HubSpot AI Data Sources: The Story of Your Products + Services

In this episode—the final installment of our HubSpot AI Data Sources series—we dove into a deceptively simple, yet critically important part of the AI settings: Products and Services. On its face, this feature might look straightforward. You plug in your value propositions, key pain points, and product or service descriptions. But as we discovered in our conversation, there’s so much more happening here—and so much potential that brands cannot afford to ignore.

If you’re thinking, “Why does this matter? Isn’t this just another spot to input data?” I hear you. But here’s the deal: HubSpot AI needs this information to function effectively as a strategic tool.

Your value propositions, the pain points you solve, and the products you sell are the backbone of how AI-generated content is personalized to your brand. Whether you’re using HubSpot’s social media agent, prospecting agent, or any other AI-driven tool, this data ensures that what it creates doesn’t just sound right—it aligns with the outcomes your business is trying to achieve.

🚀 HubSpot AI Data Sources Series:

• How to Get Started with HubSpot AI Data Sources
• HubSpot AI Data Sources: Brand Voice + Tone
• HubSpot AI Data Sources: Buyer Personas vs. Ideal Customer Profiles
• HubSpot AI Data Sources: Marketing Strategy

This episode isn’t just a feature review, though. We dug into the larger questions this raises about messaging strategies, the structure of marketing data, and how businesses should be thinking about their AI inputs. Should all this information live in a single messaging strategy tab? Or does breaking it into value propositions, pain points, and products make it easier for AI to process and apply? Oh, we had thoughts—and we didn’t all agree.

We also tackled some of the concerns that naturally come with this tool. Will teams take the time to use this correctly? What happens when brands overload AI with too much information, or worse, input the wrong stuff entirely? Garbage in, garbage out, as George pointed out. But on the flip side, this tool opens up some exciting opportunities. We explored practical strategies for filling it out, best practices, and even a few formulas you can steal to make sure your inputs are rich, concise, and effective.

Whether you’re skeptical, excited, or somewhere in between, this is a conversation you don’t want to miss.

Keywords

HubSpot, AI data sources,  marketing strategy, products and services, inbound marketing, customer engagement, holiday greetings, messaging strategy, AI tools, content creation, customer experience, data sources, implementation concerns

What We Cover

• What’s Inside the Products and Services Setting: We break down exactly what you’ll find when you open this tab—value propositions, pain points, and itemized product or service descriptions—and why each piece matters for AI tools like the social media and prospecting agents.

• How AI Processes Products and Services: AI doesn’t just “read” your inputs—it uses them to generate content that’s personalized to your goals. We discuss how this setting informs AI-driven campaigns, proposals, and messaging across your HubSpot tools.

• Why Messaging Strategy is at the Heart of This Tool: This feature introduces traditional marketing architecture—value props, pain points—into HubSpot. But is this the best way to structure it? We debate whether businesses need a single messaging strategy tab or if splitting these elements makes sense for AI.

• The Practical Role of Value Propositions and Pain Points: These aren’t just marketing buzzwords—they’re the building blocks of clear, consistent messaging. We discuss how to craft value propositions that resonate and how pain points give your AI the context it needs to engage your audience.

• Formulas and Best Practices to Fill This Out: George shares proven formulas for crafting standout value propositions and product descriptions, along with rookie mistakes to avoid. Whether you’re starting from scratch or refining existing messaging, we’ve got you covered.

• What Happens When You Do This Right (or Wrong): AI tools are only as good as the information you give them. We explore the risks of cutting corners and the benefits of taking the time to get this right—think personalized content, clearer messaging, and smarter prospecting.

• The Future of HubSpot AI and Messaging: Finally, we look ahead. How could tools like this evolve? Will there be a more integrated approach to messaging strategies in the future? And what does this tell us about where HubSpot AI is headed overall?

And so much more ... 

Episode Transcript

[00:00:00] Liz Moorehead: How are we doing, everybody?

[00:00:02] Chad Hohn: So good

[00:00:03] George B. Thomas: you can tell we're in the festive mood and season with the, this, this is going to be a real interesting episode. I have a feeling.

[00:00:12] Liz Moorehead: You know, George, I always feel really safe and comforted when you declare at the top how interesting an episode is going. That brings me, as the person who has to wrangle all of you,

[00:00:23] George B. Thomas: I'm,

[00:00:24] Liz Moorehead: tidings of great joy.

[00:00:25] George B. Thomas: I'm just reading the room. I'm reading the room. I'm understanding what happened during the intro. I'm watching Max's face. I just, we'll see.

[00:00:35] Liz Moorehead: I gotta be perfectly honest, Max has been the one on his best behavior since we got on this morning. Which I'm going to, that actually does not fill me with joy. That does not spark joy. This is, we've now, we're hitting a critical mass that I find a little bit upsetting. You know what, George, let's just dig into it.

[00:00:53] Can you, I want to kick it over to you for some housekeeping items. So our listeners know what to expect during the holidays because next week is Christmas y'all.

[00:01:01] George B. Thomas: Yeah, I, my goodness. Uh, that got here quick. Uh, I mean, listen, we're gonna, we're gonna, you know, throw a couple things at you as far as like, Uh, a season roundup and, you know, the, our favorite episodes of this year, but the, uh, the Hub Heroes crew has taken a little bit of time, um, of course, to maybe strategize, uh, maybe drink some eggnog, uh, maybe hang out with family.

[00:01:29] So this, this episode is the one where we need to just wish you like Merry Christmas, Happy Holidays, Happy New Year, whatever, all of the things that. You celebrate our, we, we say that to you today, um, because, um, we're just, we're, we're about to go into chill mode.

[00:01:46] Liz Moorehead: goodbye.

[00:01:47] George B. Thomas: Buh bye!

[00:01:48] Liz Moorehead: Peace.

[00:01:48] Max Cohen: we said that so weird? Felt like we

[00:01:51] Liz Moorehead: Now you decide to talk,

[00:01:52] George B. Thomas: Chill, chill mode?

[00:01:54] Liz Moorehead: by calling us weird?

[00:01:55] Max Cohen: No, I'm just saying we, we said it as if we were like stepping back to completely re evaluate everything we're doing. We're just taking two weeks off for the holidays.

