← Marketing Against The Grain
Loading summary
Transcript5 lines
- [00:01]
A
We got a special show for you today. Today we've got Barbara back on the show. Barbara is one of our most popular episodes of all time and she's bringing the heat. She is walking us through how to build a content engine in Claude code. And this is taking a single webinar transcript. We all do webinars and turning that into blog posts, LinkedIn posts, email sequences, everything. You need to have a full content strategy and content plan around a basic event like a webinar. This is something a lot of people can use. She is an AI and automation pro. She's a non technical builder. So we're really excited to have her come back and not tell us what she's doing, but show demo and walkthrough so that we can all build and recreate this ourselves. Let's get to today's show. Here's a quick word from HubSpot. HubSpot helped Tumblr solve a big problem. They needed to move fast to produce trending content. But their marketing team was stuck waiting on engineers to code every single email campaign. Now they use HubSpot's customer platform to email real time trending content to millions of users in just seconds. The impact? Three times more engagement, double the content creation. Want to move faster like Tumblr? Visit HubSpot.com.
- [01:27]
B
Hi everyone. I'm going to be showing you a really useful workflow, one of my newer ones. It's an entire content engine that can be generating content for you after every webinar, podcast episode or anything that your company wants to promote. This is a workflow that I use in my own agency for all of my clients who are holding webinars. We have this workflow kind of down to a T so that we can make sure that we are planning the webinar, building that promotion engine, hosting the webinar, and then taking advantage all of the content creating for promotion afterwards. Which means that when you do that entire ecosystem, you have this really great content engine for your leads that you can kind of keep them warm, that you can start converting them and you get that engagement in RSVPs. So this is something I use every single day for my clients. From planning the webinar, to organizing everything, to hosting it, to dealing with the content afterwards, which is arguably the most important part. The live event is the least important thing. Everything that you do around it is what's important for keeping those leads warm and making sure that you are slowly converting them over time, especially when you have something like a webinar series. So after the webinar when you have your content, your transcript, your recordings, here's what you can very easily do with them to make sure that you're immediately promoting that content and making sure that you're reaching your target list with that and taking advantage of it. So this is a workflow that I use for all of my clients who are doing webinars. So we work from step one to step three. In step one, we have the basic planning of the webinar that we use AI for as well. So figuring out who their audience is, taking all of their current icp, figuring out what is the best topic for them, how do we need to structure this event? Where do we need to get RSVPs, what is our kind of starting point? And then building that promotion engine based on it, we want to make sure that we are posting everything on social media. We make sure that there is a distribution, because just posting on social media is not really going to give you those RSVPs unless you have like thousands and thousands of followers. And when you kind of have that promotion and content, you can easily find ways to distribute via email, LinkedIn. And then after the webinar is done, you need to then take that content, take that transcript, take that recording, and make sure that you're keeping those leads warm. So everyone who showed up for the webinar and everyone who RSVP but might not have showed up, you still have their email. You still have to keep that lead warm. And this is the part of the content engine that I will be showing today, the post webinar content engine, and how you can do it very easily with a very small team and basically immediately after the event and have that ready. So if you want to build this for your own team, go to the link in the description and it will have everything you need to build this custom AI agent for yourself. Be creative, be original, tell it what you need and tailor it to your own workflows. Click the link in the description box and you will find everything you need to build your own custom AI agent. So in terms of the tech stack that I use, I outlined everything here for all of the steps, but we will be focusing on step three. Today we're using Claude for everything that we need to do with AI. And I will be showing how we created and automated the content engine. We are assuming that you are using Riverside for webinars. You can use anything that you might want. Zoom is great. Riverside gives you nicer recordings. So that's my personal preference. It gives you a really nice landing page for webinars as well for RSVP's so it's a really good solution to kind of handle everything. We used to use a zapier integration to take the transcript from our webinar and then automatically just generate all of the content that we need. So LinkedIn post, blog post, newsletter, LinkedIn articles, any outreach that we might want to do after the webinar, it automatically just gets generated it for me based on the transcript. This part we now automated even more with a custom agent that we built. So I'm going to demo that agent today and I'm going to tell you exactly how I built it so that you can build your own or you can just take the markdown file that we will leave in the comments or in the description and you can just recreate the exact thing that I'm going to be showing you on my laptop. All of the files can be basically generated and stored wherever your team works best. So whether that's notion, Google Docs, whatever you guys might use so that everyone on your team who needs to look at the content before it goes out, they can see it easily. If you have different people for social media, email, they can see