← The Dave Gerhardt Show (from Exit Five)
The Dave Gerhardt Show (from Exit Five)
How to use Claude Code to Scale LinkedIn Ads with Anthony Blatner (Founder, Speedwork)
00:37:30
Loading summary
Transcript73 lines
- [00:00]
Dave Gehard
Hey, it's me, Dave. Today's episode is brought to you by webflow. Webflow is a website platform for the agentic web, built for modern marketing teams to build fully custom sites that perform an AI search. No developer needed. Your website isn't just your homepage anymore. It's actually the first thing an AI agent is going to read about your company and from there decide whether or not to recommend you to a potential customer. Before a demo, before a sales call, before anyone on your team even knows that buyer exists. Which means if your site isn't structured to show up in AI search, you're not in the buying conversation. It's as simple as that. Webflow is the platform marketing teams use to fix that. You can build a fully custom site without a developer, optimize it for AI search, govern your content, and actually see how agents are reading and interpreting it all in one place. Make your website your biggest growth engine right now with Webflow. Check them out@webflow.com Exit 5. That's webflow.com for Exit 5. Hey, it's me, Dave. Today's episode is brought to you by Markup AI. AI has made it really easy to produce a lot of content. But producing content fast and producing content that's publish ready are not the same thing. By publish ready, I mean it follows your brand guidelines. It sounds like you, and it's a good fit for your audience. I'm seeing a lot of marketers right now running their content through a bunch of different AI tools trying to get to a final product. But by the time it comes out the other end, it doesn't sound anything like them. It just sounds like AI. And that's the AI slot problem. And it's why there's a growing sense of frustration with AI content right now. Markup AI fixes that. It sits in your content creation, workflow, scanning and scoring your content against your brand, voice, preferred terms and messaging standards before anything goes live. Markup AI even checks for AI visibility, best practices, and works directly inside Google Docs so there's no editing, extra tool, or handoff. Grammar was never your hardest problem. Well, speak for yourself. The real risk is publishing AI content that's generic, off brand or unsitable. True Markup AI helps you catch that first. So you're always putting your best stuff out there. Go and check it out at Markup AI. That's Markup AI. You're listening to the Dave Gehard Show.
- [02:10]
Matt
Exit. Exit. 12212. All right, so today I'm speaking with Anthony Blattner. Anthony is the founder of Speedwork Social, and his company essentially helps you with anything LinkedIn related, whether it's paid, organic, outbound, whatever it might be. But today, specifically, we're talking about LinkedIn ads and AI. We're talking about how you can use cloud code to connect to the LinkedIn Ads API and automate a lot of analysis or optimization. We talk about how you can personalize campaigns at scale. We talked about things like competitor conquests, and also how to create LinkedIn ads that are not just a bunch of AI slop. I think you're really going to like this episode. It's super tactical. We got right into the weeds of it, so hope you enjoy. All right, I'm here with my favorite LinkedIn ads guy, Anthony. How's it going?
- [03:09]
Anthony Blattner
Hey, Matt. Glad to be here. Always glad to talk about LinkedIn ads with you.
- [03:13]
Matt
Yeah, a hundred percent. You know, it's funny, today I. We're just talking about this. I wore my Texas longhorns hat, I got my bill signed in the back. And you're based in Austin, originally from Buffalo. You're a Bills fan. So we got all the good right there. That's it. That's it. I love it. I love it. Okay, awesome. Super quick. I'll give you, you know, 30, 60 seconds to introduce yourself, but you've been part of the Exit 5 community for, like, two years now. I think it first started with you coming on the podcast, joining the community. You've been to DRIVE both times. You've been to our Austin events. You've been helping Exit 5 with our ads for maybe two years now and, like, crushing it for us. Like, it's been a massive channel for us from a growth perspective. So, yeah, I'll just tee that up. I'll let you introduce yourself, but, you know, you've been super ingrained in this community. That's been awesome to see. I'm excited to get into some LinkedIn ad stuff today.
- [04:00]
Anthony Blattner
It's funny, I just posted today about the speaker announcement for DRIVE this year. So, number one, I'm very excited to speak at Drive this year. We're going to cover a lot of good topics there. But in that post, I talked about how, yeah, maybe 2024 or whatever year I got, I joined Exit 5. I was watching Dave's podcast and going through past episodes, and I was like, he doesn't have one on LinkedIn ads yet. Like, a B2B marketing podcast doesn't have his episode on LinkedIn ads yet. I gotta reach out to him. So reached out. We did a podcast together and it was fun to deep dive on that. And I remember on that, like, after the recording, we stopped recording and I was like, all right, so when are you gonna do local chapters? Like, Austin's got a great community here, blah, blah, blah. He's like, hold that thought. You'll hear more very soon. So then you guys did the event, like the panel event in Austin that was really cool. And then there's been the local chapter events, and the Austin group's been putting together some really good events. We did a rage room, I think was the first one. Very unique, but good team building activities. So, yeah, it's been part of Exit 5 for a while and it's been a great community.