[00:02:03] George B. Thomas: I mean, doesn't, isn't that what people

[00:02:06] Max Cohen: You say, the team's gonna be stepping

[00:02:08] Liz Moorehead: No, George, not everyone is you.

[00:02:10] Max Cohen: gonna be doing for

[00:02:12] Liz Moorehead: not everyone takes two weeks off and starts news companies, ends other companies. Like some people just take a nap.

[00:02:17] George B. Thomas: am, am I the only one that like, re evaluates everything that I'm gonna do before the, the next year?

[00:02:24] Max Cohen: Yeah, I'll reevaluate

[00:02:25] Liz Moorehead: some re evaluation. I mean, my big thing about the, well, here, a honest answer to your question. I do some of that. Like I do sit down and kind of take stock. Like, what did I do this year? What do I want to stop, start, keep doing. But I also know that if I stared anything long enough, I will find a problem with it and try to kill the baby.

[00:02:46] Like, I can't, can't do that. Like,

[00:02:50] Chad Hohn: Don't do it, Liz, please.

[00:02:53] George B. Thomas: Well, I will tell you this, I will tell you this, I refuse to start another company. I

[00:02:59] Liz Moorehead: well, that's a relief. But what about another podcast?

[00:03:03] George B. Thomas: uh, the

[00:03:05] Liz Moorehead: we go.

[00:03:06] George B. Thomas: still, the jury's still out on that one. Because, um, well, well, anyway, not why we're here. We're here to talk about Hubspot. We're not here to talk about the superhuman framework. But I'm looking forward to the superhuman framework in 2025.

[00:03:22] Uh, and that might equal a podcast? I don't know yet, we'll see, but anyway. Okay, so what are

[00:03:28] Liz Moorehead: You're right, George, speaking of HubSpot AI data sources, no,

[00:03:32] George B. Thomas: we go.

[00:03:33] Max Cohen: Man how many episodes have we done on the data sources now at this point?

[00:03:37] Liz Moorehead: this is our fourth one and it's our final one, my guy,

[00:03:39] Max Cohen: Oh, wow. We're at the end of the road

[00:03:41] Liz Moorehead: I know. So we have

[00:03:43] George B. Thomas: to men.

[00:03:44] Liz Moorehead: brand. I would, I was about to sing it and then I'd realize I'd love for us to keep listeners going into,

[00:03:51] George B. Thomas: Wait, we have listeners? Oh, we should probably behave then. I told you it was going to be a fun episode.

[00:03:59] Liz Moorehead: we done, gentlemen?

[00:04:01] Chad Hohn: not possible to behave.

[00:04:03] Liz Moorehead: Yes! Kick it, Chad!

[00:04:06] Chad Hohn: You thought you'd get away.

[00:04:09] Max Cohen: For a second, I thought that was chad just with like a voice modulator Anyway My farm's going great

[00:04:18] Chad Hohn: That was me with the voice modulator.

[00:04:20] Liz Moorehead: It

[00:04:20] Max Cohen: Oh it was okay

[00:04:21] Liz Moorehead: him. Well, yes, so we've

[00:04:24] George B. Thomas: I gotta figure out how to do that. There needs to be a gig at George. Okay, let's,

[00:04:28] Liz Moorehead: I will turn this podcast around. I swear, I will pull us over, we will not stop at McDonald's. That's it. Be good. Everyone stay on their respective sides. Thanks, Gigamat. No, so we've actually done brand voice and tone. We did ideal customer profiles and we did marketing strategy.

[00:04:50] But George, before we get into our final installment, let's just imagine for a moment we've got some listeners with us who are just hopping on the HubSpot AI data sources train. Can you give us a quick rundown of what those are?

[00:05:03] George B. Thomas: so the data sources train is the AI context and specificity that it needs to, uh, use the content agent, the social agent, the service agent, the prospecting agent, any of the agents in the future are going to look at things like brand kits for your brand voice, a company profile, ideal customer profiles.

[00:05:28] Hopefully that'll get better in the future. Anyway, now why we're here, marketing strategy, user profiles. And today we're talking about products and services and what, uh, lies inside of products and services and why those pieces are important. Uh, to humans, to businesses, but also in the context of AI agents, uh, as we move forward.

[00:05:48] So that is, that is AI Data Sources, a, a setting inside of Settings of Settings, uh, and we've done four episodes, so that should tell you something, yeah. That should tell you something about the potential thought of importance around these items inside here.

[00:06:05] Liz Moorehead: You know, it's interesting. I originally had in my notes that I wanted to talk about today that I was feeling a little less nervous going into this episode than I was last week's episode, because if you missed last week's episode on the marketing strategy setting singular, go back and listen to it.

[00:06:23] I don't

[00:06:23] Chad Hohn: it was property and setting were both singular.

[00:06:28] Liz Moorehead: We're just going to, we're going to, but originally I was going to say I was, I was less nervous I'm nervous, but George, you've cured me. You cured me entirely of that. So that's, so we're just going to move right in. Yeah. We're just going to move right in. I am. One thing I will say, I am excited about today's episode is the fact that last week's marketing strategy conversation was around a singular lightweight product.

[00:06:50] Property and this week we are dealing with something quite dense where I feel like we're going to have some really fun stuff to talk about here because a product and services discussion for any company is critically important because it's not just about what they are. It's about what makes them different and what they actually solve for.

[00:07:10] But George, I want to come back to you. Can you share with folks what they will find when they open up the product and services setting in AI Data Sources in AppSpot?

[00:07:20] George B. Thomas: Yeah, there's actually a couple things and I want to talk about these from a couple different angles as we move forward, but as soon as you open the page, you're going to see Um, first of all, it's literally just says products and services and it says describe your organization's value, propositions, products and services to ensure AI generated content is relevant and tailored to your organization.

[00:07:45] If we have time, by the way, I'm going to open up the, uh, social media agent and show what that actually looks like from a products and services standpoint of this, but you're going to get value proposition or a place to add your value proposition. You're also going to, uh, see that you can add what pain points does your company solve.

[00:08:06] And then you can add an itemized products or service list. So this is literally like the product name and the product description. And so that's, that's what you'll see when you get there.

[00:08:17] Liz Moorehead: So what do we love about this tool on its face?