everything in one place and just prepare those promotions and send them out. So it's really easy to sync across teams. And for email and for LinkedIn automation, we usually use Heyreach and instantly because they have an integration together. So Heyreach might be a bit of a newer player, but it's definitely by far the best LinkedIn automation tool I've used. So that's kind of our tech stack for this third part. So there's a few ways you can do this and it depends on what works best with your current workflow and how big is your team and what is generally the best way for you guys to sync on this content. But you can use zapier, you can use Nan. Essentially, what Riverside will give you right after the recording is a transcript. That transcript is everything that you need to generate the rest of the content. So you should go into Riverside, you should download the transcript. So I didn't automate the step because I already do it organically. But look at your own workflow and if something needs automating, do it. If it's going to work well, it should be custom to what you're doing. Then you can use a zap that then takes that transcript and transforms it into these forms of content with Claude that you need so you can outline exactly what your team needs. For me, that's usually LinkedIn posts, LinkedIn articles, blog posts, newsletter recaps, email that gets sent out to attendees, and then I also want to two separate emails that are for cold outreach because you should be using that webinar and any upcoming webinars to reach out to people via cold email. It's a really good strategy for anyone who is maybe a lead or an existing warm lead that you have in your pipeline and Claude would generate all of that for you. Now the other way that you can do this is a bit more custom, but I think it's well worth putting a little bit of more effort in with Claude. It is basically your own custom agent. So this is the one that I built specifically for my use case and for my clients. We technically build each of these for each of my clients, which might not be like the most effective or scalable way, but again it takes you like maybe two hours to build it. So I don't think it's that big of a lift. So this is what my content agent looks like. It's fairly easy. I build it through cloud code. I am not technical, but if you just go into cloud code, if you describe what you need, it will help you build exactly what you need. So for example, this does exist on this domain and I can share it for everyone to test it out to use it just so you can see what you can do with Claude code. But I highly suggest you or anyone else on the team go into cloud code and explain what you need, what your current workflow is. If you've already been hosting webinars, if you already have any kind of transcript that you need turn into content, just describe what you need and what your final result should be and it will tell you how to do that. Again, I'm non technical and I built this. It just tells you what you need. You need to have a little bit of patience, you need to have a few, you know, back and forth. It's never a one shot, but if you have any issues, worst case scenario, you just take a screenshot of what you're seeing. If it's an error and you just put it into Claude and it will tell you how to resolve it. It's a very creative and organic way to kind of communicate with AI. And if you find Claude code a bit intimidating, go into cloud chat, explain what you want, tell it to draft what the tool you're trying to build should look like and then just download that as a markdown and put it into cloud code. So you kind of have what's essentially like a prompt for it to start with. But my favorite part about all of this is that it is essentially just a conversation, so you can just go ahead and explain what you need. We had a very clear idea of what we need to happen after each webinar, so that was pretty straightforward to us. But this is a test that I built just to demo here. Let's take it up for a spin. As it was building it, it was also telling me exactly what it needs. So this is not something I would be able to figure out myself. It just cloud told me what to do and what to input and it kind of guided me through it. So let's upload a transcript here from one of my recent webinars. It's a transcript straight from Riverside. And this is also something that you can automate. So for each client I basically have an agent that does the same thing. And it already is familiar with who the speaker is, who the host is. It can pull that from the transcript as well because Riverside gives you name and everything. And it's also familiar who the target audience is, what the brand voice is. But obviously that changes sometimes from webinar to webinar. So I like that I can kind of input it and control it that way. And then we're just waiting for the doc to generate. Takes about two minutes. And then we will show you the final part. In the meantime, you will see the Google Doc that generates as a result. Most of my clients work with Google Docs versus Notion, so that's kind of what I defaulted to in Notion. You can also have it set up basically an environment that you always use. So like a full Notion page with, you know, specific pages for LinkedIn posts, for blog posts, like anything that your team needs to kind of align on things. And for me, honestly, a Google Doc is easier. It's easier for companies to share that you can really tailor it to exactly what you need. And this is why I think everyone should kind of try to build their own versus just adopt whatever I give you as a download. You can play around with it, but essentially you should be able to create something that's custom for you, because that's going to be the best automation for you as well. And the benefit of automations with Zapier and whatever is most teams know how to work with it. At least in tech, they're very easy to use. Their UX is great, but it's not the simplest workflow either. It looks very nice and simple on Zapier, but all of those steps have to go in and be configured and zaps sometimes break. You might not be aware that they were broken for a few days. Then you have to kind of go in and figure it out.