- [05:05]
Matt
Yeah, it's funny, like, quick side note lesson before we get into it, like, you just saying, you look through the episodes and then like, pitched doing an episode in LinkedIn ads. Just like a quick little lesson on if you want to pitch yourself or anyone to get in a podcast. Like, actually just look at what the podcast is about and look at past episodes and position yourself as someone good to fill a gap there. Like, I still remember your email to this day, which is funny. I think you said, like, hey, I noticed you guys have no episodes on LinkedIn ads. I'm one of the nine certified LinkedIn ad experts in North America or something like that. And I remember seeing that and like, we get a ton every week. And I was like, oh, like, this makes a ton of sense. And I was like, hey, Dave, we should do an episode on this. And we just ran it from there, so sometimes it's just that easy. And look at. Look at where we are now.
- [05:51]
Anthony Blattner
So covered a lot of ground since then.
- [05:53]
Matt
Yeah, yeah, yeah, Exactly. Cool. So, yeah, give us the quick 30 seconds on you, speedwork.
- [05:58]
Anthony Blattner
Yeah, quick 30 seconds on me. I'm based in Austin, Texas. I head up Speedwork social. We're a LinkedIn agency, so I spend way too much time on LinkedIn, but I love it. So I'm mostly in the ads platform myself. Speedwork is an agency. We also do organic content and outbound and all that good stuff. So cover all parts of the platform have been focused on it for a while now. So, yeah, I was part of that original certified marketing expert program. There's like 10 of us in North America. And then I've created the LinkedIn learning courses on LinkedIn ads, so done everything all across the board, and it's a very exciting place to be in.
- [06:32]
Matt
Cool. Awesome. Yeah. So today I think we're going to talk about a couple things. I think one of the first things you wanted to chat about was like how to use cloud code to connect to the LinkedIn ads API. So tell me a bit about like why you want to chat about that, maybe some benefits that people can get from it and then we can talk about how to actually do it.
- [06:50]
Anthony Blattner
Yeah, so it is very exciting times these days. Claude code now lets you as a marketer talk to any API out there. So one of the first big unlocks and we've been working with the API for a very long time and we've always known that the LinkedIn Ads API offers a lot of data that you can export from the API and get in a very granular level to the level that gets you more data than you can get in the UI. If you're just a login to LinkedIn ads, go to Campaign Manager. It gives you more of a summary view in a lot of those. So like for example the demographics and the performance reports, you get a lot more granular data when you pull it from the API. We've known that for a long time. And now the exciting part is now using Claude code to go build your interfaces, build your web apps, do this analysis for you, really simplifies that process. In the past you had to build your own web app, obviously now claude code can do a lot of that. So it helps you automate and do a lot of this to the next level. So cloud code has opened up a lot of new capabilities. You know, using an MCP is one level where there's some pre built MCPs out there that can do some different functions. But as with any API, if you use the API itself, not just the mcp, you get access to every endpoint that it offers and you have the full range of capabilities. So very exciting times, like with any API, once you use that, you get a lot more data. So we've been plugging in Claude code to do a lot of these automations analysis and really like what that turns into is how can I go build maybe a lot of campaigns or ads and launch them using the API? How can I do bulk analysis, bulk management in a way that that's going to get me more information to make those advanced optimizations and then implement it so that we have a cost savings in the end for our advertising? Yeah, so that's like the summary of it. Now how do you go and actually do this? Of course, number one, you need cloud code and you got to be comfortable using that. And then I won't go into full detail on how you build it, but I have a video on our YouTube channel that walks through how you actually can build it. So go check out our YouTube channel. It's under Speedwork Social and there's a Claude code one right at the top there. So go watch that. You'll see how you can build your API interface to Talk to the LinkedIn ads API. The short and quick is you can go download all the documentation of the API, give that to Claude code and say, hey, build me this interface and basically it'll do that.
- [09:10]
Matt
Okay, cool, cool. And what's the like you said, because I think one of the interesting pieces about this is like, so automating some regular optimizations and also like some bulk analysis even like to this day. But I know like in the past what I would always do is like once a week I would just do a sweep of like the metrics inside of the UI and just say, okay, this is happening this week, so we're going to do X because of it. So now it looks like at least like 80% of that work is going to be whether it's automated or at least like you could surface it up to yourself to make some recommendations is what it seems like. Tell me what that end product looks like on like a weekly basis.
- [09:46]
Anthony Blattner
Yeah, so a lot of different things that you can implement. Maybe I'll give a couple examples that'll help. So you know, if you're managing ads on any platform, bid management is a big part of it, where your bids are typically changing over the course of a month, a quarter, a year. Especially on LinkedIn, there's lots of seasonality that will happen. So you want to be adjusting your bids so that you're ideally spending the budget you intend to spend and spending it most cost efficiently. So over the course of a month, you're on average, prices tend to rise over the course of a month. But also you might be launching new ads and you might be getting better engagement rates and then you can bid lower when you have that. So it's all to say in the end it's good to do like a bids optimization maybe once or twice a week. So that's an example of something that will implement for a lot of clients and it'll just kind of run in the background twice a week it's going to go analyze. Hey, are there any opportunities to lower bids or are we underspending anywhere that we need to raise bids? Because maybe we're getting towards the end of the month, prices have gone up or Maybe we launched some new ads that are performing well and we can lower bids and we get some cost savings out of that. So a bids optimization, just have that kind of run as a workflow in the background. And then it's great if you have like a human in the loop step where, you know, first I want to be able to approve those because maybe I know something else that the algorithm doesn't know just yet. We just launched something today that I just want to be careful not to change it yet, or I know something's coming out tomorrow, something like that. But bids analysis is a good one. One other one is companies analysis. Well, a demographic analysis where also maybe on like a weekly basis, report back to me, who are the companies being reached by our ads, what are the job titles, so that if there's any deviation from what I'm intending or any changes in that, I can catch that quickly and get a campaign back on track.
