Scrape & Auto DM Every Social Media Platform Using AI (2025)
1:00:29

Scrape & Auto DM Every Social Media Platform Using AI (2025)

Nick Saraev 05.04.2025 36 507 просмотров 1 135 лайков

Machine-readable: Markdown · JSON API · Site index

Поделиться Telegram VK Бот
Транскрипт Скачать .md
Анализ с AI
Описание видео
Join Maker School & get automation customer #1 + all my templates ⤵️ https://www.skool.com/makerschool/about?ref=e525fc95e7c346999dcec8e0e870e55d Want to work with my team, automate your business, & scale? ⤵️ https://cal.com/team/leftclick/discovery?source=youtube Watch me build my $300K/mo business live with daily videos + strategy ⤵️ https://www.youtube.com/@nicksaraevdaily Link to whiteboard ⤵️ https://link.excalidraw.com/readonly/dbLZYNzJiSfJ1LZ8Q8MX Summary ⤵️ I used AI and automation to scrape leads from *every* social media platform, including Facebook, Instagram, Twitter, LinkedIn, X, TikTok, and YouTube. My software, tools, & deals (some give me kickbacks—thank you!) 🚀 Instantly: https://link.nicksaraev.com/instantly-short 📧 Anymailfinder: https://link.nicksaraev.com/amf-short 🤖 Apify: https://console.apify.com/sign-up (30% off with code NICK30) 🧑🏽💻 n8n: https://n8n.partnerlinks.io/h372ujv8cw80 📈 Rize: https://link.nicksaraev.com/rize-short (25% off with promo code NICK) Follow me on other platforms 😈 📸 Instagram: https://www.instagram.com/nick_saraev 🕊️ Twitter/X: https://twitter.com/nicksaraev 🤙 Blog: https://nicksaraev.com Why watch? If this is your first view—hi, I’m Nick! TLDR: I spent six years building automated businesses with Make.com (most notably 1SecondCopy, a content company that hit 7 figures). Today a lot of people talk about automation, but I’ve noticed that very few have practical, real world success making money with it. So this channel is me chiming in and showing you what *real* systems that make *real* revenue look like. Hopefully I can help you improve your business, and in doing so, the rest of your life 🙏 Like, subscribe, and leave me a comment if you have a specific request! Thanks. Tools Mentioned ⤵️ Twitter - https://www.drippi.ai/ Instagram - https://autoigdm.com/ Instagram (email scraper) - https://apify.com/bhansalisoft/instagram-email-scraper LinkedIn - https://expandi.io/ LinkedIn - http://heyreach.io/ Facebook - https://reachowl.com/ Facebook (group scraping) - https://chromewebstore.google.com/detail/instant-data-scraper/ofaokhiedipichpaobibbnahnkdoiiah YouTube - https://apify.com/scrapestorm/youtube-email-extractor---fast-cheap TikTok - https://apify.com/clockworks/free-tiktok-scraper Chapters 00:00 - Introduction 00:35 - Twitter/X Scraping 14:26 - Instagram Scraping 29:33 - LinkedIn Scraping 40:20 - Facebook Scraping 48:58 - YouTube Scraping 51:33 - TikTok Scraping 59:38 - Outro

Оглавление (8 сегментов)

Introduction

Hey, today I'm going to show you an exhaustive list of ways to scrape every major social media platform currently known to men. That includes Facebook, Instagram, LinkedIn, Twitter/X, Tik Tok, and we're even going to do YouTube. I'm also going to show you how to automate outreach on those platforms. So, you're not just going to be able to get a list of leads, you're going to be able to put those leads into a sequence and actually sell them stuff, which is what I think most people really care about. I have a giant flowchart here that shows you how to get up and running on every platform. So, if you guys want that, just head over to the uh link in the description to find it. Let us start with the very first, which is going to be uh Twitter.