[00:08:19] Max Cohen: get, is it, is there any transparency about how it actually uses this information or is it just kind of going into the, You know, You know, the, uh, the AI machine and saying it's giving it context for content it's creating.

[00:08:33] George B. Thomas: I mean, it pretty much the context is that top part of the page where it's going to be using it for, at this point for, uh, the different agents or content

[00:08:44] Max Cohen: Oh, the agents got it. Okay, cool. All

[00:08:48] George B. Thomas: I can draw, I can draw a straight line for you right from this to my social media agent. And also we know as soon as you talk about value proposition and pain points, prospecting agent.

[00:09:01] I mean it's just, duh, like,

[00:09:03] Max Cohen: Yeah. Yeah, yeah, yeah. No, I, so I know I'm usually, um, Hella skeptical about these tools, right? But I'm actually kind of excited about this one. Just

[00:09:17] Liz Moorehead: Oh,

[00:09:18] Max Cohen: when I did implementation, I know I say that a thousand times on the show, but like, you know,

[00:09:24] Liz Moorehead: wait, did you do implementation?

[00:09:25] Max Cohen: did implementation. Well, no, it's just interesting.

[00:09:28] The question I would always ask on the first onboarding call is, I would just say, tell me about how you make money, right? So I could like fully, fully understand what the end zone is. Right, because if like we're putting together this like, you know, giant strategy with HubSpot, right, it's really important to kind of like figure out like what the immediate end goal is, right, in terms of how do we get people to buy things from you?

[00:09:53] How do they buy from you? What do you actually sell? Because I felt if I knew that everything that they would have to do with HubSpot would slowly come into focus, right? But if I don't know what like the end point is, and I'm just throwing money. Generic inbound mumbo jumbo at them and I don't have anything to kind of like tie it into or or you know Understand again the end point that the strategy is like leading to and I know it's not the ultimate end point like buying something Right, but at least at the very beginning when you're like just setting it up and it's like how do you get to value quickest?

[00:10:24] And understand how you support that for me. It was really important to understand that right? So what I think is really neat here is that it feels like they're at least at the beginning You Of getting all these different agents and hopefully influencing the way that content's being written to understand what the End goal is right is to Get people to buy this thing in this like certain way so they can accomplish xyz Right, and I feel like if the agents have that context They can probably do a much better job in the way that they communicate or suggest things or whatever it may be.

[00:11:00] So, I'm actually kind of stoked on this.

[00:11:04] George B. Thomas: Yeah, I, I love this piece too. And, and it's funny because you, Max, are jumping into AI agents. I'll say even for humans, this is important, right? Uh, and, and one of the things I did is I love to dissect things with AI. Um, I have quickly over the last year and a half, almost two years, become a super AI nerd.

[00:11:25] And so I took a screenshot of this product and services with it filled out of value proposition of what pain points is your company sold for us sidekick strategies. And I said, why is this information important for a business to know? And why is it important for an AI agent to know? Because I wanted to see if there was any differences.

[00:11:44] And it's, it's quite interesting because for the businesses, it talks about clear value proposition. And the fact that a well defined value proposition helps businesses articulate what sets them apart, enabling them to attract the ideal customers. Now, I grabbed that word like I did, because literally in one of the AI data sources, we talked about ideal customer profiles, right?

[00:12:07] And so value proposition attached to ideal customer profiles. It talks about for businesses addressing customer pain points, understanding and documenting the pain points the company solves, make sure they focus on delivering solutions that matter most to their target audiences. Now, here's the fun part because we're literally talking about prospecting agent target audiences ideas.

[00:12:30] So you can kind of see how even from the human side or business side, it talks about clarity on products and services and strategic growth differentiation. Now, when you get into. Um, why is this information important for AI agents? What's fun is it talks about personalized and relevant responses. An AI agent can generate tailored content answers and campaigns based on the business's value proposition, pain points, and product offerings.

[00:12:58] This avoids generic or irrelevant output, which we don't want that. We want it to feel very much like as effective sales and marketing assistance, the AI can assist in creating marketing materials, email campaigns, or pitches that directly address customer challenges and align with the company's offerings.

[00:13:17] Customer support optimization, streamlined content creation. AI can generate blog posts, social media content and proposals that showcase the business's expertise in solving complex challenges. Consistent messaging is a big piece of this, so like it's important to both, but important in different directions.

[00:13:39] When you think about the documentation and having this in un, in HubSpot, and in your organization. But I literally then my brain jumped to this other piece that I want to mention, and then I'll shut up for a minute and we can kind of go over this a little bit more as I was like, well, is there a formula for like, crafting a value proposition that's going to work for AI assistance and your business?

[00:14:04] And so what I would want the listeners to realize is there's a, who you help, what you do, how you do it, and the transformation slash results formula that you could start to pay attention when you're when you're crafting this. So. When I think about this, like, who you help to identifying in this statement that you're writing.

[00:14:23] Uh, the specific group or persona that you serve. Uh, for business, you know, small to medium sized businesses, enterprise companies, whatever it is for the agents, teams, individuals, marketers, like things like that. What do you do the services the products? How do you do it? Human 1st approach advanced data driven strategies, tailored solutions, powered or assisted optimization.

[00:14:45] These are just words that we might use, um, the transformation or the results. Save time, increase ROI, unlock new opportunities. So you want to think about these things and Liz, maybe I'll get you the formula so we can put the formula on the show notes and people can be like, I want to steal this and use this when I'm crafting the value proposition, because again, do I just want you to come and just type some small stuff in here?

[00:15:11] No, I want you to create a rich, robust, very powerful value proposition. I want you to be able to have a, uh, I'll talk about pain points formula later, but I want you to have a very, Valuable, value proposition, if you will, inside of this setting.

[00:15:28] Liz Moorehead: So this is where I want to hop in and have a bit of a conversation, because what's interesting about this is that I love that we have things like pain points and value propositions. We have brand voice and tone and stuff like that. But what's interesting is that these are assets that are often created.

[00:15:43] By brands in different ways, like value propositions and pain points. Those are like the 1st 2 sections of every miss messaging strategy. I ever write like that. That is, you know, I pulled up 1 just so I have it here next to me. You know, I'm looking at 1. I did for a client recently where there are 5 different specific.