- [11:29]
Matt
Very cool. Okay, those are both super useful. How long does it take to set that up through cloud code?
- [11:35]
Anthony Blattner
Pretty easy, pretty fast. So for a lot of these, you just need to know the right API endpoints to use and you just tell cloud code, hey, go look at the API documentation to figure this out. And it's always funny to watch, like, Claude code start to learn something where it's like, it's going to fail. It's going to fail. And then like, oh, yeah, I figured this out. It's a different format of the parameters and then it learns it. So we've built lots of different skills and workflows to do this, and that kind of all gets it all baked in like that. If you're doing this for the first time, start with API documentation. It all depends what you're building for your company and your account. So it might be a little bit different, but cloud code's pretty good at, like, learning that. And then you just tell Claude, like, hey, what do we need to document based on what we just did so that you know the right formats and stuff to use? Next time it'll save it to a memory file or a Claude MD file, or even better, just bake it into a skill that you're going to reuse over and over again.
- [12:24]
Matt
Very cool. Okay, so yeah, it's like, in short, you're essentially creating like a little, little intern that is just like analyzing your LinkedIn ads, doing a sweep of it once a month, telling you how it's performing. Or not once a month, I guess, however often it's going through it, but telling you how it's performing, telling you what you could potentially change, telling you who you're reaching through your ad. So it's just surfacing up the best information and then you can make the decision on like, here's what I want to do about it type thing.
- [12:50]
Anthony Blattner
Yeah, anything that you think that you find yourself doing, like anything that's like tedious or manual or like you have to repeat doing a lot, that's a great opportunity to go, hey, build a workflow, build a skill or something that will do it for you.
- [13:03]
Matt
Okay, cool, cool, cool.
- [13:04]
Anthony Blattner
This is all of marketing, not just LinkedIn ads, but it's great when you use it. Save your. Save yourself money.
- [13:08]
Matt
Yes, yes, exactly, exactly. I know, that's great. That's a great tip. Good way to kick this off. You said something which was like one of the use cases was looking at what companies are engaging with your ads and being able to decide whether those are the right companies or not. I think this segues us nicely into the next thing we wanted to talk about was like, how can you get a bit more segmented and personalized with your ads using AI? Personally, in the last six months I've spoken to hundreds of CMOs as part of our CMO community and ABM and targeting super specific people and enterprise companies, for example, is a big focus for a lot of people right now. So yeah, we'd love to just kind of dig into where you've seen that work and some, some tips and tricks you have for that.
- [13:52]
Anthony Blattner
Yeah. So this kind of combines ABM as a strategy with economies of scale using LinkedIn. So the marketing principles that the more you can closely align your marketing materials with who you're targeting, the more residents is going to be there and every ad algorithm is going to give you a cost benefit when you have that. So like LinkedIn evaluates primarily based on engagement rate. So the higher the engagement rate you can get of your ads, your ads matched up with your audience, the better cost savings you're going to get on your campaigns. So this is the kind of the basic principle, how do I save money on my ads? Get your engagement rate up and then you can bid lower and save money. So now we're thinking about how can we get further segmented on what we're doing. And a lot of this is based on what is the production limitations from a company. They usually need to create landing pages, they need to create ads, they need to write ad copy, they need to set up the campaigns and they need to analyze those campaigns. In the past that would work up to like you know, 5, 10, 25 campaigns or so, then you kind of really have a lot of moving parts past that. Nowadays you can have Claude code build out a lot of this for you and then also help do the analysis. So really we're kind of like smashing through the production barriers where now we can set up some templates and then have cloud code go build those out for us and then boom, we can replicate that and scale it, you know, from that three to five to easily 20, 50, 100 or more. Just all depends on your budget and how many companies you have to reach, how many industries or how many Personas you have to reach. So that's like the very basics of the strategies. Now let's talk about kind of what that looks like in practice. So very often we build out our marketing matrix in terms of whatever the approach is. Let's say we're doing a one to one ABM campaign and you have a list of companies you want to be reaching. So for each of those companies you have what's the targeting I need to set against that? And also like what's the campaign I'm building that's going to reach those people? What's the ad copy going to say, what's the image going to look like, what's the landing page going to look like? And then is there any kind of final offer or downloadable that they're going to get by going through that? So lots of different ways you might set this up. You might have a demo request funnel or you might have a personalized report funnel. Both of those can be two good examples you might use. And you know, in the past you might have had like a finance report or a free trial for finance companies. Now you can take that just another level deeper. And maybe a lot of times we're thinking about what's the Persona and the industry that we want to be reaching. And then we're going to segment into all those that the client wants to serve. And again, also like an ABM approach at a company level, you can have, you can have a specific landing page that speaks to that specific company, you can have images, visuals and copy all that speak to them. End of the day we're just trying to build that resonance between who are we targeting, what are we offering, that's going to improve our conversion rates and that's going to improve the cost that we're getting out of LinkedIn. So very often you're going to start with like a template and you're going to probably write this yourself or you're going to use Claude to like generate that first template. Once you have that in a place that you're happy with, you're going to then have Claude go replicate that across all these different segments and kind of scale it out for you from there. And then once we're using the API, we can have the API go build those campaigns, build those ads for us, set up these campaigns, and then get them all launched in parallel. Now, what's the limit to this? Well, LinkedIn's limit is 10 bucks a day per campaign. So you just need to do that multiplication to see what kind of budget you can support here. That said, if you have a lower budget, you can cycle campaigns. So maybe you have one week this industry or this set, and then you swap the next week. Sometimes we'll do it by days of the week. We'll have odd days and even days, and we'll flip flop campaigns if we're trying to fit more into a certain budget level. So that's kind of a good way to use it to scale out your approach there and then be able to achieve a lot of cost savings and then conversion rate optimizations by doing that at scale.