Twitter/X Scraping

And I'm actually going to go as far as literally signing up to the platforms for you guys. So, I'm going to whip out my super cool uh AX and we're literally going to be paying uh money and signing up to these platforms. All good in the hood. I want to show you guys how this stuff actually works if you were to go out and do it for a client today. So, starting with the first, which is Twitter, the number one simplest and easiest way to scrape and then do outreach to Twitter profiles right now is the drippy. ai. Uh, I'm not affiliated with Drippy at all. In fact, the vast majority of the platforms I'm going to be talking about today, but it is a very straightforward way of essentially just plugging in a couple of parameters and then getting a list of leads that you automate the outreach to. So, Drippy is right over here. You just type drippy. ai. It's an AI assistant for your Twitter or XDMs. They pitch it this way because this is obviously not cool or funky dory with the Twitter terms of service. So, they have to kind of pitch themselves as an assistant more than anything else. Let me put in my email address. Next up, they're going to send me a verification email. So, I'm just going to open up my email. Cool. Then going to verify. Verification looks good. I should now be able to just refresh this. Okay. Awesome. Now we are in Drippy. So we're going to do some onboarding here. I'm just going to add some basic answers, I should say. I'm now going to add or connect my Twitter account. So in order to do that here, you obviously need to actually manually connect your Twitter. So I'm going to do that. You then have to sign into Twitter. And then I'm just going to save my password. They're going to then email me an additional verification code because obviously they don't want you to um be doing any sort of scamming. Paste that in. Should now be connected. We have an onboarding session that is optional. I'm just going to skip that. And then okay, great. Now they're going to walk you through the process of adding a lead list and so on and so forth. Uh I'm just going to do that myself sort of next. So, it's not very difficult to do. A lot of the stuff is self-explanatory. So, to add a new lead source, I'm just going to click new lead source. And you have a couple of options here. They're going to allow you to scrape leads from a tweet or a Twitter profile. You can also just import leads from a CSV file, which I'll talk about in a second. And then you can do lookalike leads, finding leads similar to your existing customers or ICP. So, in my case, um, yeah, I'm just going to click this to start, but I I'm also going to run through the other options here. Lead scraper basically means when you click this button, little modal is going to pop up that's going to allow us to let's say we want to do uh I don't know an account source. All you do is you literally just put in the account that you want to scrape the followers of essentially. So um I don't know. I mean there's a lot of options I have over here. Uh why don't I just do I don't even know if we could do myself, but I'm just going to try doing myself. That way I don't have to muck around with anybody else's. So I'm just going to add this as my source. Cool. Looks like it pulled my data successfully. And what do I want to do? do followers or following? Maybe I just want to, you know, contact the followers of some automation dude. Okay. Now, what's going to happen is it's going to start importing that list of followers. Now, um this is a preview uh of 400. And I believe they do 400 or maybe 10% or something like that. So, we're just going to let that preview run. After that preview is finished, uh we'll click next. And here we have a couple of options. We can um continuously like we can add more. So, we could filter leads by the number of followers that they have. So, we could grab my leads and then we could say, hm, I only want to, you know, reach out to leads if they have more than 100 followers. We could also filter by the number of following. We could exclude leads based off of whether their bio says something or whether their bio um like doesn't have something. We could have a number of tweets filter. And then we could also just reuse other saved ones that we have. So, in my case, I'm just going to filter leads and I just want them to have more than 100 followers because if they have less than 100 followers, I don't know, odds are they're just new profiles or maybe they don't really have any valuable things to say or whatnot. I want people that are a little bit more socially engaged. So, I'm just going to click apply. Okay. Then I'll go down to next. And then um we now are essentially just previewing our changes. So, I have one data source. I have one filter applied. Brought my 4,000 down to 1,400. Then they're also going to give us a lead credits amount. And in this case, I'm just going to say Nick Rives followers greater than 100 followers. Then going to save. And now voila, we have our drippy. ai lead source. Um, it's already done all the lead scraping for us. Uh, now it's currently importing, which is pretty nice. Obviously, if you want to actually do this, you are going to have to, you know, upgrade your account. So then you just head over to their um app. drippy. ai/ settings/plans. And here you'll have access to a bunch of them. There's like a starter plan, an unlimited plan, then custom plans. So, um I'm just going to click start free trial on the starter plan cuz that's just all I need for the purposes of this demo. Nothing super fancy. And then I'm going to go ahead and then start my trial. Okay. So, while all that stuff is importing, I'm going to head over to automations. And then what I want is I want um to set up my cold DM automation essentially. So, what I'm going to do is I'll use cold DM 2. 0. This allows us to send outreach messages to users you've imported as a lead source. So, I'm going to click next. Then the source I'm going to use is going to be Nyx followers. Then we have the option here to either layer AI personalizations into an existing script or provide the AI with instructions to write an entirely personalized message. So you kind of have u um you know two options here. In my case, I'll just use script personalizer. And to be clear, I actually recommend you guys just like use a AI to personalize an element of an already pre-existing script because uh somebody that's read, you know, sent several hundreds of thousands of emails at this point, that's typically how you get the highest ROI. You don't have AI take care of everything for you because AI can just kind of be flamboyant and annoying and the tone of voice tends to suck. You actually just like have a a snippet of text that works and you know it works. You just have AI customize a couple of parts about it. So here is where you have the option of setting this up and then they actually allow you to test it which I think is really cool. So there's standard and then there's advanced. I'm just going to use standard. And essentially um what you do here is you provide a little bit of context. So who are you? Who is your business? Um what's your offer? So, uh, I'm just going to put my context here. I'm going to say, you know, my name is Nick. I run a YouTube channel with not my anime list. net search with 60,000 subscribers where I teach AI automation, cold email, um, cold outreach, and growing your online business. I do this by drilling deep into process optimization and doing live builds, showing people how to put complex automations together in a simple fashion. I don't