[00:16:02] Value propositions that they provide because they are tied to 5 very specific. Pain points, right? So we have this interesting thing where we're starting to see some more traditional marketing architecture make its way into HubSpot. What I find interesting about it, and I can't decide if it's something where I like it and this is what it needs to be, or I wish it was a little bit different, is that why isn't it just a messaging strategy tab?

[00:16:29] Why isn't it just a single document where that can all get uploaded? Or is this the best way for these AI tools to process this information? Like you need to pull value propositions and pain points out and have it directly paired with products and services. Because like I said, like I have, I'm literally sitting here with a document in front of me that goes through value propositions.

[00:16:51] Pain points at great length and detail as well as their differentiators, which is a 3rd thing that is neither value propositions nor pain points. So I think we're getting into this interesting spot where I'm loving what I'm seeing, right? I'm loving this idea that we're bringing more storytelling to it.

[00:17:10] Right. Um, I wish our listeners could see live what is happening. I'm sorry, Chad. Did, did, did Santa Chad just come and help fix the fire for you behind you?

[00:17:22] Chad Hohn: Uh, no, that was Linus from Linus Tech Tips. I'm

[00:17:27] George B. Thomas: I love that channel, by the way.

[00:17:30] Chad Hohn: Yeah. That was legitimate. I mean, this is Linus in this stalking, and then that's Dennis. Who's like one of their content creators and not stalking

[00:17:39] George B. Thomas: Which, by the way, if you're listening to this podcast and you don't know what this is, I'm sorry. But, it's, it's cool. And you should watch, by the way, because you'd see Chad's background. I'll just throw that there too, so.

[00:17:52] Chad Hohn: We do a live

[00:17:53] Liz Moorehead: have newer ways to watch. We'll have newer ways to watch in the new year. Just saying.

[00:17:59] George B. Thomas: Yeah,

[00:17:59] Chad Hohn: If we if we pass the test

[00:18:02] Liz Moorehead: if, well, we've done it before. We can do it again. I believe in us.

[00:18:05] George B. Thomas: We'll see.

[00:18:06] Liz Moorehead: But to go back to my original point though, this is where I find a lot of this stuff fascinating, right? There are already methodologies and ways to craft messaging strategies that encompass all of these things.

[00:18:19] I, I personally have GBTs that I built using these messaging strategies built in a very structured way. So again, I love this idea of seeing all of these things come in. And do we need to break the structure? Is this the way it should be structured? Why can't I just have a messaging strategy tab that has certain things in there?

[00:18:42] I still really love it though. Like that's the thing. I can't understand Chad. Maybe this is where you might be able to answer my question. Is it better that value proposition and pain points are pulled out and put directly with products and services? So the AI chat agents Better understand how to process this information, because how messaging strategy documents are built for humans may not be the best way that they are processed by these AI tools that we're now becoming so reliant upon.

[00:19:08] George B. Thomas: Chad, after you go, I have thoughts.

[00:19:10] Chad Hohn: Yeah, sure. Yeah. I mean, I think it's uh, you know for for a couple different reasons, right? It's this way and one of the reasons is You know, HubSpot has a bunch of teams of people working on a bunch of different features, right? And so, it's not like they got one dude where they just, you know, slide some pizzas under the door every now and again and all of a sudden they get Breeze Copilot, right?

[00:19:32] You know, that's not happening. Like, they got so many people working on it, right? And so, because of that, We have all of these different data source functions in different tabs because each different team is working on it all at the same time, most likely, right? That's what I would imagine. Anyway, obviously they're all, they're all over some, or there's a product lead over kind of the whole thing and, you know, they have meetings and whatever, but like people are building them.

[00:19:59] Individually, right? And traditionally we're used to like trying to make our own model or our own GPT with memory where we are putting in the guardrails into our ginormous plain text document, you put in things like, here is a value proposition and here is a company pain point, like in your documentation, it's pretty structured.

[00:20:24] But if you give. People the opportunity to put in traditional documentation and then they just feed it a big pile of poo. Like it's not going to know how to pull the data out of that necessarily, right? I mean, it'll be able to try to look forward to the best it can. So what I'm imagining is on, on their side, they feed value proposition as a variable into A prompt like a larger prompt architecture where it would go in normal documentation.

[00:21:02] It's almost like if you could see, like, if it spit out a document after it was done doing what it was doing, or it understood what it was, you know, was extracting the context from what you gave it, then I bet that output would look a lot closer to your traditional documentation, if that makes sense.

[00:21:24] George B. Thomas: Yeah. So, so I love a couple of things. One, it is very much a recipe, a little bit of X plus Y equals the, you know, the muffin or the cake that you're trying to get. And so looking at these items in AI data sources as a spice rack of it can X plus Y times whatever equals this thing is beautiful. But also what I, what I find very interesting is that we, um, we really want to paint the picture or paint the stories from who we are and what we know about the world.

[00:22:03] And I can tell you HubSpot is not selling a blogging tool, therefore not focused on messaging, um, but they are selling a CRM and a customer platform, and there are hundreds, if not thousands of humans who are Buy the software, and don't give a shh, oh my, my goodness, there, there we go, don't give a shh, about,

[00:22:27] Liz Moorehead: And he was worried about us

[00:22:28] George B. Thomas: yeah, yeah,

[00:22:29] Liz Moorehead: us cancelled. I

[00:22:30] George B. Thomas: that, I, I pause, I stop myself, I stop myself, they don't give a shh about messaging.

[00:22:36] But what they do want is they want their freaking prospecting agent to work or they want their social media agent to work because they're literally coming at it as we're buying this platform so that we can do these six things. Um, they are loosely tied. To inbound, they're, they're not owners or creators of value based human centric content, like, they're not coming from the place where we come from Liz.

[00:23:06] And so they're not looking at this as the formula to create the best content. Um, what they want is the formula that gets them the best outcome, uh, in things that they can do inside of HubSpot, like email template creation, social media agent, uh, you know, optimization, like these things. And so if we put it under a messaging tab for some of us who love that, we'd be like, yeah, um, for others, they'd be like, I don't need that.