- [17:44]
Matt
Gotcha. Okay. Super cool. So essentially, from what I'm understanding, like the premise of this is like, you know, a lot of companies have multiple segments that they target multiple industries, whatever it might be, right? They have a handful of Personas. Like, for example, like at a previous company of mine, we sold software to small business owners, and they're all small business owners, but it would be like plumbers, electricians, chiropractors, doctors, dentists, et cetera. So from my understanding, especially for companies like that, probably for many, but especially for companies like that, now we can have offers, essentially ads copy that's super targeted to all those different Personas, and we would break out our campaigns by Persona and then launch it from there. Is that generally like what this is?
- [18:30]
Anthony Blattner
Yes, basically.
- [18:31]
Matt
Got it.
- [18:31]
Anthony Blattner
So you'll have your ads that have an electrician in it and you'll put that in the visual. You'll have one that has a doctor in it, you'll have one that has a dentist in it. And then that's going to be broken out between all those. And you know, you could do this in the past and it'd just be very manual to do and maybe you get yourself up to like, like I said, like 25 or so. And then it's kind of really hard to go past that. So you can use AI to go scale. You can do a hundred of these very Easily, yeah.
- [18:55]
Matt
I think it's like the entire funnel now you can just use Claus to do the entire thing. Like, obviously you want to apply some of your own intuition. Company data that is important, that's going to influence that. But like the other day, for example, and lately I've just been playing around more with clogged design and it's actually like, even if you don't go into design, it's really good, but if you do, it's like ridiculously good. So to your point about if it's an ad for electricians, having an ad that is specifically for them, you could literally grab a stock image of an electrician and then give that to cloud design. And within your branding and the copy and text you want, you could just create the ad image right there.
- [19:34]
Anthony Blattner
Yeah, exactly. So I like in Claude code, you can set up directories and stuff. So you'll start to build out your folders. They're basically okay, all the segments I want to be reaching. I'm going to have a folder for each of those. I'm going to start with my brand guidelines at the top line. Some people have those and people don't. If you don't just tell Claude, go build you some brand guidelines and then it'll go reference that as it builds out the rest of your assets there. So, for example, like landing pages, there's always like a lot of work that goes into those. So I like to build one really good one iterate on that design. And then once I got into a good place, then I can say, hey, go replicate this landing page for all these different industries. And it could be industries, it could be target companies, they might just be swapping in logos into all those target companies. So that's very good for just on the landing pages and then also for the ad level too. That's like one of the best ways to get some of the highest engagement rates that you can is put that logo into your ad or write it out. You know, write out that company name in your ad just to maximize all that.
- [20:30]
Matt
Very cool. Okay, this is great. I feel like one thing that does get overlooked sometimes is just like the offer or lead magnet, right? It's like it'd be pretty easy to. Not easy, but, you know, you can go create a different ad for each of these Personas. Right. And we know how to do that. Like as marketers, it's like we know how to write the copy that we need to write the imagery, et cetera. The targeting is, you know, fairly straightforward. I'd like to Think, but talk to me a bit about like the offer and lead magnet. Like, how can you personalize that? Should you personalize that part? Like, let's walk through that a little bit.
- [21:04]
Anthony Blattner
Yeah. So a lot of companies will have like industry based reports. So whether you're doing an industry based report or there's a lot of examples that we've been running lately where it's like you're going to do a report for a company and it's already generated, you're going to pre generate it and then the ad is basically get the report for your company and it is impressive when you can deliver that right away to them versus, you know, a lot of people think like, oh, I'm going to submit this and then like your salesperson is going to get back to me in a couple days. But when you deliver it to them right away, that's kind of an impressive workflow. And they're like, all right, I got some immediate value out of this. The longer you wait to get that to them, the more they're going to forget about you, all that. So the more you can instantly give them that report. Boom. That makes it a lot warmer for them. They got some instant value out of it and then that's going to be more likely to take next steps with you.