know, just pulling stuff out of my ass here. Hopefully you guys can tell. Uh, then you have your own script. So, here is where you would actually write your script and then you'd put in like your personalization variables and stuff. So, I would probably go, "Hey, first name love. " And then here I'd go some specific aspect about their profile. Refreshing to see some refreshing aspect of their profile that's different from other people's. Thought I'd reach out. For context, I run a YouTube channel at around 60k subs and I'm looking for, I don't know, let's hypothetically say like clippers for my YouTube content, which is actually something that I am currently doing right now. So, I run a YouTube channel around 60. I'm looking for clippers. Your profile came up as a good fit. short form. Given your experience with short form, would there be any interest in giving this a try? I have a challenge set up here. I'm just going to go to my challenge. Uh, and the way that this works for people that are unfamiliar is basically I just have a bunch of content and I'm offering people the ability to clip and post my content on their own websites for on their own um social media platforms for a dollar per thousand views. So, I'm going to paste this in. Let me know. Thanks, neck. Cool. Check the settings. Creativity. Nope. That's all good. We can include website context, include tweet context, and then uh yeah, we're basically good to go. I'm just going to generate a message to this perpetual comics person here who love the fact you're a comic collector, especially someone with a part-time role at Oracle. Thought I'd reach out for context. I run a YouTube channel at 60k subs. Your profile came up. So, this is like the standardized one, right? Let me know. Thanks, Nick. Cool. And I used 002 credits for that. Very nice. Here we have the ability to generate some message variants. So, we can actually have AI do it. So, I'm just going to click generate variants. It's now going to generate a couple of different messages. The reason for this, uh, why you might want to do this on Twitter anyway, is just because, um, you don't want to send the exact same message over and over again. You want to like vary the messages that are being sent out so that Twitter doesn't think that you're just like spamming a bunch of people via DMs, which, as I mentioned previously, is not exactly part of their terms of service. AI is going through the whole rigomearroll. Maybe while it's doing so, I'll just make a little change. I'm looking for video. Let's do video editors instead. And I'm just realizing like if I'm going to run this, I should probably change the primary audience so that it's video editors and not necessarily people that follow my channel, right? But I'm sure you guys can kind of see how you'd piece stuff like that together. Maybe you follow the followers of some video editing channel that uh teaches people how to edit videos. Well, odds are people that follow that person are going to know how to edit your videos. So you just perform customized DMI reach for them. Cool. Looking for great editors and are a fan of styles like um let's just do screenblade. Cool. Let me know if so. No pressure. Cool. That looks pretty good to me. Then we have some follow-ups that we can add. So, I'm not going to add any follow-ups. Um, well, maybe I will. I'll just say just checking in on this. Let me know. And then after that, we could also schedule how many messages we want to send. So, maybe I want to send 10 messages per hour or something like that. I don't know. Not really. Um, you generally want to keep your limits pretty low here. I think on average, you want to send like 50ish DMs a day. So, about 1,500 a month max. You can also set a delivery schedule like, you know, Monday to Friday between 9:00 a. m. and like 6:00 p. m. PT, for instance. I'm just going to leave that delivery schedule off and just send it all the time. And then if I've already previously contacted somebody, I'm just going to leave that off. I can autofollow people if I want to auto like them, auto retweet. Uh, you know, there's a lot of options here. Um, but I'm for now I'm just going to leave all of this as next. Finally, then we enter the automation name. So, I'll just say primary DM automation. And then the sending account is going to be Nikki. I'm just going to create. Once I've created this thing, obviously we are good to go. um it is going to pull uh leads directly from my leads section here and then start sending uh outreach to all of the people that I've listed here. Right? Import is now complete. We're basically good to go. Um so that's basically all you need to do. Um once you're done that, it is now live. The automation is sending. You are good to go. Um you can obviously turn off the new cool DMs or follow-ups if you want. You also have the ability to um pause the whole outreach campaign here uh if you so wish. But um yeah, that's literally all you need to do in order to get this up and running. Now, let's cover a couple of those other lead generation methods. So, you see how earlier um when I was creating my little lead source, I could do lead scraper, CSV import, or lookalike leads. Well, CSV import is a pretty cool mechanism that you could use in order to find leads that I don't think a lot of people know how to use, but you can actually drill much deeper down than their um their defaults. The way you do so is you do something like Apify, which is a scraper that enables you to scrape large volumes of profiles without necessarily having to use Drippy's interface. I can see this probably being more beneficial from a credit perspective as well, but basically it's a big marketplace where you could scrape any platform, and I'm going to cover how to use it in a few of my future ones. You could do Tik Tok, you can do Instagram. Um, in our case, we're going to do Twitter. And so, what you can do is you theoretically could actually scrape, let's say, based off keywords, based off tweets and stuff. So, you could extract all of the information that you could ever want based off of, I don't know, you could put like Elon Musk's Twitter profile or something, and then you could get a big list of all of the tweets that he's ever made. Then, based off all made, you can get a bunch of people that have commented on his Twitter posts. Then from that, scrape all the people that have commented on his Twitter posts and then use that as your list. Um, which would essentially be a customized one. Um, that Drippy doesn't allow you to do so organ uh organically, right? So, a lot of cool options there. Um, but I think I'm going to leave it at that. There's a lot that you can do with this and could obviously customize. Um the last mechanism is you could also do a lookalike leads. So what you do is you paste a link to the account that you want. So I might do my own. Then it'll actually go and then it'll find lookalike leads based off of people that follow that account. So, it's not going to get the exact people that follow your account, but what it's going to do is it's going to find patterns in their followings, their followers, the keywords on their profiles and stuff like that. Then, they're going to give you different people. So, this is useful obviously if you're scraping large volumes of people from one um account and you don't just want to do it from that one account anymore, or you're getting good results from that account, but you want to get people that are similar and you don't really have any more followers to scrape, you can just create a lookalike lead list um instead of like actually