[00:23:34] So I'll never look at it. Unfortunately, very much how I feel like, and why we did these episodes as people look at data sources and be like, that's nerd shit. I don't need it.

[00:23:43] Liz Moorehead: Well, the challenge I have with that, though, and I, because I want to throw a couple of things out there. 1. The idea of having a messaging strategy is a lot more traditional corporate than you might think. That is not something that I came out of like inbound and things like that. That is very traditional corporate stuff.

[00:23:59] The other thing I will say is Let me finish. The one other thing I will say is that value propositions, when I think about this stuff, you know what the last thing I'm thinking of actually is storytelling. This is a very much a practical manual to make sure we're just all saying the same stuff. It is literally like when we say what this company is and what we do, we're not all saying 15, 000 different answers.

[00:24:24] It's when we say like, what are our top value propositions? We're, we're committed to compliance. We're transparent and reliable. We have uncompromising ethical practices. Dot it like it's, I'm running through one from like an example client, right? And so everybody's always saying the same thing. Now, what I will agree with you, George, is that I think a lot of times people do not realize, did you know that there's a thing called a messaging strategy that has all the same stuff in the same place?

[00:24:51] Completely agree with you on that. But I would, I, the only thing I want to be very careful about is this idea that Some of these principles are coming from these very kind of soft, fluffy bait places when it actually is very just like, if you want your sales team to all be saying the same stuff in the sales process, having a messaging strategy out of the box is not entirely unheard of.

[00:25:15] George B. Thomas: And, and I don't, I don't disagree. What I'm saying is, if you go back and count, you can't count on both hands what I'm about to say. The amount of times that we've made fun or talked about people buying HubSpot so that they could send email.

[00:25:32] Max Cohen: Idiots!

[00:25:34] George B. Thomas: or, or buy HubSpot so they can, like, do one thing. Or get HubSpot and turn it into their existing system.

[00:25:44] Chad Hohn: Oh yeah, that happens all the time. People just don't want to change sometimes.

[00:25:50] George B. Thomas: and so there's this huge disconnect of like, this place where we want people to live, and this place where we hope that they have the things that we think they have, and know the things that we hope they know, and then what is reality.

[00:26:06] Chad Hohn: But they don't.

[00:26:07] Liz Moorehead: one of those they don't. And then this is one of those things where it's kind of like, how much do we want to. Cater to like bad habits and how much do we want to bring people into correct habits? Now I know that's a much broader discussion and I don't want to sound like I'm the person trying to be on the high horse.

[00:26:28] We should all be doing these things exactly this way. Like that's not what I'm trying to say, but like if I have a bunch of tabs and one of them says messaging strategy and it says value propositions. Pain points, like certain things like that, that is not outside the realm of possibility to me, but I totally get what you're saying, George, because we are so inside the platform, we are always looking at it from this holistic humanistic kind of way.

[00:26:56] And we do have people coming into the platform who are using it and much more like. I got this because it was an email platform. I got this because I don't like Salesforce anymore as our CRM, and I would like to not use it anymore. Like, we're having people come in through more transactional avenues, and I totally get that. are running into a situation, though, where it's like, A little bit of messaging, a little bit of handholding, a little bit of anything from HubSpot here would have been helpful, but I think what is very exciting to me about this conversation is how much potential there is within this specific tool.

[00:27:30] Now, I do want to ask, go ahead, George.

[00:27:33] George B. Thomas: so let's be honest, if I had my, um, like if I could wave a magic wand, um, right under value proposition, there'd be a link to a HubSpot Academy video teaching the what's whys and hows of

[00:27:47] Chad Hohn: hmm. Mm hmm.

[00:27:48] George B. Thomas: dopest value proposition for this box. Underneath the what pain points does your company solve, there'd be like the dopest piece of content that, and by the way, this might be something that other agencies are listening to.

[00:28:00] And if so, and you have something like, uh, supered, uh, it would be my suggestion that you would create an internal video, uh, targeting the word value proposition on said page of HubSpot settings or what pain points is your company self and create a freaking supered, uh, video to actually teach the way that this ish needs to be done to, to be helpful.

[00:28:24] Anyway, I digress. Go ahead, Liz.

[00:28:26] Liz Moorehead: You know, you say things out loud like that in front of your content strategist who is now just gonna be like, so George, unrelated to anything, could you whip up some videos for me walking through each of these tools?

[00:28:37] George B. Thomas: Yeah, maybe, maybe

[00:28:39] Max Cohen: sounds like George is tiptoeing down into the super industrial complex. Are you sir?

[00:28:45] George B. Thomas: I, I, well, I've, I've, I've, I've, I've talked to them multiple times. I'm, I've talked to them multiple times. I haven't pulled the trigger on it yet. Um, but I think there's some interesting use cases for it in the future, maybe, but not why we're here. 

[00:29:00] Liz Moorehead: Okay, so let's flip the script here. We've talked a lot about what we like, but I would be curious to hear if any of you all have concerns. George, let's start with you.

[00:29:09] George B. Thomas: I don't, well, I mean, my first gut response is I don't really have any concerns with this section in general. But I will say that I guess I might have concerns that people won't do it right, but we've already kind of been alluding to that. Like, listen, AI garbage in garbage out. Right? So if you just slap some like, Oh, I think that's good enough in here.

[00:29:33] It's probably not good enough. And the reason I want to bring that part up too is because this is one place in the AI data sources that we don't have a character limit. At least that we can visually see there's nothing saying that has to be under 5000 words, which means to me, you can get rich and robust, which means you can add a lot of specificity in context to the things that we're talking about here.

[00:29:57] Um, so I've just, I would hope people would take the time. To figure out things like where my brain went, the, you know, when we were prepping for this is what's the formula for the dopest value proposition for an AI agent, what's the dopest formula for, uh, explaining the pain points that your company solves for an AI agent.

[00:30:16] What's the best formula to create a description, uh, that AI agents really understand the who, what, why, and where of the service that you're providing when you put it in here for your products and services, um, and taking time to Do this right again. I don't know if that answers your question, but that's what my brain goes.