- [21:50]
Dave Gehard
Hey, it's me, Dave. Today's episode is brought to you by Optimizely. Optimizely is a platform built for modern marketing teams, helping you create content, run experiments, personalize experiences and optimize your website. All powered by agentic AI. Most of us are using AI to write faster copy emails, briefs, and honestly, that's fine, it's useful, but it's also kind of the boring part of what AI can actually do for a marketing team. The shift that's actually worth your attention right now is using AI to handle coordination inside your team. Not just content creation, but the routing, the flagging, the context tracking that eats up everyone's week. There's a better way, and that's what Optimizely's agent platform is built for. With everything going on in AI right now, building agents into your marketing workflow is one of the most useful places you could be putting your energy. And optimizely is a great tool to help you do so. Plus, it's no code, so every marketer can create agents without developer dependencies. So if you're agent curious but not sure what the next step actually looks like, optimizely put together a resource that gets into real examples of teams using agents and what it could look like for you. Go and check it out and learn more about Optimizely@Optimizely.com exit 5 that's Optimizely.com exit 5 hey, it's me, Dave. Today's episode is brought to you by Walker Sandstone, an integrated B2B marketing and growth services agency. Here's something I see happen all the time in marketing. The team is busy pumping out campaigns, shipping content, managing channels. The list goes on. But when you zoom out and look at all that activity together, the results don't add up like they should. It's not always obvious how the daily work connects to business growth. That's the challenge that Walker Sands solves. They help B2B marketing leaders get clear on what actually matters business outcomes. They work with you to build a go to market strategy around your desired outcomes with the right messaging, the right channels and measuring the right impacts so your team's work ladders up to that big goal. Walker Sands has been helping leaders turn marketing strategy into measurable business impact for over 25 years. And they work with brands like Amazon Business, Globant and Commerce Tools. Their whole outcome based marketing model is built around the concept that brand demand and revenue are one conversation, not one. Three separate silos. I love that. If your marketing feels busy but a little disconnected from the bigger picture, Walker Sands is worth looking into. Go check them out@walkersanz.com exit5 to help your business drive measurable growth. That's walkersanz.com exit5 yeah, got it.
- [24:21]
Matt
I like that. And even another thing with this, so like even if you didn't drastically change the offer, like if you have an industry report for example, like that would be ideal. But let's say you're a company where you don't have specific content for each industry. And let's just say your offer lead magnet is the same. It was what you were saying earlier. Like you could still personalize the landing page itself to be more tailored to that person to make it think like it's for them. Even if like the end thing they get in their inbox or the action is the same as it would be for the other Persona. I really like that piece too. I think like even just using any version of Claude, like you can so easily whip up a nice landing page. You get the code for a nice landing page like in a minute or two maybe. So it's like it's kind of scary fast.
- [25:07]
Anthony Blattner
How fast it is?
- [25:09]
Matt
Yeah, it's scary fast. Like it's, you know, back in the day you'd have to go find page. I mean, you still have to do this. So it's not like you should skip the work, but it's like you'd have to go find pages you like, maybe look at some elements of them, screenshot them, and then you'd have to, in a document, write out all the copy you want. And then eventually you take all that, you give it to a designer and you say, hey, can you create like a V1 this for me? And you have to wait, and you have to wait for the designer. And then when they finally send it, you'd have to give them more feedback. And it was this process of like, okay, a week later your page is out. Right now it's like the beginning part's the same, but where it changes is in the end, the feedback is just instantaneous. So if you can literally go create a good version for one of your Personas, you can create 10 landing pages in under an hour, 30 minutes for all the other Personas you have and then just pass those to your ads, right?