Instagram Scraping

doing that primary lead list. All right, so that's Twitter/X. Let's now get into Instagram. The way that we are going to do Instagram is we sort of have like a truncated fork in the road. We can choose one of two paths and I'm going to show you guys both paths, but you could use Appify and Make. com. Appify is the tool obviously that I showed you a moment ago. And I'll run you through what that looks like. U Make. com is sort of like a visual automation tool that lets you um drag and drop various steps, which is, you know, very powerful tool that you shouldn't just consider in the context of lead scraping. You can do a lot with it. Um, and then auto IGDM is sort of like an all-in-one package similar to drippy. ai, but I'll show you what both of these look like. Let's start with appy and make. com. Basically, what you can do on Instagram is you can actually like extract email addresses from profiles pretty easily. And, you know, if you feed in a,000 profiles, you realistically be able to get like 200 300 email addresses from that, which are pretty good numbers. And then you can actually like reach out to those people directly over email. So, you could, you know, use Instagram as your scraping source and then you could reach out to people over email. Um, or what you could do is you could use Instagram as a scraping source and then reach out to people over Instagram using direct DMs. Now, this is kind of a nuanced point that I don't think a lot of people understand, but you know, social media platforms, they are controlled by a third party entity. And because they're entity, they obviously have very harsh rate limits in place. These third party entities don't really like when we game the systems like we're doing right now. So, this is kind of like growth hacky, right? The thing is, email is not really like that. With email, for the most part, assuming that you build up um the infrastructure, you can send out as many emails as you want in a day. So, in the case of Instagram, Twitter, LinkedIn, and all the other platforms I'm going to talk about, you know, like with Twitter, maybe you send 50 cold DMs a day or something like that, right? But if you hypothetically spun up an email campaign that sourced emails using Twitter data, you could theoretically send a thousand emails a day. There's really nothing stopping you. Instagram is really similar. So, in our case, I mean, we're going to be able to send realistically maybe 30 to 50 a day using a tool like Auto IGDM. But if you wanted to augment that, um, what you could do is you could just get the emails like I'm about to show you here, then use that to add to a cold email campaign and then kind of like communicate with them off platform, if that makes sense. Okay. So, how exactly do you do this? Well, as you see, I have a million in one tabs open here on Appify. What I did is I just typed on Instagram and I have a variety of scrapers that I could use. Generally speaking, I like um scrapers that are payer result. Those just are a lot easier for me and they're a little bit more cost effective. Pay per result just means if I get a thousand leads, they'll charge me for a thousand leads. If I get 10,000 leads, they'll charge you for 10,000 leads. The alternative on Ampify is you could like pay per usage or you could pay per server um uh per hour or rental or something like that. I find it's just not really that scalable for our purposes. So, I always just go pay per result. So, the one I'm going to be looking for is Instagram search. We're going to be using this one right over here. And this one costs $2. 30 per thousand results. Obviously, pretty reasonable and reliable. And uh the way that this works is you basically just pump in a query and then it returns a bunch of profiles for you. So users. So I'm going to go query um AI automation. And for the purpose of this demo, I'm just going to do like I don't know 50, let's say. Let's just make sure this is 50 users. Then I'm just going to click save and start um on Appify. And now what's happening is on the back end, this is actually spinning up a server on Ampify's um hardware and then it's using that server to like run their little private scraping thing. The cool thing about Appify is it has a bunch of built-in proxies. That just means that has a bunch of like scraping protections that minimize the probability that um that maximizes our ability to get these things and they just take care of all of it for us. We don't really have to do any work ourselves. Okay, great. So, as you can see, I got a ton of profiles and the really cool thing is on in addition to those profiles, I also got a bunch of websites. How cool is that, huh? Pretty sweet. All right. And it looks like I don't know, we scraped all 50. So once you're done with that scraping, you can just click export results. And what are you going to get as a result? A CSV file. CSV stands for commaepparated value. So this is just a quick and easy way for us to get a giant list of things in a format that is our own. We're not using a third party tool for this anymore. When we import a oops, sorry, I downloaded that by instance. Uh what we want to do is we actually want to import that CSV that we just exported. Um and the value here is this is our own tool. This is ours, right? This is our Google sheet. It's our data. We can do anything we want with this stuff. It's just voila. It's just all here. So, I'm clicking import. Once the data is inside, you can see that it's I don't really like the format. The um the sizes of these cells are really big. But if I drag this and make this a little bit smaller, what we'll see is we actually have access to the biography, the business address, city name. We have all of these fields basically with data in them, which is awesome. And the value in having all these fields with data is obviously we can then use this to get um email addresses from these accounts. Okay. So we're not going to be able to get all of the email addresses from these um of the 50. I don't know. My inclination is we're probably like 15 or 20 emails. Usually it's about like a 30% enrichment rate. But um yeah, we have now officially gotten a bunch of uh a bunch of profile data and we've put it in our own service. So what do we do from here? Um pretty simple and straightforward. you need to use what's called an email uh verific or sorry an email finding tool. They have a bunch of email finding tools available to you. Um I personally really like one called any mail finder. Now obviously all these cost money so there are variety of services out there that you can cross compare the costs with. I find any mailinder is a great mix of both cost and then reliability and I'm usually able to get like a fair number of the emails that I I look for. Um but yeah, so now that we have all of these um all I need to do is just head over to my anymailfinder. com account. I'll go to my app. And the way that an email finder works basically is, and I always recommend using the UI before you try and like upload a bulk file is you can search for um a company, you could search for a person at decision makers also do LinkedIn URL search. I'm actually just going to do a company search. The way that it works is you basically feed in the domain name that you're looking for, and then it'll go and it'll see if there are any accessible email addresses just by doing a bunch of math on the back end. like it's basically scraping this website uh for any email things formatted as an email. It's scraping all mentions of this website on the internet. It's doing a bunch of like what's called nominative email verification where they combine your name and stuff like that. And uh yeah, you know, if I type in my own website, left click. ai, you can see it literally found my email address right there, right? Pretty neat. So, what do you think is going to happen when we feed in a bunch of these Instagram profiles? Well, we're going to be able to find a bunch of their emails. So, I'm going to head over to bulk search and then click new bulk. Then, I'm going to go to company. And then what I'm going to do is I'm going to upload the CSV file that I just downloaded. So data set Instagram search scraper. That's it right over here. It's now processing the file. Um it looks like in our case we are exceeding the maximum allowed length. So what we need to do is we just need to grab some of these and then shorten the hell out of them. That's uh fair. We do have some of these that are just like unnecessarily and unreasonably long. So really, if I'm honest, the only column that we actually need here cuz we have a lot of data here, right? The only column that we actually need in this entire thing is just the um domain name one. So, I'm actually just going to go in. I'll grab the URL. I'm going to go all the way back here. Just click on this. here, very beginning where it says external