- [12:07]
A
We'll be right back, but I want to tell you about another podcast I love. The DTC Pod, hosted by Ramon Berrios and Blaine Bolas, is brought to you by HubSpot Media. DTC Pod is a podcast about all things direct to consumer. Ramon and Blaine cover everything from starting, growing and optimizing e commerce stores and direct to consumer brands. They talk with founders, marketers, platforms, creators and marketing and growth agencies to cover topics like brand building, social media, influencer marketing, website conversion, paid media, Facebook ads, and much, much more. If you're interested in the stories behind your favorite consumer brands, this podcast is for you. They did an amazing show called Meta Ad secrets how top DTC brands spend 300k monthly profitably. You can listen to the DTC Pod wherever you get your podcast.
- [12:57]
B
So there it is. Content Package Ready. Let's Open the Google Doc. So here is the test that outlines LinkedIn post, newsletter recap, short form video scripts and email follow up sequences. Now that's usually what I need and everything else depends on the client. But this is kind of what I want to see as a baseline LinkedIn posts newsletter recap because I want to send that out the same day short form video scripts that helps the editor know how to take short videos and then you have the email. Now one note. If you build your own agent, if you build anything with cloud code, it is connected to the folders that you have on your computer, which means that those folders can have all of the context docs that you need and that AI needs to make this from a pretty whatever generic AI output to something that is very, very usable for social media. So for example, if I show you my folders for clients, I have writing examples of the founder, I have case studies, I have blog post guidelines, I have ICP lists, marketing blacklist words, demos, ICP writing examples for both the founder and the brand and that has all of their writing in it. So when you're building your agent that's going to take the transcript and generate this content for you. If you have a lot of context docs in that folder, it will pull from it and the output is going to be much, much better. This is basically working with just a transcript. So that's kind of the difference between really great quality that you can truly take from this Google Doc and mostly use it as is and kind of generic AI writing. So take a second make sure that those Folders have a lot of context, as much as you can give it. This will matter both for writing design anything you're trying to build, the difference is truly staggering. And then additionally to the context docs, when you're building an agent with CLAUDE code, you have what's called a markdown file where basically everything about that agent is stored. So you can see the implementation guide that basically describes exactly what this agent is. Now you can take this file and you can put it into cloud code and you can tell it to build what is in this markdown. And you will have exactly what I have here and you will be able to use it on your laptop. It will take you 30 minutes. With cloud code, every single agent has its prompts and references. Now, if you look at this prompt, you can see that it's very, very simple. Just take the webinar transcript, run it through four separate ABA calls, each with a specialized system prompt. LinkedIn post like, this is not a substantial prompt. What I would do now, if you have this built in Claude, simply continue to have a conversation with it. So say I would like you to follow this writing style. Maybe you have like a full list of don't use EM dashes, don't use this sentence structure. Like, just tell it what you needed to fix and just tell it to deploy that. And the next time I refresh this page, it will account for that. So if you keep tweaking it, the better and better the output will be. What you can do is just kind of one shot it and give up on it because it didn't do great on the first try. But honestly, this took what, two minutes to generate that entire doc. I'm sure you have a few minutes to work with AI a little bit to be like, okay, I like this, I don't like that. Let's lean more into this. Let's not do that. You know, work with it a little bit because this gets you to a really good 70 to 80% with just a transcript in two minutes. And this entire agent took me around two, three hours to build. So in terms of time and energy spent, I think it's a pretty good output. So this is what it looks like when you have a really solid folder with a lot of contact stocks, but both for writing and for visual elements of the company. So a lot of brand guides, all of the brand elements, colors, fonts, everything like that. And if you upload a presentation that you use during the webinar, which it can directly pull some of these visuals from. So this is what it Looks like webinar Series Part 1 of 5. And this is the blog post that it was generated. It did both the text and the graphics. I didn't touch anything. I had to make very minimal adjustments. And I just have this blog post ready to go. And in the same breath, it also generates a LinkedIn article for me, which I've instructed it that it needs to be a bit more concise than the blog post. The format is a little different, optimization is different, and it needs to be a bit more engaging for LinkedIn. So in one go, I get all of the graphics that I need for the blog post itself. I have all of the text and I have all of the LinkedIn articles, and I know that everything is optimized. And then additionally, you can keep working with it and telling it what you need. So, for example, I want to have multiple touch points via email for cold outreach or for warm leads. And I instructed it what kind of format I like, I said what kind of links I need to be included there. And then it just follows that format every single time and basically adjusts it based on the latest transcript that I have. So after the webinar, I immediately have LinkedIn posts that I can start scheduling. I have video scripts that I can give to the editor to make some short videos edit the full webinar. It basically takes a snippet of the most interesting parts that would perform really well on social media. After each webinar, you can also take social media posts from your