[00:30:37] Liz Moorehead: No, I like that a lot. I mean, there is something to be said for the fact, and this is something I've experimented with outside of obviously the HubSpot tools, but with other AI tools, I've noticed that the more, because I have to use all of my messaging strategies for my clients and I build GPTs that help are, that are constructed on those and there, it has forced me in very smart ways to rework some of the things that I do because you can't, some, if you make things too complex, it gets contradictory.

[00:31:10] You may not always realize like you are overloading, like more isn't always better when it comes to the context that you're giving AI tools. So like I found like it's actually helped me refine things just me personally where it's like, well, maybe I don't need 10, 000 words to say we're good at stuff, right?

[00:31:29] Like, it could just be we're good at stuff, right? But, no, I completely agree the fact that the, I did notice that there were no character limits on it. I found that kind of surprising, but I also understand, you know, maybe it's because they understand that people have value propositions documented and they may be approved at many bureaucratic levels that cannot be unapproved.

[00:31:51] So it's, it's a little bit of column A, column B, but this is where the concern I have is that education. Like I know they have blog articles about value propositions. I know they do. I've always been fascinated by the fact that they don't integrate linking to knowledge based articles or blog articles that they have in some of these tools, but maybe cause they go out of date.

[00:32:09] I don't

[00:32:09] George B. Thomas: Oh Liz humans don't like to read

[00:32:12] Liz Moorehead: I know, but I'd least like to know that they're there.

[00:32:15] George B. Thomas: Yeah, I I don't disagree I I was kind of making a funny but um, yeah the See HubSpot does is to attract Even a lot of the education HubSpot Academy is to attract we need, we need, this is why I brought up Superd, we need content that actually, uh, is, you know, I'll be, shut up, not why we're here, uh, just, yep, do that for the rest of the show,

[00:32:49] Chad Hohn: The old flatline.

[00:32:53] Liz Moorehead: Concerns, grievances.

[00:32:55] Chad Hohn: I mean, I've been thinking about, uh, you know, just a lot of where this is going is like, you know, we're talking about how like they give us a few certain things in certain spots and like, you know, we were kind of talking about earlier. Um, when you. You know, have something, should we have like a fallback or like some sort of a mechanism in case people don't do it right or the way we're expecting them to.

[00:33:23] And like, at least when you're engineering some sort of integration or like architecture of something that talks to something else, it's like all the time I scope out things with. People and they're like, Oh, yeah, well, this data should always be here if the process has fallen. I'm like, Yeah, what about when it's not?

[00:33:43] And they're like, Oh, well, it should always be there. And if it's not, then there's a problem. I'm like, Yeah, but like, what about when it's not? it's not gonna be there more than you think it will, if it relies only on a process to get it there or something along those lines. So like, we I think some of the reason that it's architected the way that it is, is to make sure that they have something in the right spot.

[00:34:12] Rather than just to be completely missing huge chunks that they might absolutely critically need for the architecture of the solution, right? As a whole. Um, so I don't know if that's like a thought or a concern. That's just more of a thought, like, than a concern. I think my, you know, I know this is like AI is getting so much attention, right?

[00:34:34] But I still have a little bit of a concern that like, This could go a little bit the way of, like, HubSpot projects in certain parts of it, like, I mean, could it just, like, get stuffed under, under a settings tab, under your, you know, click on your face, and then all of a sudden it's AI Data Sources, and then it's there since 2017?

[00:34:56] George B. Thomas: Which, when you say that, what you mean is it ends up with cobwebs and like two inches of dust, that's what you

[00:35:01] Chad Hohn: yeah. And like, it just stays this way forever and they don't go further. Cause it, you know, either the project got abandoned or whatever. And I mean, like, I really doubt that that's going to happen because of how much attention AI has and how huge of a focus it is, but like, there's the potential for that in the event that like.

[00:35:21] Some of these things are like too abstract and like, there's not enough, again, I think going back to like the handholding with people through it to understand what to put here, even just simple stuff, right? Like simple. This is what's gonna, you know, like action, input, output, sort of helping people see the picture because like right now they see fields, they don't see like, um, a vision for what putting good data in this will help you do outside of like, it will affect your contact content agent.

[00:35:58] I need to get a voice mod that will be Liz's, Liz voice.

[00:36:02] George B. Thomas: Oh,

[00:36:03] Chad Hohn: It

[00:36:04] Liz Moorehead: be as amazing as I am.

[00:36:05] George B. Thomas: yeah,

[00:36:06] Max Cohen: affect your content agent. What do you mean?

[00:36:09] Chad Hohn: anyway, so that

[00:36:11] Max Cohen: I, it's um. Do all these agents require you to complete all of this stuff before using them?

[00:36:18] George B. Thomas: Well, so, so the, uh, products and services one, it literally says you have to have one product or, or, uh, one product because they don't call it. Well, yeah, they

[00:36:28] Liz Moorehead: I really hope if you're on HubSpot, you're selling something.

[00:36:31] George B. Thomas: and you have to have well, no, but I'm for the agent to work. You have to have at least 1 in place.

[00:36:37] Max Cohen: Okay, good. My concern is that people would ignore this and just start flipping the agents on, right? But it sounds like you've got to at least have something, right?

[00:36:46] George B. Thomas: least got to have something for it to start to work because again, it needs that context.

[00:36:51] Max Cohen: yeah, um

[00:36:54] George B. Thomas: could look at, uh, data sources as like AI agent onboarding.

[00:36:59] Max Cohen: Yeah, that is, yeah, space, that's pretty true. Um, yeah, I mean, I've got less concerns because, you know, um, I was always scared about AI lacking all of this context and just leaning on shit it knows from the internet to be able to create content and write stuff and, and all of that and, you know, cause it, I mean, it kind of goes back to the, Argument that I would have with, uh, you know, customers when it came to like hiring, a partner for content creation.

[00:37:32] I was always very skeptical that some outside agency would be able to come in and understand your customers and your product and your industry the way you do. So I always thought like, Hey, it's really important for subject matter experts within the company to actually be the ones generating original content.

[00:37:49] Right. And then, you know, when, you know, the whole, oh, hey guys, right. And your blog posts now stuff came out and people said, Oh, see, it's so easy to create content. And then I was just like, Well, I mean, I really, really doubt that like the AI is going to understand your business and your customers and your process and your goals very well.