URL here. I'm just going to delete literally everything else cuz yeah, there's like 2,000 columns in there. That's a pretty big ass file, right? Then I'm just going to remove a bunch of the other information that I personally don't really think is super relevant. And then voila. Now we have external URL here. We have URL, we have the username, and then we have a quick little biography. Um, I think the biography might be a little bit long. I'm just going to remove that, too. Obviously, play around with this. And various different types of Appify scrapers that you guys use are going to have different uh they're means of delivering this data to you. The one that I used was just like crazy long. So anyway, um now I'm just going to export this updated Apify search. Then I'm going to go back to my bulk scraper and I'm going to upload it again. Okay, we are currently doing some scraping here. Looks like we found 14 fully verified emails. Seven of them were partially verified. So actually that's like 20. Wow, 27. Not bad. We also have some that we just didn't find, right? Which is understandable, but um still pretty good. Second that it's done, I can then preview this. If I preview this, it'll actually show me the email addresses. Um, keep in mind that some of these are I'm want fully verified and partially verified as a CSV. I'm going to continue this download. That's okay. Uh, and then I'm just going to go back here and I'm just going to make a new Google sheet and then just visualize all these email addresses. So, I'm going to go import. I'll go upload. And we're just going to paste these puppies in. Okay. And now, as we can see, we actually have a ton of email addresses. Wow, very nice. Super cool. Now, as you see, some of these um some of these links aren't actually links to like the person's own website, right? Like school. com is not that person's website. School. com is like a aggregator that they just happen to have a thing hosted on. Same thing with I don't know like Linkree and Cal. com. So, these these emails obviously like they don't really matter to us, right? But they're still emails, which is pretty cool. And now all of a sudden, you have a big list of emails. You could pump this into some sort of email campaign, like an instantly campaign or something, and then get out there and start sending. But that's only one half of it. Let's hypothetically say you actually wanted to send DMs to these people instead. How would you go about doing that? Well, there's a dedicated tool that essentially does a lot of the heavy lifting for you in the back end. The only issue is it's, you know, it costs a little bit more than doing it peace meal like we did back here. And then, you know, it actually allows you to send it via Instagram as well. So, it's called AutoIGDM. Um, AutoIGDM has pricing plans that are monthly, quarterly, and then yearly. So, I'm just going to sign up to the monthly pricing plan and then just give this a go. Um, but let me walk you through the entire process. First thing is click start now and then enter your email address and then I'll do nick at leftclick. ai. Very cool. Create my account. Then you have to obviously pay for the service. So, I'm going to go through and I'm going to pay for my service. This allows you to send 1500 DMs a month per account. It's 100% cloud-based and it's also fully automated. So, there's no work kind of that you have to do on the back end. You just kind of add all the stuff you want once and then it just does for you. So I have to verify my email address as per usual. Let us go back here and then I'm just going to type autoigm verify. Create my account. I'm just going to um copy this link actually because I don't want to do it in this browser. Go back here and then paste it in. And now I go through the signup process of my account. So I'm just going to go this and then I don't know. Um let's not do that. Choose my own please. Let's do that. Then going to sign in. Save. Uh, now I just connect my Instagram account. So I'm just going to go Nick. Then it'll go through and attempt to connect to my Instagram account. So I need to log back in. And then I also need to add my own password. Um, if you add the wrong password, it'll, you know, kind of screw around. So just give me a second to go through and actually add my password to my own Instagram account. Okay. And then after that, you need to do what's called a secret key setup. So, they actually have a tutorial video that shows you um all of this. Basically, you need to go to account center. instagram. com. It'll take a list of all of the profiles that you have available to you. You go down to password and security. Then you go two-factor authentication. You go to the account of interest. It'll open up this two-factor authentication screen or page. When you click on that button, it'll then generate a key for you. You copy this key. And then what you do is you go back to AutoIGDM. Then you paste the key in here. Okay. Once you're done, you click add account. Then you have now connected your account and everything is perfect. So the system will eventually log into your account for you. Um you may get a notification saying we suspect automated behavior in your account. That's completely understandable. And then what you do now is connect your account to a campaign. So the way that you do this is click connect to campaign. Click add new up here. Now you have your campaign. So we need to link the account that we just added. So just click that button. Then we go in and now we can actually go um they're going to obviously give you a bunch of information about this, but that's okay. Um you then go in and then you uh reference specific hashtags or you reference accounts. Okay? So for instance, you could look for Instagram accounts that are engaged with a specific Instagram account. So in my case, I have another channel called Nick Sarafh here. I know that that's pretty good. Then I have Nick. Surf. And then I don't know. Let's just go Instagram AI automation. Let's just see what's the first thing that pops up. Um, this person's account, new link app. That looks pretty cool. Just going to paste that in. And now I have these three accounts that I am targeting, which is pretty cool. So from targeting, you can then go to filtering. You can add negative keywords and positive keywords. Now I'm going to be doing basic filtering. I'll be looking for people with profile picks, bios. Um, I don't know if I'm going to look for website links. I will exclude sexual content. Do definitely do not want to be DMing those. I get enough of those as is. Do not DM the user if you already had a conversation. I'll do that, too. Then you can also set a minimum number of followers. I'm just going to do 100. Okay. All right. From there, I then um have a couple of other options. Obviously, you need to create a message. So, what I'll say is there. And then um here is where you would add your specific message text. So, hey, first name. just thought I'd reach out. Love your posts on AI automation and um in the same field. So that's obviously a pretty trash outreach message. Ideally, you would want some sort of reason why um that's a little bit more justified than, oh, I think you're really cool, but I just want to make sure you guys can see how this works without spending God knows how long crafting the perfect message. Um I'm then going to send follow-up after 2 days. And then I'll just say, "Hey, first name. Checking in on this. No hard feelings if you don't want to add me. " And then because I'm cheeky, I'm going to add a funny little face. Okay, great. After that, you define the schedule. So maybe I want 9 until 5 Mountain time. Okay, pretty nice and simple and easy. Then I submit my campaign. Once I've submitted my campaign, it is now um basically in their system. And once it's in their system, we are queued up to start doing our outreach automatically, which is pretty sweet. Okay, great. And now it's going to ask us if we want to submit the campaign. You just click yes. And then voila, it has now been um submitted successfully, which is fantastic. I'm also just going to save a draft of this campaign just in case. Yes, I'm going to leave that page. Cool. So, we have our account connected. We also have the campaign that is going. Um, when you submit, it will not begin immediately. I believe either their team reviews it or they have some process in the back end to take care of it for you. Um, but more importantly, there is a statistic page where it'll actually show you all the leads that you've been messaging. It'll show you results on those leads. Um, I think you get like a nice little visualization as well. And yeah, that's a quick walkthrough of how to accomplish that second big step, which is a direct DM path where they basically take care of all of it for you. All right, now let's get to the motherload, which is LinkedIn. Now

