How to Integrate AI Into Your Marketing Workflows (3 Systems You Must Try)

Everyone's saying AI is the biggest shift since the internet, which sounds cool. So why are we just using it to rephrase bullet points and rewrite a headline or two? There's obviously a gap between what this thing can do and what we're actually doing with it. So I called it an expert. Stuart Hillhouse, who's been working with the folks at StoryArb to integrate AI into all kinds of different workflows. I've spent the last number of years helping companies sort of figure out how to build. operational content systems and have really seen the evolution go from doing it all manually with no AI to then, you know, AI is okay at certain things, but really not a huge piece to now as we're recording this, you know, AI is a fundamental part of any content and marketing strategy. So the way I approach using AI, I kind of come at it from three different angles that helps a marketer in their day to day. So the first one is AI is unbelievable at distilling and compressing huge amounts of data that you otherwise wouldn't have the ability or likely time to ever sift through. Second, it's really useful as a creative partner. So something to help you brainstorm new creative angles or hooks or pitches. And then lastly, AI is fantastic for that last mile. getting your idea from in your head to in the market, which is super important for marketers. The velocity in which you can get your ideas out there in the public, the faster you're able to learn from it and adapt. So at the end of the day, you still need to be really good at marketing and you still need to understand all the foundations, but AI is there as your assistant to help get things rolling and to be able to go from something in your brain to something tangible way, way faster. And that's really the heart of Stu's approach. AI isn't here to replace marketers. It's here to do the things we can't do at scale. And nowhere is that clearer than the first workflow he walked me through. But before I jump into these workflows, I want to share a four-step crash course from HubSpot on how to leverage AI. It gives you the basics you need to understand LLMs. You'll see the principles from this crash course come to life and Stu's workflows. So it's a great place to start. Okay, back to the first workflow. Leveraging sales calls for messaging. Your sales team is having hours and hours of conversations every single week with prospects and customers. They're really getting into the details of why they use your product or need your product. And so when you listen to these calls, you actually are given all the answers. You know exactly what you need to say because you're hearing the language that your market is already using. And I think that should be the foundation that helps shape the rest of your marketing messaging. The problem is sales transcripts are packed full with useful data, but no marketer has the time to read through all these calls. So that's where Stu's workflow comes in. Start just by going and gathering between three to 10 sales calls. The more the better. You're going to take those transcripts and you're just going to use a general AI chat tool and dump them all in there and start asking questions. Stu writes the following prompt. You're a B2B marketing analyst. Given these sales call transcripts in the ICP below, extract number one, top seven pains, number two, buying triggers, number three, must have outcomes, number four, common objections with rebuttals, and number five, 15 verbatim quotes with timestamps. Then synthesize into three POV pillars, a benefit ladder for each pillar, problem value proof, five hook headlines per pillar in the customer's language, and lastly, a one-page messaging sheet. The first thing I noticed about this prompt is how specific the setup is. It's really important when you're crafting your prompts that you include specific details about your business. And these need to come from you to be really crisp about who your ideal... customer persona is so that's who you're trying to sell to another thing here that i was specific about was i wanted to think as a marketing analyst it's going to be very you know analytical and only give me answers based on the transcript so when prompting you need to bring a lot of that context but once you define them and give it to the ai it now will include that context in the answers it gives you but knowing what to include is only half of the puzzle the other half is figuring out how to structure the prompt itself. Stu had an answer for that too. You can reverse engineer any type of prompt you want from the outcome you're trying to get. You can say, hey, I'm looking to rewrite my homepage and I need some quotes and I need some pain points and I need some objections. Write me a prompt. that I could use to extract these from a transcript. That's a big unlock that you can have is realizing that you need to manually write prompts. You can actually ask AI to create the prompts for you. When you run Stu's prompt, here's what you give back. A messaging map with your core pillars tied to real quotes. An objection sheet showing what customers push back on and how to respond. And a library of copy starters that you can use across campaigns, website headlines, whatever you need. These deliverables give everyone a shared understanding of how to speak in a way that actually resonates with customers. You'll know it's working by tracking these KPIs. And the best part is you can run it monthly to keep your messaging relevant. I think what's really unique about now having AI workflows that we can use again and again is you can make lightning strike twice. Like when you do, like when AI has an output and then you put it out in the market and it gets a positive response, you know exactly the ingredients that went in to make that happen. You can then rerun the prompt and say, here is our marketing from a year ago. I really liked a b and c can you give me a modern version of this given the new data that we have this workflow makes use of data buried within your company to craft messaging based on customer truth these aren't headlines from a brainstorm that are informed by what customers are already telling you but messaging on its own can only go so far you also need to make sure people can find you especially as the way people search is changing that's why stew designed a workflow for aeo or answer engine optimization It's an evolving industry, but where it stands right now, AEO is less scientific than SEO. You're not trying to look for keyword volumes or difficulty of ranking. It's more of a game of trying to see around the corner. What else can we answer for? our prospects. What else does a buyer need to know in order to consider our product? What else do they need in order to kind of complete the purchase? So you want to have this like very holistic view of what are the questions that people might have? In what ways do I need to package that information to make it really useful? to the person but also very accessible to the llm that is scraping that content so step one here is we're actually going to use those sales calls again and ask ai to extract all the questions that they prospect asked and then step two is we're going to extrapolate on those and what this is going to do is it's going to explore all of the angles and all of the positioning that a prospect might have and give me 10 more questions that digs deeper into that line of questioning. So now what we're left with is a hundred questions and these all now become on-ramps for your content to discuss those and you can create a ton of different marketing collateral that your prospect is already likely thinking about. Stu recommends figuring out what content to prioritize with the following prompt, act as an AEO strategist. From these sources, output a CSV with columns, stage, question, intent, evidence needed, priority one through five, existing content URLs. For the top 10, draft briefs with a direct answer, outline, citations needed, a CTA suggestion, and three title variants. Stu stopped short of asking the AI to create the content, only asking for briefs because he believes a human point of view is a differentiator. Technically, LLMs aren't actually thinking. They are just putting words together. The AEO is desperately looking for new content and points of view that people have. So rather than just creating AI-generated content, from other AI-generated content, you're actually rewarded for creating hyper-specific content that's tailored for a very particular audience that uncovers net new ideas. And so the more quotes you have, the more expert opinions you interview, the more data points and evidence that you reference around a particular idea, your content will perform better because now you're giving the LLMs novel new information that it can reference in its answers. So use AI to set the scaffolding, but keep the storytelling human. Your goal should be to create content that's less interested in keywords and more interested in being complete, credible, and sight-worthy. And the most credible and sight-worthy people at your company are the C-suite. People want to hear from them, but they rarely have the time to post. Their best ideas end up trapped in internal meetings and on sales calls. This final workflow from Stu fixes that. For this one, you're going to want to start... by gathering some transcripts. And these can kind of be a lot of different forms. They can be leadership interviews. They can be customer calls. They can even be internal all hands, assuming there's no proprietary data being shared. Even 20 minutes of a recorded conversation between you and your manager or you and your CEO is enough to generate a week of posts. So the second step then is now to run a prompt like this. From this transcript and our point of view document, propose five angles. For each, include a short thesis, who cares, one number to prove it, and the cleanest counterargument. The idea here is not to come out with like a finished draft. It's just to define a really unique point of view that that person can write about that will be interesting to see on social media. And that is like the number one thing that any individual can do when posting on social media is to have a strong point of view. Once you have the angle, you tailor the voice. Take existing writing from your leadership and feed them to the AI with this prompt. create a tone of voice guide that replicates this style if that material doesn't exist have the ai analyze their transcripts for patterns diction cadence tone and output a style guide now you have a system for writing in their voice and you can action these tools to create thought leadership that's consistent in tone and appealing to your target customer to make it sustainable keep it simple you need a content lead to run prompts and draft posts your leader for approvals and context and someone in comps to check accuracy and compliance you can take any one of these workflows and implement them separately but Stu designed these to all work in parallel. You can have the messaging going on over here and then you can have executive thought leadership going on over here. But a truly unified marketing strategy works best when one activity feeds another activity and then gives you real market signals that then goes back and helps inform the strategy at the top again. And so that's why I like to see these all as running together, but also could be independent setups. Stu showed me how AI does the work marketers rarely have time to do to enhance the marketing we do. open to the imagination if you want to see what else is possible check out the ai workflows crash course i mentioned earlier it'll give you everything you need to start experimenting with your own systems and let us know in the comments how you're using ai in your marketing we're all figuring this out together and the best ideas usually come from seeing what other marketers are trying