[00:38:08] Now we're taking a step to really alleviate that, um, you know, that fear, right? And just like anything else is a baby step, right? But I think it's a pretty big baby step to be honest, when you look at all this stuff together in, in, you know, You know, summation, right? So, I'm less concerned. I'm more hopeful, um, you know, that we're gonna arm AI with exactly what it needs to act like a team member and not just like a word generator.

[00:38:35] Um, and I think that's cool.

[00:38:37] Liz Moorehead: I have to throw something out there that I think George is going to be proud of me for saying, because normally I'm, I'm the little butthead who's like, but content is art, you know, all of these different things. And it is right. Like I, I've had extensive conversations on here about like, you gotta be in the driver's seat, right. And this is something that is not a conversation that is just specific to this tool, but you said something Max that I find really interesting about the idea of like outside parties coming in and being able to tell your story better than you can.

[00:39:08] Max Cohen: mm hmm.

[00:39:09] Liz Moorehead: You have a and I've heard about this like you can't outsource your content can't do this.

[00:39:13] You can't do that. Oh, that's malarkey No, but here's why I want to speak to that fear at a high level max. I'm not saying that's what you were saying I will just want to share this because it kind of hit on a point that I think sometimes we miss if it requires Someone to literally be inside of your organization in order to speak to the soul of it about who you are, what you do, who you're built for, and how you do what you do better than anybody else.

[00:39:38] What your value propositions are all these things. If you don't have the ability. To articulate that to a third party outside of yourself in a way that makes sense that can be repeated and shared. You have larger problems as an organization.

[00:39:53] Max Cohen: Yes, I think that's true, but I think more so when I was a kid. So like i'm not I I would never discount the ability for a third party to come in And be able to tell that story about a company, right where it gets tougher for me Is them creating the content for the folks out there just trying to solve for a specific goal or challenge that they had Right.

[00:40:17] That don't even, they're not even in the framework of learning about a company and learning about what they do and who they are. Right. They're just going, Oh, my something, something is wrong. And I'm trying to figure out how to solve for it. Right. When I'm in that mode, I don't give a, I don't give a about, you know, who, who a company is and why they are who they are.

[00:40:39] And like all this other, you know, stuff, like I'm just trying to find an answer to my question. Right. And. That to me, I think like the nuances of doing that really well to me, it was always, I always felt like it was like a little bit tougher to like, expect the third party to come in there and really be like a subject matter expert on the goals and challenges of your customers, because that company that built a product to solve for the goals and challenges of the customers probably have a much more deep, intimate understanding of what those things are.

[00:41:09] Right. Um, then a third party that's not. Well skilled in that right now. Yes, I I can agree You can you can communicate that to people and have them create the content and everything, right? um But you know it just totally offloading it to them and just saying like oh, yeah Like you know create which is like kind of what happened in the past Hopefully it happens a little less but you know i've always just kind of been someone to say like listen when it comes to that true like Awareness stage, uh, you know, that, that true, like awareness, consideration, you know, top of the funnel content that is all like purely educational and really pointed and targeted, like the nuance goals and challenges that the type of people you're trying to attract have, I always, always want subject matter experts who are passionate about that shit and understand that stuff.

[00:42:01] Deeply involved in that content creation process. Right. Um, but that's just me.

[00:42:07] Liz Moorehead: We need to have the outsourced content conversation on this podcast. I don't think we've ever done it because I think that is, I'm going to say one thing. We're going to put a pin in it because George, I want to come to you to come to you about best practices, but to put a pin in it, I think the outsourced content conversation is a breakdown on both sides.

[00:42:24] Because I think in some ways most content shops or agency offered content services do not do the work to integrate themselves into organizations the way they should. So yeah,

[00:42:38] Max Cohen: I agree with you. And it's more, that's not me saying there isn't an agency that can't do it. I'm, I'm saying the ones that are really, really good at that are very, very rare and probably super expensive, right? To have that detailed and skilled of a process and nuanced understanding of it, right?

[00:42:57] The run of the mill people that are going to come in and be like, top, top 10 things you need to know about company ABC and have that be your content is just like, you know, bonkers. But anyway.

[00:43:07] Liz Moorehead: Yeah. George, I want to talk to you really quickly. I want to hear from you because you have actually built this out for sidekick strategy is your agency, right? You've gone in, you've done all the tools, you've done all the build out. And I'd love to hear from you. What are some of the best practices folks should keep in mind while filling these pieces out?

[00:43:25] You've alluded to formulas. You've alluded to all of these different things. I'd love to hear. Best practices and maybe some rookie mistakes folks can avoid as they're going through this process.

[00:43:35] George B. Thomas: Again, I'll, I'll say something and I'm going to say it in a little bit of a different way than I usually do this. And by the way, Liz, as I'm saying this, my brain also is transporting me back to when you say. Some organizations already have these things because it's part of the fabric of who they are.

[00:43:54] And so if you have these things, great. Um, look them over. Make sure they're not outdated from when you built them seven or 17 years ago because your business changed the world changed like So let's, let's make sure they are really what they're supposed to be for today and then put them in, put them in.

[00:44:12] Um, but where I'll go if you don't have that and you're trying to get this in here is one, you don't have to do this alone. Um, so many times I've watched super admins just be super admins and think that they have to do this alone. Um, it's almost going to pair with the content, uh, outsourced content.

[00:44:31] Interview some of your top sales folks, your SVP, your, you know, marketing CMO, interview them, ask questions, get the transcript, like, start to craft this at a human level if you want, interview them, get the transcripts and work together with an AI assistant to be like, Hey, if I take these interviews and I try to build a robust value proposition that is going to be used in HubSpot, take screenshot, paste in, boom, like help me create, use this framework that Hub Heroes gave me and the information that I've taken from people in my organization and help me craft this draft that I can now take to the next level.

[00:45:13] In other words, don't do it alone. Like, tagging other humans, tagging AI, um, because yes, between humans and AI, you can use those to actually do what we're here to do, is program the AI that is HubSpot, the version of HubSpot that is AI, um, it doesn't have to be a huge hurdle. Like, Liz, when I, when I crafted these, like, 30 to 45 minutes, boom, done, loved it, and Moved on, but also came back and said, hey, this is what I created.

