also, just a quick note: I am hitting shift enter, so that way I’m not like, oh, let me just doing this one at a one piecemeal at a time. I’m a big fan of just giving a whole swath of feedback and then hitting enter so it can take it all.
Chapter 7: Finalizing and Uploading Your CSV File
[00:13:55] All right, so I’m gonna go ahead and give it that feedback, and now it should make some updates to the captions and give me a new file that I can mess around with. I’m gonna go ahead and close this file out. Here we go. I appreciate the fact too that it’s telling you what it’s doing. If you’re ever interested in like kind of the things that are going on behind the scenes, you can always take a look at that and see now it’s ready to give me the file to my user. Notice that it is giving me some logic to cap FOMO as is. That’s an acronym. Let’s see how it did. Okay, great, great, great.
[00:14:24] Download it again. We’re gonna get on version three. All right. This looks like it might be correct. Now, the last thing I’m gonna do: let’s give it one more prompt. All right, so I took a stab and said, “Hey, why don’t you give me your thoughts about what date and time that second post should be going out?” So notice that it did do some thinking about that for LinkedIn. It’s gonna say, let’s publish that MenopauseIt post on the 6th of January at 11:30 AM 6th of January, if you weren’t aware, happens to be a Tuesday.
[00:14:59] All right, so lastly, let’s check our file version number four, and we’ll talk about this in a second, about all the iterations. Okay? So, let’s spread everything out. We’ve gone through this exercise. You may be asking yourself right now, or asking me, Rachel. Wouldn’t it have just been faster had I just gone through and written all those captions myself in the correctly formatted file, and then I can go bulk upload? Sure. Maybe the first, second time. Absolutely. What if you had 30 links to share? 60. What if you’re scheduling for a whole quarter?
[00:15:39] I don’t know your work process or workflow, but based on your brand and based on your whole content calendar and strategy, you may have a whole lot of content that you really do need to bulk upload and listen. We just did two posts that would go out on LinkedIn. You might have 20 posts, 200 posts that need to not only go out on LinkedIn, what if they also need to go out on Twitter and also need to go out on Facebook? You can use this method, though, where you’re grabbing these links and saying, let me at least give you a list of links and based on where you grab those from, you may not have to do it where you do it like I did, and go copy them out of a URL bar. You can actually export those out of your website, your CMS, or I actually wind up using this method.
[00:16:29] I have Claude integrated with Visual Studio that is connected to my client’s GitHub, and I have actually asked it saying, Hey, grab the last 30 days of blogs with this specific tag, and let’s set this up where they’re in a bulk upload file and it knows how to do it. It can save you loads of time. So if you’re asking me, asking yourself, asking Claude, why am I spending all this time going back and forth with you to get this particular thing right. It’s just like training another human who’s gonna be doing a job for you, it’s worth it to you to train them to do it correctly, even if that takes some back and forth and iterating.
[00:17:09] Because once they know how to do it, that’s saving you time from doing it yourself. That is the whole point of this. And please note the human being here was very important to have. We could not have just taken the first product, the first outcome of it, working on that file and uploaded it. Agorapulse would not have known what to do with it.
[00:17:29] So it did need those iterations and the checking. But once you get that process set, and this applies not obviously to just social media bulk uploading, this can apply to a whole lot of stuff. Use your imagination.
Chapter 8: Conclusion and Final Thoughts
[00:17:42] Hopefully this video got your imagination going and may help you do an epic bulk load process in 2026 and beyond using your favorite AI program. This one happened to be Claude, working with Agorapulse and of course Excel. But you can do whatever you want. I hope you found this video helpful. Lemme know in the comments. Lemme know your questions. Let me know your feedback, and let me know how it worked if you tried it.
[00:18:09] You’re still here. It’s over, but you should hit like on this video before you head out and subscribe if you wanna get “Moore” videos that’ll help you realize how much “Moore” there is to marketing.
But wait! (Yep, there’s more for subscribers.)
Exclusive content is available with Rachel’s livestream videos: Boss Level Mondays and Hot Mic Fridays. Start your week off with irreverence and real talk as you face down the work week of a marketer with Rachel at your side. Then, on Fridays, head to Hot Mic Fridays for unfiltered takes from the week, as well as podcasting and video tips you can always use. Subscribe here on Substack to access and join Rachel live!