- [26:01]
Anthony Blattner
Yeah, exactly. And now let's, let's talk a little about how to do this. How to produce like good assets and not just like AI slop assets. Yeah. Because, you know, sometimes listen to this and be like, great, I'm going to go have Cloud go do all this. And then like, you know what it outputs, it's like, okay, that's pretty good. But I think there's like another element that a lot of people now can like look at something and be like, that's AI generated. Or like read something and be like, that's AI generated. And there is new research out right now that people don't necessarily want to be reading AI generated stuff. And I'll also, you know, in the LinkedIn feed, you see lots of posts and comments that kind of are clearly AI generated and they always perform worse. Because when you're reading something on LinkedIn, somebody doesn't want just an AI generated post. They don't want to just read what a machine output. They want to hear from other humans. So you have to kind of keep this in mind as you're creating all your content. AI can produce some really cool looking stuff, but we're starting to see what are the trends and the design trends and the writing trends that I'll kind of give it away. So you want to avoid that. Number one, you can go do a little bit of research. And a lot of people have posted like prompts and stuff like that that help get around a lot of that. But what I want to talk about is like the content itself. So if you're thinking about a one to one ABM campaign, or if you're thinking about an industry based campaign or another super common one we'll do is competitor conquesting where you have a series of competitors. You might have 10 competitors that are your most common ones. You're going to build a campaign like this and target the companies you know are using each of those competitors in each of your campaigns. So this is another good kind of at scale ABM approach is the competitor conquesting one. So some of the best sources you can use to plug into your content generation are your CRM data. So what history do you have there for like for example, HubSpot saves a lot of good information there. Emails, phone calls, stuff like that. You might be using a call recording software like Fathom or another platform like that and you can pull in transcripts. So now we're taking that matrix that we were talking about for ad landing page, all that and we're going to start with what's the inputs that we want for each of these. So you want to organize your Fathom call transcripts by whether it's by competitor or by industry that you're targeting. And then you want to organize your HubSpot data that you're pulling out. Like emails are great, you know, going back and forth, talking with people, that's all good stuff to put into your Claude so that it's going to generate really accurate content to feed into these campaigns. So back to your Claude directory that you're building in Claude code. You're going to have a folder for each one of these and then you want it to. One of the building steps is to have Claude go organize all that data. You know, a lot of this is like too much for one context window. So it is important to kind of save your work in the steps that you're working through here, where you might have several different Claude chats open and certain ones are building your directories and your knowledge base here. And then by the time you get to generate those reports, you want it to go access each of those directories so that I can pull in the associated information and create the best asset for you. Because if you just do it all in one chat, you don't save anything, then it's not going to have enough context to really build you good assets. So important to break this down into kind of baby steps here where you're going to have your kind of setup steps where you're building out that matrix, you're building out your directories and you're organizing your content so that when you finally kick off Claude to go create all your reports, for example, it's going to follow the template, but then it's also going to go to each of those directories and pull in the associated content for each one of those segments and then it can build the best asset that way.
- [29:35]
Matt
Very cool. That's incredible. I love that. Do you have clients here you're running this with right now?
- [29:41]
Anthony Blattner
Yeah, starting to roll it out to more and more of them. And one hand depends on budget level. The more budget you have, the more campaigns we can spin up, the more content we can create. But once you have a good template, it's very easy to roll out a lot more.
- [29:53]
Matt
Gotcha. Okay. Yeah, that was going to be one of my next questions. So there's a lot to unpack about what you just said for sure. But one question I had was like, there's a varying range of people who listen to the XFI podcast. There's a people who have a lot of budget, there's people somewhere in the middle, and there's also just in the marketing world, a large percentage of people that don't have a smaller budget to play with. So do you think something like this is viable for somebody with a more modest budget? And if so, how do you execute something like this if you are in a more modest budget? Like, where should you prioritize?
- [30:26]
Anthony Blattner
Yeah, that's a good question. There's very little downside to adding more segmentation. You know, more segmentation almost always is a good thing to do for the smaller companies. Like, you probably have an idea of like, who's the best industry you want to be going after, who's the best Persona you want to be going after? You know, it's a marketing principle of if you go to everybody, you're talking to nobody. So, you know, instead of using that big broad audience, you want to be very specific on who you're going to reach and what you're going to say to them. So if you're just getting started, another thing I'll throw in here is thought leader ads are still a good tactic to do with this. And if you've heard any of my other content or listened to any other LinkedIn ads episodes, like, it's all about thought leader ads these days. So you can now also, as one of your ad elements, it can be a thought leader ad. And if you're planning to build your matrix or whatever, it's going to be if you're a small company you might have three and you might just have three different posts and then each of those posts gets boosted to the associated industry. So that thought leader ad will get you the best engagement rate and then therefore the cheapest cost. And then you're still building the ABM approach, but you're going to do it with thought leader ads and then that'll help you get even more cost savings out of it.
- [31:34]
Matt
Love it. Yeah, even one thing that could be interesting too is like if you are a company who has many Personas and is interested in executing something like this, like is there a world where you just, you try this approach with one, right? It's like let's just take one, one of our more popular ones and just say we're going to go create these hyper specific ads and copy and land pages and offers and if this works well and does what we needed to do, then we're going to go replicate that with the next one and just that could be our how we vie for more budget too.
- [32:06]
Anthony Blattner
Yeah, definitely. Some companies have like hey, here's the one company we want to be going after this quarter. I'd say on average I like to like start with at least three. Kind of gives you like a sample set there. Maybe maybe that one company you thought was going to be the top one. Maybe you get a quicker bite from a different company. So I like to start with a small grouping. But yeah, start with a small grouping, test it out and then from there you can kind of scale it out to the rest of your list afterwards.
- [32:27]
Matt
Very cool.
- [32:28]
Anthony Blattner
And then I should also mention as you start to scale all these out, say you have 50 campaigns running. This is also where some of those other workflows I mentioned become more useful. When you have 50 campaigns it is very difficult to manage bidding and budgeting on a regular basis. So this is where you would want Claude to go, hey, do a scan twice a week. Other places where I can save on bidding, save on my cpcs and other places where maybe my demographics are on point or off point and it can give you some feedback that way. So as you go to scale things out, this is where other workflows get more important. Bidding and budgeting and then demographic review are like two of the most common ones.
- [33:01]
Matt
Okay, cool, cool. On your guys end, are you like from an image creation perspective, is it mainly cloud code you're using or are you still, do you still have like a design team on the inside that's like creating some assets? How are you doing that part of it?