*Get our free crash course on AI for business builders and learn the LLM fundamentals behind these workflows.* 🔗 https://clickhubspot.com/478bfd

Most marketers are using AI to rewrite a headline or two, but there's a massive gap between what AI can actually do and what we're doing with it. In this tutorial, I brought in Stew Hillhouse from Storyarb to walk through three AI marketing workflows you can implement right now to transform how your business operates. 

We're covering how to:
Turn sales calls into messaging gold 
Optimize for answer engines like ChatGPT and Perplexity 
Scale executive thought leadership without burning out your C-suite 

These aren't theoretical frameworks, they're practical systems with prompts you can copy and KPIs you can track. If you've been wondering how to actually integrate AI into your marketing beyond the basics, this is where you start.

TIMESTAMPS:
0:00 - Why we're underusing AI in marketing
0:16 - Meet Stew Hillhouse: AI workflow expert
1:58 - Workflow #1: Turn sales calls into messaging strategy
6:13 - Workflow #2: Turning buyer questions into discoverable content (AEO)
9:07 - Workflow #3: Turn executive convos into thought leadership content
11:06 - How these workflows work together
11:38 - Next steps for implementation

Subscribe for more HubSpot
🔗 https://clickhubspot.com/t90m

📙 FREE Certification Courses
Digital Marketing Certification:
👉 https://clickhubspot.com/od6
Social Media Marketing Course:
👉 https://clickhubspot.com/Social-Media-Certification
SEO Training Course:
👉 https://clickhubspot.com/SEO-Training-Course
Email Marketing Course:
👉 https://clickhubspot.com/Email-Marketing-Certification

About HubSpot
HubSpot is a customer platform that provides education, software, and support to help businesses grow better. The platform includes marketing, sales, service, commerce, operations, and website management products that start free and scale to meet our customers’ needs at any stage of growth. Today, thousands of customers around the world use HubSpot’s powerful and easy-to-use tools and apps to attract, engage, and delight customers.

⚠️ Disclaimer
The videos on our YouTube channel are for informational purposes only, and are not intended as an endorsement for any of the products or services that we feature.

#HubSpot #AIMarketing #MarketingWorkflows #B2BMarketing #AnswerEngineOptimization