LinkedIn and show the analytics and just put them in the folder so that AI knows what post performs best and can kind of optimize all of the following posts based on that. And I have the newsletter and I have email outreach, not just for people who attended the webinar, but for anyone that we want to start targeting via cold outreach, we want to start targeting via LinkedIn outreach and basically any single target audience. If you have a webinar and you know that the main attendees of that webinar were three different roles within three different industries, you can also tell it to tailor an email response to for each of those target audiences, because you might have different landing page that you want those people to land on, there might be resources that you want to share, then you can really tailor your approach there and your communication without spending a week of, you know, coordinating with the team and trying to extract all of that. It just does it by default. And then you get to really use the momentum from that webinar and kind of do the outreach and promotion and make sure that you're staying top of mind to those leads that just saw you on the webinar. And then this is how the LinkedIn article, this is also all generated with AI. Cloud code works much better for graphics, especially when you have really good brand guidelines, because if you generate a graphic, you can just download it as an svg, upload it to your Figma, which means that you can move all of the elements. That was an issue before with ChatGPT. You kind of get an output, but you can change it. And if someone leaves feedback on that that you're like, oh, well, I can't really change it, I have to reprompt it. You don't really have to do that anymore. Figma now has an integration directly with Claude as well, so it's becoming a more and more seamless process. And then you can kind of load all of the social media posts to go along with the short clips when you create them. There's also different tools online that can help you with that. Riverside as well helps you generate some shorter videos and clips from the webinar itself immediately after it's recorded. I work with pretty regulated industries and most of the time AI doesn't really give me what I need with short videos. So I do often edit them myself, but I edit them in capcut and it has a lot of really good features. So it's not really that big of a deal. But I am excited for AI to remove that part as well. So this is your entire content engine that you can basically automatically generate any content that you need after a webinar, a podcast recording. You can even upload an insight call with your founder here if you need to generate some content from that. And it will do everything for you and make it very easy for your team to collaborate on it. It will land directly in your Google Docs or your notion, wherever you might want it. And you can very easily kind of take the next step and do whatever your team needs to do with it. If you want to automate any of the next steps, you can do that as well. As soon as the LinkedIn post is done, it could be scheduled. Nan is great for that. I personally like to review most of the content before it goes out, but it just depends on what you and your team need and it creates a really good content engine that you can just keep rolling. If you have a webinar every one or two months. This is a really beautiful kind of content ecosystem that you can sustain pretty easily with maybe just one person on your team. So yeah, if you want to get this exact webinar content engine that I'm using here. We will have a markdown file in the description that you can download and put it into cloud code and just tell it to build it. It will guide you through the steps. There's maybe a few things you were going to need to help with web and give access to, but you can have this exact thing on your laptop. Just tell it what you want fixed, tell what your use case is, explain what you're trying to build, and you will have a really, really useful tool that you built within an hour or two that is completely custom to what you needed. I really encourage you to also try to build it from scratch because you know your workflow's best currently and you know what you need. So my solution might not be the best for you, but it's a really good starting point and we will dive into the entire webinar workflow in one of the upcoming episodes and keep you updated with everything that we are building with cloud code and automating.
- [22:57]
A
Wow, that was awesome. Huge thank you to Barbara. If you want to go and build that yourself, we've got everything. Click the link in the description below. Get hooked up with everything to kind of build what Barbara just showed you. Please hit like, please hit subscribe. We're doing more and more awesome videos to try to help you build an AI first, marketing and business. We'll see you next time on Marketing Instagram. Hey everyone. You know Kieran and I have been doing the podcast for a while now. We've been at this for a couple years. We love it. We could not be happier to be doing this, but we wanted to take things to the next level. We want to level up the impact we're having with Marketing against the Grain. So the next step of our journey is something we're really, really excited about. We're going to launch the Marketing against the Grain newsletter. And Marketing against the Grain newsletter is going to be amazing. If you are a marketing leader, practitioner, you're in the trenches doing marketing every day. This is for you. We're going to deliver right to your email inbox and you're going to get all the behind the scenes frameworks, practices, tutorials from us, from guests we have on the show, and from people even beyond the podcast that we think are going to be helpful and really have an impact on your day to day, week to week doing marketing. You're going to love it. It is something we've been talking about for a while. We're really excited to have it out in the world. We've already got a hundred thousand marketers who are on this newsletter. Please join. It's completely free. We'd love to have you as part of the Marketing against the Grain community. And it's easy. You can click the link in the description below, or you can head to marketingagainstagrain.com subscribe.