- [33:16]
Anthony Blattner
Yeah, a little both. Kind of depends on the client we're working with, but have been using AI to generate some images, Been using Nano Banana a lot for generating the images. Usually we'll give it the template frame we want to work with. And then if we're putting a list of companies or if we're putting a list of logos over it, put that in there and then have that generate the images and it's pretty good.
- [33:37]
Matt
Cool. Cool. Awesome. Honestly, I feel like that was pretty solid. This is just a side note now, but I think that's pretty solid. I don't know. Do you feel like there's anything else, like, important that you want to cover that's pretty solid?
- [33:47]
Anthony Blattner
Yeah, pretty good.
- [33:49]
Matt
Yeah. Because, like, I figured if we can just make it 30 minutes versus 60 and pack a ton of value into it, like, I. I don't see why that would be a negative thing in any way.
- [33:57]
Anthony Blattner
Yep.
- [33:58]
Matt
Cool.
- [33:59]
Anthony Blattner
Maybe we can talk about list building real quick. Using Claude to build some lists that you can't use from other systems. That can be another kind of tip and trick.
- [34:07]
Matt
Okay, yeah, let's get into it. Cool.
- [34:09]
Anthony Blattner
All right, so other topic is targeting. So, of course, building your target lists is one of the key parts to getting started. Sometimes you have a very clear industry you want to go after, and, you know, maybe Apollo or Clay has that industry and boom, very easy for you to build that list and export it. I'll even say, like, one of the nice things about using a data platform is they usually don't charge you on the company level if you build a list, a list of companies, but they'll charge you on the contact level. So one of the nice cost savings for LinkedIn is build yourself a company list and then upload that to LinkedIn and then do the company list and your set of job titles, and then, boom, you're able to export those lists, usually for free, and then use LinkedIn to find the individual people. One other thing to note is if your target audience is something that's not a super clear filter, lots of companies out there are trying to reach maybe a certain type of company or a certain person, and that just. There's just no clear industry filter for that, or it's about something that LinkedIn doesn't have a filter for. So what are some examples of that for? I had somebody ask me about cannabis companies and it's like, okay, well, there's no cannabis industry on LinkedIn. You'd have to build yourself of your own target account list. So Lots of examples like that, maybe a piece of software they use or something else like that. So these are other cases where Claude code can go build you lists. Do be careful with your credit limits. It can eat up a lot if you're asking it to go grab a bunch of URLs and do a bunch of research for you. So best to build kind of the list of names and website URLs. If you're asking for more than that, it might spin up way too many agents to eat up your credits real quick. But it can be very effective for building you target account lists in some of the segments that maybe there's not a super clear industry or super clear definition of. So that's another kind of very interesting workaround we've been using lately because yeah, lots of times companies have like a certain customer segment and it's like, okay, there's no like specific industry for that. One thing we're looking to target in the past it's like, well, you're going to kind of have to deal with getting some bad leads and some good leads now we can get a lot more targeted with what we're doing.
- [36:02]
Matt
Right. So you're giving it. You're essentially using these tools, I don't charge you to build company lists and then those, you're just going to go plug those into LinkedIn and say these are the job titles at these companies that you're targeting and it's only going to go after those people.
- [36:15]
Anthony Blattner
Right, right, yeah, yeah. LinkedIn is the only platform that can do the company based approach like Met is all contact level. So yeah, yeah. If you're using Apollo or Clay, they'll usually charge you by contact, but you can export as many company lists as you want then yeah, if there's not a super clear industry in Apollo or Clay, like if you're just not finding what you want in Apollo or Clay, that's another good use case to go to Claude and say, hey, build me some lists for this.
- [36:40]
Matt
Cool. Very cool. Yeah, I love that. I feel like in the past like that was always one of the most difficult parts about just ABM in general is just like the list building part of it. I mean there's a whole side of like agreeing internally as to like what that list is and that's probably where like a lot of the work actually is on this front. But like it sounds like at least when like internally you've decided on what some of those companies are, what they might look like, then that's where you can plug this process. In and it's super easy to just grab some company names and target from there.
- [37:10]
Anthony Blattner
Yeah, exactly.
- [37:11]
Matt
Cool. Awesome. Awesome. Well, yeah, I feel like what I loved about what we talked about today is it. It just feels like super tactical stuff that people can go execute on tomorrow. Like, I feel like a lot of the stuff that I read with like how to do X with AI is like maybe not super practical or it's like I wouldn't actually go and do that. Like, when you actually go try and execute on it, it's not really what you would have hoped. And I think in this case, like, that could happen if you don't put in the effort to make it good. Right. So it's not saying this is like the like golden ticket to like crushing it on LinkedIn, but it's like if you put the effort into creating the good stuff up front, you can just scale your ad campaigns easier with less resources, really, and get a lot more targeted if you plug some of these tools in. Yeah.
- [38:01]
Anthony Blattner
And just to give people some bullet points is if you're able to take this and double your ctr, you can usually get your CPC in half. So you might save half your ad cost by doubling your CTRL by taking this segmented approach. So give yourself a 50% off discount on LinkedIn ads.