LinkedIn Scraping

there are variety of different ways that you could do LinkedIn outreach. I've talked a lot about them across my channel. have obviously created LinkedIn automation systems in NAN and make. com and stuff like that. I'll have that be one of the options. Then I'll show you some other options that you guys could use as well. So on the far left hand side here we have LinkedIn automation using hey reach or expand. So just like drippy and just like um auto IGDM this is like a third party tool that you sign up to. You pay some premium fee then they basically just deal with it all on the back end for you. In the middle here there's this LinkedIn AI outreach system which I developed um that uses a noode platform called NAD and then another platform called Phantom Buster. kind of just connects the two together like glue. Um, and then you know the end result is you could send customized DMs. And then the last one is just going directly using sales navigator and email. Um, I think instead of sales navigator I'm probably just going to do h maybe I will do sales navigator. I don't know. I might just use appy for that. I'm not entirely sure. We'll kind of play it by ear. But um, in reality like you can mix and match a lot of the tools that we're talking about and it's really just a matter of finding the most cost effective one. Okay. So let's start with this tool. Hey Reach or Expandy. You can use either one. I'm just going to use expand here. Um, and the reason why I'm going to do it is I think the pricing is a little bit cheaper than the alternative. It's like 99 bucks a month or something. Um, you end up with uh a lot here. You know, you could set up templates. You could do campaign prioritization. Tons of different campaign types. Um, country based IP addresses to minimize the probability of you getting like blocked and stuff and so on and so forth. So, what I'm going to do is I'm just going to click generate leads and then I'm going to go through their little signup process and then I'm going to show you how to do this. Under account type, I will use Expandi for my company or personal goals and click sign up. They're now going to send me an email which I can obviously then verify just by checking out the Expandi verification. I'm going to copy the link address, just open it in this tab. The reason why I always do this is it just I don't know. I don't want to like be constantly all tabbing between different tabs. Just want to have it all here. After that, it will run me through some onboarding. So, I'm just doing sales. I'm heard about Expandy, but I've never used it. Let's go. This is Nick's workspace. You can see they're onboarding the hell out of me. I like it. Now, I'm going to connect my LinkedIn account. I'm going to use a Chrome extension for this. So, I'm going to click this button. I'm going to download this Chrome extension. And basically, instead of you having to enter your username or password, which is fine and all, um it it's going to use a Chrome extension which will automatically connect to your LinkedIn for you. So, it'll take you to the LinkedIn page. You'll then go here, continue to Expandy. Then, it will redirect you back to the app um page that you were on before, except this time it'll also have your cookies, which will allow it to access everything that it needs to access to automate the hell out of your processes. Cool. Um I'm now going to choose the country that I want to run this in. Now, I am Canadian, and so my usage up until now has been from Canada. So, I'm going to want to minimize the discrepancy between the usage and then obviously where I'm going. Uh, I'm now going to connect my card details. Okay, it's now creating a workspace subscription for me. After that, it's going to synchronize my account. Basically, run through that Chrome extension again, do some connecting behind the scenes. Then, it's going to ask you to invite your teammates. So, I'm just going to do that later. And then, voila. You know, we have some goals that we can put in here. Um, get your ideal customer profile. That's what ICP stands for. So maybe um I don't know I help exa uh let's say sales chief sales officers with driving leads affordably. My goal with this campaign is to book more sales meetings. Nice. So I'm going to take that and then AI use it to like do something cool. Um and then we can do a basic search. sales navigator search or we can do a recruiter search. So it recommends that I do the basic search. So, um, what I can do is I can find prospects in LinkedIn, then copy the link, then paste it directly in here. So, I could just actually just type this in, and it'll actually go through and like it just used AI to generate the search. Um, so you can obviously rank this down and get super into the nitty-gritty here. You could add a bunch of filters based off location and where the person lives and stuff, but this is the link that you are going to be providing expand that it is going to be doing the scraping from. So, we're searching for chief sales officers using this link. Okay. After that, we can just enter the name. We can also auto reload this. Um, and it looks like you can't actually autoreload it right now. You need to do it the second time for some reason. Maybe it's just a little bug of the way that they do things. And then here is more or less how that works. Um, how the entire platform works. You reach out to second and third degree connections with personalized connection requests and follow-ups. But you can also select a variety of other types. You could do sequence campaigns, groups, event invites, mobile connectors. You could actually try and do event participants. So you could do messengers, company follow invites. You have a bunch of options essentially, which is pretty cool. Um, for now it's just going to use AI to like, you know, whip up the copy of your campaign, whip up everything you need. So this is the quote unquote personalized message. Do I think these personalized messages are great? No. If you guys want more insight on that, just check out some of my copyrightiting videos. I think you can realistically do a lot better than this. Generally, you just want to make it seem as personalized as humanly possible. Um, you can allow it to send with or without the connector message. Um, I'm just going to leave it at as this for now. And I'm going to leave the follow-up messages as default. Follow-up messages as default and all the spacing and stuff like that is default as well. It is nice to allow you to get um kind of like I want to say 80% of the results just out of the box with AI. Um but I do find that last 20% is super valuable. So I would highly recommend if you're going to launch any sort of campaign here, definitely make sure to do that yourself. Make sure you know what the offer is that you're selling. Make sure you know how to pitch yourself. Make sure to put yourself in the prospect's shoes. you know, if somebody were reach out to me, you know, how could they make that outreach look as personalized as possible? Um because, you know, if it's not, likelihood of somebody getting back to you um is a lot lower. Okay. And once that's done, the leads are now ready for action. You can let the networking begin, they say. And then from there, we are good to go. So, you can create your workspace based off your activity. You know, add your working days and so on and so forth. I'm just going to skip that. And then you're good to go. This will start scraping leads on the back end for you. Once it's ready, all you need to do is just confirm by giving this a click and then going to the top right hand corner and pressing activate. You won't be able to do that until the leads are done scraping, but that happens relatively quickly, I would say. All right. Aside from that, there are a couple of additional ways that you could scrape LinkedIn. I'm going to be covering the LinkedIn AI outreach system by the yours truly, Nick Saraf. Um, I use a variety of tools to that end. The big tool that I use is called NADN, which is a no code drag and drop, no or low code drag and drop builder that allows you to like automate a lot of processes in the back end. And so I've actually created a flow here which you guys can check my video for. Just check the um the link up here or just look to I think it was my third last video that just says Naden LinkedIn system or something like that covers how to do this. But essentially what we do is very similar to what Expand is doing. I'm just going to go um on form submission. Give this a quick test. It'll now ask me to describe my audience. So, I'll just say founders, partners, or um co-founders at creative agencies between 1 to 100 people in the United States. If I submit this, um I can then just run through this test. The reason why I'm showing you guys how to do this in a no code tool is just because I think it's important that you guys reduce your reliance on third party tools cuz these third party tools do like take a fair cut off the top, right? Like it costs a fair amount of money to use them. As you saw, I've already spent like 300 bucks. We we're barely uh you know, it's barely noon today. Um, when you know how to use no code tools like this, you can usually do a big arbitrage on the value of these tools and then you can get up and running for cents on the dollar really. Anyway, um, I generated a search URL in Apollo and basically I just got the same sort of thing that you would have gotten in LinkedIn. So, a big list of people. From there, uh, what you do is you basically run what's called an appy actor. So, if you aren't familiar, an actor is the term for a scraper used in the tool that I was showing you guys a moment ago, Apify. And this no code tool nan actually allows me to spin up and scrape a list of leads using that completely automatically. So I don't actually have to go through and you know like manually do it or anything. I actually just do it all automatically. And what it's doing is it's now scraping a bunch of leads for me. So it's actually getting a bunch of email addresses of these people which I can use to do any sort of outreach that I want. Okay. And then once all that's done the scraper output it's find over here like we were doing before. Or the way that this works is we actually have access to all of that data now just in NAND which is pretty cool. Um, and then we can actually go and we can do like some personalized outreach. So, my personalized outreach is a little bit better, I think, than most of the personalized outreach out there. I basically take a sentence from their profile data and then I use it to customize a little bit of outreach. Um, so love seeing your creativity, love seeing your knack for fresh ideas. You know, I'm looking for X, Y, and Z. Would love to connect. From there, I add it to a Google sheet, which is uh not super complicated. Essentially, I'm just adding a new row to a sheet, which I think we've probably all done here. And then um finally, at the very end, I trigger what's called a Phantom Buster agent. And this is the cool part. Um this is where we use another third party tool called Phantom Buster. But I find it's just a little bit um more cost effective than the other ones. If I go to the Phantom that is running, basically what I have is an auto icebreaker connect. This is one of the many tools that Phantom Buster allows you to use, but essentially what this one does is this one actually just sends a connection request along with like a customized message for you. So, I go down to results. I've actually already sent a bunch of connection requests to people and um the messages are like this one. Hey, Frank, loved seeing your creative director role. I'm also crafting stories. Thought I'd connect. And then, um I actually automatically send a connection request alongside the DM. So, if you think about it, under the hood, exact same thing Expand is doing, but I can do it all for substantially less money than that ongoing membership. Okay. And then the last way to scrape things from LinkedIn is you scrape the data from LinkedIn and then you just do outreach over email instead. And this is actually the primary way that most um AI automation agencies, uh recruitment companies, and other services do things. So instead of actually going to LinkedIn directly, what we're doing is we're using a third party service, Apollo, which I showed you guys a moment ago. The thing is Apollo just scrapes other leads from LinkedIn. Anyway, so we're scraping a service that scrapes a service. So the value there is um you know, it's just a lot cheaper than if you were to go to LinkedIn Sales Navigator directly. And uh what we do is um we run the exact same flow that I showed you a moment ago. So, we export these 500 results. And now we have a bunch of email addresses, right? So, with these email addresses, we then just pump them into a cold email tool like instantly. ai or uh maybe Smartle as well. Smartle and Instantly. ai are probably right now like the two big competitors here. Um, and what you can do is you just paste the email in, then you write your own sequence just like we did earlier, and then it automatically sends them out um based off of some mailbox uh settings. So, if you guys want a guide on how to do that, I've made many videos on my channel about how to get up and running with cold email very quickly. Um, but this is another way that we could do a data source. All

Facebook Scraping

right, let's cover the lovely Facebook. Facebook has a couple of different routes that you could take. You could do um Facebook group scraping using a show, or you could use a third party tool called Reach Owl to do the connecting for you. So, just to keep things simple, I'm just going to start with Reach Owl and then we'll talk all other stuff later. Um, in order to do this, just type in Reach Owl. This is the platform that we want to use. Go to their pricing here. You'll see that their basic plan starts at $29 a month with one social media account. So, I obviously want to connect Facebook. Uh, what I'm going to do is I'm going to click register up here. Then, I'm just going to go through their sign up process. Once that is done, they have a little onboarding sequence they're going to take you through. Then you can continue to dashboard. And now you just connect your social media account. As you see, they also do Instagram, but I think their main value is Facebook. I think autoigm does better. And I'm going to log in with cookies. Now, what you need in order to make this work is you need a Chrome extension called Cookie Editor. Cookie Editor just allows you to extract cookies on various websites. So, I'm going to be doing that for my Facebook account right now. Going to grab cookie editor, add it to my Chrome menu. I'm then going to pin it. Then I'm just going to go like Facebook. Actually, I think I already have a Facebook tab open here. Right there I do. And then what I want to do is um I just want to go into cookie editor. Um I need to request permission for all sites. Then down over here, you see I have a variety of cookies. I believe what I need to do is just export this as JSON. And then I go back to reach and then I paste my cookies. And let me just double and quadruple check that that's good. Okay, looks good. And then what I do is I just set an account title, Nyx Facebook. Then click next. Okay. Um, then you have to find what's called your proxy URL. And then after you're done with that, you just need to connect your proxy. So I just scrolled down from select from shared proxies and used 181214117123. It's much easier way to do things than actually having to go through the rigomear roll of like setting up your own proxy account, but I want you to know that you can set up your own proxy account if you want to. Okay. From there, it'll ask you if you want to start warming up the account. So, I say yes. Um, then you can set your minimum intervals. These are all recommended intervals. And then you can also just turn off the campaign if a bunch of your uh friend requests or messages fail in a row. So, I recommend turning that on as well. This is these are obviously all just safety procedures. Once you're done with that, you can actually set up a campaign. So, we go over here, create your first campaign. And here you have a variety of different things that you could do. Okay? So, for instance, you could um I don't know, you could extract Facebook profiles from a custom list of URLs. So, I'll show you guys all about that later, but then you could find Facebook friends of somebody that you're already friends with. You could do Facebook video comments. So, you could actually pull a video on Facebook and then extract all of the people that have commented on that post. You could also do ad reactions. So, if somebody has posted or given a thumbs up to an ad or something, you can actually have that um you can actually scrape the list of people there and then send them a message or whatever. Then you can do group members, you can do um Facebook post ad comments as well. So, I'm just going to do post ad reactions. I need to go find basically a Facebook ad URL. So, I'm just going to go over here and hopefully nobody that I'm connected with has done anything super embarrassing. Okay, so this one from Grant Cardone looks pretty good. I think in order to do this, I don't actually remember exactly how. I think I click embed. Okay, now we have access to that specific ad which if I click on I'll now go directly to the post and then what I do is I go back to hey reach paste it in here. You'll see that there's a the a bunch of extra fields here like question mark refbed. Um that's okay. What you want is you just want it facebook. com/post/post ID. Okay. So just delete this number. Have the post ID there. Click next. Okay. And then from here you can set up a message sequence. So do you want to set a friend or a follow request? You could be like these two lovely gentlemen. Or you could send a message sequence. I'm just going to call this example reach owl campaign. I'll run this from my Facebook. Um daily limit is 30. I'll filter by audience or location. I guess I'll go like USA. Um and I'll also do United States. Um the reason why is cuz I think this is just like a string. So some people are going to write USA on their profile, other people are going to do USA. I'll do both genders. Send message to new prospects only. Uh I'll say don't send messages to people that already have conversations with me. Then I'll do my pitch. I loved this. Hey, I figured you were interested in um Grant Cardone given that I saw your like on one of his ads. I run a YouTube channel at around 60k subs. I wanted to see if you might be interested in XYZ. Maybe I'll delay one minute and click save. And then I can start my campaign. If you just click start your campaign, it'll start. So voila, done. We are now um essentially reaching out to people as we speak. Uh well, first we're we need to warm up our account and sometimes that takes a while. Um but I'm just going to leave it at that. My campaign's all good to go. Obviously, there are additional settings that I could screw around with. Since I don't actually want to run that or send that exact message, I'm just going to turn mine off. Okay. And then the other way is we're going to use Appify to scrape a bunch of Facebook group members. Um, and then based off of that, what we can do is we can look through their profiles to see if they have like a website or something and then if they do, we can add them to an email sequence, same as what we've seen before. So, the way that we do this is we go back to Appify and I just type Facebook. Facebook obviously can scrape posts, comments, groups, pages, ads, ad library. The one I want is group scraper. Um, I think this one is $45 a month for rentals. Let me just see if there's one that might be paper result instead. Not seeing one specifically for groups, unfortunately. H Yeah, you might just have to do the $45 rental one if you wanted to do groups themselves. Regardless, I think they might offer you a trial. Yeah, they do offer you a free 7-day trial. So, I'm just going to sign up to this and we'll give it a try and then I'll just make sure to turn it off after I'm done. Okay. So, from here, I can actually enter in my group scraper. Just going to go to this group URL and see if um this is a maths formulas one, which is public. I think it needs to be public. So, we're just going to paste that in. Number of posts I want. I don't know. Let's say 10. Uh 20. Sorting order. New post. Search by letter. Search by keyword by year as well. And then you can do a time frame. I'm just going to click start. We're going to rent the actor for the $45 a month. And then I'll just make sure to turn that off after. Okay. So, after this is done, you can see we got a big list of all the posts that were made. Now, in this specific group, which I don't think is a good example, most of the posts are just made by the same person. But ideally, you'd have a group that's active and dynamic and you have a lot of other people that are commenting and stuff. Anyway, um what you'll end up with is you end up with this column called user with two fields and in it you'll have an ID. Okay, so this is actually the ID of the Facebook profile that left the comment. So if I go facebook. com and I think what I do, just scroll back here. I think I basically go user and then ID. So, I'm actually just going to feed in the ID and then you'll see it'll just go directly over to the specific person that left that post. Okay. And this person, Ellie Surya, that you know, runs a bunch of profiles and stuff. Um, the cool thing is when you have these, you can then go to about and sometimes under their contact or basic info, they'll have websites and social links. And the second you have a website and or social link, you can just do the same thing that we did earlier with LinkedIn, right? That's pretty simple and pretty easy. Um, in order to get all of that information, all you need to do is you need to take this data that you've gotten. So, we'd export this, okay, as a CSV, and then feed it into another scraper on Appify, in this case, called Facebook profile and page scraper. And what you do is you just feed in a bunch of profile URLs. So, it' be um facebook. com/ whatever that number is. And then you could actually feed that in. And then that would go and it would extract you all of the website data, the contact info, everything essentially that you need. Um, and the benefit there too is you can actually get very creative with this. you could scrape a bunch of their posts after to look for websites that are mentioned as well. So, I mean, obviously the rabbit hole here in scraping is really deep. Uh, my rule of thumb when I'm doing scraping is the more work that I have to do to get the information, usually the better the list is, but I kind of need to weigh that against the cost that it takes for me to do so, right? Like you can get a super nuanced, extraordinarily deep list that nobody's ever contacted before, but if it cost you like $10 a lead to do so, you need to make sure that the ROI works in your favor for whatever you're selling. Okay, let's talk a little bit

YouTube Scraping

about how to scrape YouTube. The simplest and easiest way is using the Appify YouTube to email scraper. I do this all the time myself and it's about as simple as we just go to back to our actors, back to our store. Um, just go YouTube and then type in the word email and you'll see that there'll be an email extractor channel email scraper mass YouTube email scraper. I mean there a bunch that people have set up. Basically, the way that this works is a lot of the time on people's YouTube accounts, myself included. You'll actually find an email address listed somewhere. I don't know where. I mean, it depends. Sometimes people put them directly in the bios. in like a little contact. Sometimes people say for business um for business purposes, you know, feel free to email this if you have an inquiry. But basically, we'll just scrape all those automatically using one of these and then just um add them to a big sheet for you. So, yeah, I see a bunch here. Uh let me just see if there's a paper result one. Um let me see. This one looks pretty good. advanced and cheap by Stormcraper. So, I'm just going to give this a try. And then what you can do is you can actually add a keyword for the people that you want. You're not even adding in a specific um account or something like that. What you're doing is you're actually just scraping emails for marketing. That's probably one of the easiest ones. So, why don't I type like AI automation? I'm cool with any of these email domains. So, these ones are all fine for me. Then you click save and start. it'll actually go and it'll find um email addresses on YouTube profiles that rank for the term AI automation. So, that's pretty cool. Feel free to use that if you guys want to um experiment with sort of like an all-in-one solution. There's also uh if I go back to all actors and then store over here. I think if you type channel email, it'll actually um allow you to put in a list of URLs. So, um I don't know, maybe I put in my my URL here, Nick Sariah for something. Um and then it'll go and it'll scrape specifically emails at that URL. Okay, so we just pumped in a bunch of AI automation channels in and you can see we've actually gone through and we've extracted emails for them um simply using the scraper. Um but yeah, aside from that, there are a couple of other ways that you can go a little bit more nuanced, a little bit deeper. um just type in YouTube channel email and then from there you'll get like a YouTube channel bulk email scraper, YouTube channel email scraper, YouTube email finder and so on and so forth. So what's the value here? Well, once you have um once you have that uh you can just pump this into an instantly or smart lead sequence and start sending cold outreach. Um there's no inherent DM mechanism on YouTube unfortunately. So I'll leave it at that. But um yeah, it's probably one of the simplest and easiest ways to get up and running sourcing data off of YouTube, especially with like more niche keywords and stuff like that. And then last but not least, here is how