- [38:20]
Matt
Yeah, that's super cool. Yeah, I think that's something that like, people who aren't in the weeds of LinkedIn every day may not be considering as much as, like, the more you can improve the quality of your ads, it's actually going to be cheaper for you to run them. So it's actually in your best interest to invest that time in doing that versus, like just always thinking of net new. Right. It's always like if something's underperforming or not where you want it to be, it's like, how can I go and create new stuff? And sure, some of this is going to be net new, but like, the right way of thinking of it is like, how can we just make it better for that audience? Because with that we're assuming that it's going to perform better and cost us less in the end.
- [38:55]
Anthony Blattner
Yep, exactly.
- [38:56]
Matt
Love it. Cool. Well, yeah, super great tactical stuff. I feel like people can literally take this and go run with it and do a bunch of new cool things. I think, like, if someone were to start somewhere, like, obviously we gave a bunch of great stuff, but it's like, you know, into Speedworks, YouTube and just like go try and create that. What do you call, like your LinkedIn ads? Agent? Pretty much, yeah.
- [39:17]
Anthony Blattner
I just call my API agent and you just have it build you an API wrapper. And then from there you can just chat with that wrapper and it can kind of do everything for you. So it can do everything we talked about today, but then there's like lots of other, maybe bulk updates, maybe bulk renaming, bulk edits that, like, would take forever if you were just doing it yourself in the ui, but now the API thing can just do it all for you.
- [39:39]
Matt
Sick. Okay, that's very cool. Yeah, I think that's like a great next step for anyone listening is, like, if you can just go and create that, I think it helps you get your feet wet between, like, cloud code and LinkedIn ads. And then I think, like, once you do that and you've maybe automated or made your life easier, I think that's going to give you then the confidence to go and run some of these other things that we talked about today.
- [39:59]
Anthony Blattner
Yeah. And then you can start to do this for every ad platform and every other marketing system you're using. So really just kind of opens the door to, like, everything else that you could do.
- [40:07]
Matt
Yeah, exactly. And it's. It really is an easy one because it's like the LinkedIn data is all there and a lot of people have been running ads for many years. So it's like there's already a really great foundation to build this tool on. So I think it's a really great project for any marketer and seems like it wouldn't take super long anyway, so.
- [40:24]
Anthony Blattner
Yeah.
- [40:25]
Matt
Awesome. Cool. Well, Anthony, this has been awesome. I know you're, you know, super involved in the Exit 5 community and on LinkedIn, et cetera. So I know anyone can go there and hit you up if they have some questions. You've always been a great resource, Tags of five, like our team, internally, the community, and whenever ask me something, I always know I'm getting a really good, trustworthy answer in return. So I'm sure you'll do that for others. And yeah, I appreciate you coming on today and we'll see you around the community.
- [40:49]
Anthony Blattner
It's all good.
- [40:50]
Matt
Awesome, thanks.
- [40:55]
Dave Gehard
Hey, thanks for listening to this podcast. If you like this episode. You know what? I'm not even going to add, ask you to subscribe and leave a review, because I don't really care about that. I have something better for you. So we've built the number one private community for B2B marketers at exit 5. And you can go and check that out. Instead of leaving a rating or review, go check it out right now on our website exit5.com our mission at Exit 5 is to help you grow your career in B2B marketing. And there's no better place to do that than with us at exit 5. There's nearly 5,000 members now in our community. People are in there posting every day asking questions about things like marketing, planning, ideas, inspiration, asking questions and getting feedback from your peers. Building your own network of marketers who are doing the same thing you are so you can have a peer group or maybe just venting about your boss when you need to get in there and get something off your chest. It's 100% free to join for seven days so you can go and check it out risk free and then there's a small annual fee to pay if you want to become a member for the year. Go check it out. Learn more exit5.com and I will see you over there in the community. Hey it's Dave. I want to give a quick shout out to Knack for sponsoring today's episode. Knack is a purpose built email and landing page platform and they're also one of our longest running sponsors. When I create our newsletter each week, I spend a bunch of time more recently with Claude, my friend Claude as my editor. But once I'm done editing the newsletter, it's not as simple as just getting my copy from a Google Doc and hitting send. If you're a B2B marketer, you know that. So what happens? Someone has to take that output and turn it into an actual email that renders an outlook, follows brand guidelines and ships. You know this story. The last mile still feels slow and manual. NAC has made this a lot shorter. They just launched an MCP server that connects your AI assistant assistant directly to their platform. So now you can describe the email you need in cloud or chat GPT and draft it like normal. But it automatically starts building a Knack for you. You get an email that comes out following your brand rules automatically. No manual cleanup, no broken HTML, and even better quality than anything your team built by hand. The Marketing Ops team at OpenAI is actually running this workflow right now. They intake internal campaign requests from Slack, an AI agent structures it into a ticket, NAC MCP generates the email and a marketer refines and ships. This is the future of marketing. You should go check it out at nac.
- [43:19]
Matt
Com.
- [43:19]
Dave Gehard
That's knak.
- [43:22]
Matt
Com.