TikTok Scraping

to scrape Tik Tok. Now, variety of different ways to do so. I'm going to be using Ampify scrapers. I'm not married to Ampify here at all. They're a great platform, but there are variety of other ways you could do this if you wanted to. I just continuously rely on this because they just have a bunch of like out of the box ones that work. And basically, you know, you can use Appify and then you can use a no code tool like make. com to take that data and then review it and then I don't know, add it to some email sequence like I was talking about earlier before sending them out. So uh first things first, we choose our scraping method. As we see there are variety of different things we could do. Hashtags, usernames, videos, relevant keywords, right? Then we just run this one actor called Tik Tok data extractor. I don't know where all of these arrows are coming from, but they are certainly annoying. So we're going to go back to our actors list store and I'm just going to type in Tik Tok. What you want is this data extractor one over here. This is by far the best. The value here as well is it's results base. So $4 per thousand. pretty easy videos with this hashtag. Let's type AI automation. What this is going to do is it's going to go and extract a bunch of videos for me. But there are also a bunch of other options. I could do profiles. Okay, so I could actually add in a username and then I could scrape all the videos and profile data from people that um that person follows or something like that. I could do um a direct search. I could actually enter video URLs. And then I could also download Tik Tok videos, too. Pretty crazy, right? Lots of options here. Anyway, I'm just going to try scraping, let's say, 10 videos per hashtag AI automation. I'm going click save and start. Now, what's happening is this is obviously spinning up that server like we talked about. But this is probably one of the best actors on the Appify platform. So, you get a lot of data and all of it's really relevant, which is nice. All right, so what did we end up doing? We got a big list of data here. We got the author. We even got the author's avatar profile pick, right? We got the text of the post. We got the digs, the shares, the number of plays, the number of comments. We got the duration in seconds. You can see that being valuable. We even got a bunch of metadata about music. But that's not all. If I click all fields, we got even more data. Okay, so now we have things like play count. We got metadata related to the music. So if there's a song, we actually know the song that was used. You can use this to like scrape like really good songs to use in videos, whether it's pinned, slideshow, the location, um the ID of the post, uh the various h the other hashtags that they use. So maybe you use a automation as like your primary hashtag, but then you scrape all the other ones and then search for those. Again, the rabbit hole goes pretty deep here. Um, but anyway, so now basically the thing that's important for us is we have the uh we have the profile URL. We also have the name here. Uh, and then we also have a bunch of information about the person. So, you know, join the AI ranking community, join this, join that. You know, we have a bunch of data that we could theoretically pass through some sort of system to generate um an icebreaker for, right? Which is pretty cool. Once you have a list of all of that, you can then search profiles. Okay. So, I'm just going to go back to that run and I'm just going to show you guys an example. If I wanted to run Nikico AI ranking, then extract people that rank for hashtags like a automation and then I wanted to extract a bunch of detailed profile data about them. Okay, this is what that would look like. Let me just delete this. And now I'm only scraping the video. Uh profile itself. So now this is going to be spinning up a server just like it did before and every other time and it's going to be getting us a bunch of data. All right, after you get a big list of all of the authors, it's kind of a clever and hacky way of doing things, but what you have to do is you have to connect chat GPT to this list. Then what we're going to do is we're going to feed in all of this author information into um chat GBT and see if we can have it return us a website. Let me show you guys what that process actually looks like in practice. So, first thing I'm going to do is I'm just going to use the get data set items. And this isn't a make. com video, but I just want to show you guys that it is actually feasible. We're going to get data set items from this data uh Tik Tok data extractor. So, I'll go to storage and this data set ID is right over here. I'm going to feed this in. And basically, all this is doing is this is giving us access to all of that data that we just had earlier. Okay. Now the field that we really care about is under author meta and it's the name and then maybe the nickname as well. So what I'm going to do is I'm going to grab all of that data. I'm going to feed that through an open AAI API call. So I'm then going to call OpenAI. The specific model I'm going to call it's going to be a search. This search model essentially allows you to look up results. So what I'm going to do is I'm going to use a developer/system. So you are helpful, intelligent data extraction and search assistant. Then under message, I'm going to go roll. Your task is to take as input a um Tik Tok profile JSON and look for and return a website for that person in JSON. It should be their either their company site or their personal site. Use this format. Then I'm going to say website. If no website is found, leave it blank. Only return your results in JSON. Your first character should be this. Okay. I'm then going to let's just go down to show advanced settings and go JSON object parse JSON response. I'm not going to do any of these other settings. What I'm then going to do is between this I'm just going to go transform JSON. What this will do is this will allow me to get this object which I care about which is author meta and just turn it into a big JSON string. The value in there is I'm just going to be able to get all of the all the data about that person which should help my search. So I'm then going to do text content JSON string. Then we'll just see if we can get the results. And I'm only going to return for now top five results to show you guys what this looks like. Um, looks like JSON objects not supported. Oh, my bad. You can't actually use Yes, that's why I um hardcoded it. Uh, you can't actually use this JSON object as a um, yeah, you actually just need to go empty there. Okay, now we're going to run this. It's now going to perform a bunch of searches for us under behind the scenes and sort of under the hood. Okay, so if it can find the website, it will. Now, in this case, it's finding school's website because this person had a specific offer that went back. This person did not have a website. This person had a school website because they had a school community. This person actually found their website which was nice. Then this person uh looks like just actually get the output of that. This person had their website as well. It looks like out of I don't know five profiles we got two. What you do is you then go to a Google sheet and then you just add that data. If I go to sheets. google then go down to Nick I leftclick and then make a new Google sheet. I just call this Tik Tok data. Then here I go website and then I go profile uh I don't know Tik Tok profile. Then I go back to my make. com scenario. If I then choose that Google sheet, I'll actually show you guys what it looks like when we fill it in. Just go sheet one table contains headers and the website will be the parsed version of this. So I'm going to parse this out. Feed in that result. Then from this, what we want is five. json. Cool. And then Tik Tok profile for that specific record would just be Nico ranking. Let's see what that looks like. As we parse it out, it's going to again perform those searches. And then if I open this up, you can see is we're now finding the website for the people that have websites. So we didn't find one for Automation Universe. We found one for Nathan Hodgson. And then we also found one for Callum Carver. And you know, as we see, we've gotten two out of the five. So about a 40% enrichment rate. You could also obviously improve the quality of the prompts if you wanted to do things like, I don't know, tell it not to put school communities in there because school's obviously blowing up right now and that's a big place that a lot of creators are monetizing instead of having their own website. But yeah, that is uh more or less it on Tik Tok. I

Outro

went through all applicable social media profiles today. A lot of fun putting this video together. If you guys have any questions about how to get up and running and scraping your own social media platforms on your own, let me know. For now, I'm going to hastily go through and uh remove my credit card from a bunch of trials and then also cancel a bunch of these because I just wanted to show you guys what this looks like and I was happy to pay a little bit of money to make sure that I could. If you guys like learning about scraping and the idea of getting an automation business up and running that does this sort of thing professionally for businesses, definitely check out Maker School. It's my 0ero to1 road map to show you guys how to launch and monetize your own AI automation business. Uh, if you guys already have an established company and maybe you want to take it to the next step, check out my community that shows you how to take something that's working and then scale it up even more. Thank you very much for watching. If you guys could do me a big solid like, subscribe, do all the fun YouTube stuff. I'll catch you on the next video. Bye.

Другие видео автора — Nick Saraev

Ctrl+V

Экстракт Знаний в Telegram

Экстракты и дистилляты из лучших YouTube-каналов — сразу после публикации.

Подписаться

Дайджест Экстрактов

Лучшие методички за неделю — каждый понедельник