[00:45:47] Is there a way to make it better? Okay, let's figure out a formula now. Let's update it. Right? But at least get something that's good in there. Have a conversation with humans in your organization to get the information that you might not have if you don't have it. And then work with a GPT, a clod, a perplexity.

[00:46:09] I don't, I don't give a crap. Um, to then expedite or streamline or speed up, uh, getting this into HubSpot. And by the way, part of this might even be just collecting what you already have. With the products and services, hopefully you have a products and services page. I literally just had, uh, my very programmed, trusted GPT, like, look at the page and say, give me titles and descriptions for all the products and services that we provide, and then enhance them, and then did a little bit of copy and paste, and again, 15, 20, 30 minutes, products and services were in there, so it didn't have to be this massive hurdle, uh, like some of us feel like this, these things might be.

[00:46:54] Chad Hohn: hmm. I think this takes me to the place where like I think people who, you know, are, about whether or not they need to be, like, using AI in their day to day. in any capacity. I mean, like the part of the reason you just like, Oh, here's an image and tell me to fill in some things, you know, it's because like, you've been just like cultivating and caring for your AI assistant for quite a while, right, George, like you've just been like working on it.

[00:47:20] And this is what we do and hone your craft and it assists your business. And like, it starts to feel like people, you know, like, well, it's like you're like a little Tamagotchi or whatever, right? It's like a

[00:47:31] Max Cohen: playing cat playing catch with it outside

[00:47:33] Chad Hohn: Right, playing catch with AI. Anyway, but it's like, if, if you're not at least starting to do that, then, you know, you're just going to be starting with blank GPT threads all the time.

[00:47:45] And like, every time you got to tell it exactly what you want. And it feels a little bit more laborious to deal with like your AI assistant. But I think, A really exciting place in my brain. And like, you know, this is just again, seeing the vision of the future is like pop open co pilot on the page inside HubSpot and ask it, Hey, I don't have a marketing strategy, like, you know, to get to the point where it's going to help like walk you through that, because like, how cool would it be to.

[00:48:15] To like, talk to it about what you need for this page that you're on in your AI settings to help train your AI assistants and your agents in HubSpot and your content with what will be most impactful, especially on what they're wanting, and then ask it to help you, you know, maybe to get that information from your team.

[00:48:38] And, you know, it'll give you some questions to interview so that it has the data that it needs as like part of that interview. To, you know, bring that back home. So, I mean, I think that's like the, you know, almost, um, I'm almost hoping that like inside a HubSpot co pilot is going to have some of that so that everything related to your business, you don't have to like, so manually go after all of this context input into your model, like, you know, George has been enjoying doing all this time.

[00:49:10] George B. Thomas: I hope they get there. I don't know if it's there yet, but I hope

[00:49:14] Chad Hohn: Right. No, I hope they do too. I think that's the vision of the future that I would love to see. Right.

[00:49:19] Liz Moorehead: I love that. Max? What about you, bud?

[00:49:22] Max Cohen: I feel like I kind of uh Gave it all away my last answer. Um

[00:49:28] I mean, I

[00:49:29] Chad Hohn: it all away.

[00:49:30] Max Cohen: i'm sure

[00:49:31] Liz Moorehead: I should have saved some. I should have squirreled away for

[00:49:33] Max Cohen: yeah, I gave everything I had for it um Yeah, I mean, I'm just generally excited about the direction that it's going. I mean, it's there. I think they're clearly being very thoughtful in the way that they kind of like brick by brick build this whole, you know, AI.

[00:49:48] I mean, whatever kind of breezes today, I can't even imagine what it's going to be like in the future. Um, but I, it, it, the, the thing that makes me happy about it is that they are really getting away from. This is an AI easy button to here's how you build. Here's how like you, you truly support an AI strategy within HubSpot.

[00:50:15] And here's all the things that you can do to make sure you're using it the right way. And not just something that is just supposed to be, you know, spitting out a tons of high volume, low quality stuff to make it look like it's this. Giant time saver. It's more so like, how are we turning it into a creative tool that has the context it needs to operate like a true teammate versus some hyper productivity hack, right?

[00:50:45] Um, It's, it, it, it, they're, they're, they're, they're building it with care and they're, they're building it in a way that people can use it but not use it the wrong way, I guess? I don't know. I'm, I'm I'm just glad it's not like, I mean it's just very clear it's not gonna be this like flash in the bucket AI feature thing. It's gonna be a very, you know, good approach to how you should be using it within the context of HubSpot and the greater strategy it supports, right? So I'm, uh, I'm very optimistic and I'm starting to watch it closer than I ever have before, so.

[00:51:25] George B. Thomas: mean, I've said it before, I'll say it again. I think HubSpot is building the world's largest business agent.

[00:51:32] Max Cohen: Mm. 

[00:51:33] Chad Hohn: Yeah.

[00:51:33] George B. Thomas: So this, this whole thing that we've been talking about, all of these settings inside of a setting, I don't think they'll end up like projects because they're too dang important to make the business agent be the best business agent that can be.

[00:51:48] Max Cohen: Yeah.

[00:51:48] Chad Hohn: Agreed. Mm-hmm

[00:51:49] Liz Moorehead: Well, George, as Chad alluded to, I think once before, it was very difficult for us to land the plane without you that one time we were left to our own devices. So I'd love to just turn it back to you and have you help us officially land the plane for the last time in 2024. What do you want listeners to take away from this episode?

[00:52:07] And I guess quite frankly, the whole AI data sources series, because we just wrapped that up. very much.

[00:52:11] George B. Thomas: Yeah, 1, uh, take some time to chill. Happy holidays. Happy New Year. Merry Christmas, whatever. Just take some time to reflect on the journey that you've had in 2024. Definitely look ahead for what's going to happen in 2025. Maybe, uh, set some goals and some strategies and all those good things that you should do as far as 1 of those goals and strategies.

[00:52:33] Go and spend some more time with, uh, data sources. Go think about things like value propositions, ideal customer profiles. Um, think about the agents that you want to turn on or use. Think about should you be creating educational value based content if you're not. Um, think about the experiences that you can create in 2025 because, let's be honest, At this point, that's what we humans care about, is experiences.

[00:53:04] So, go enjoy, go focus, and we'll see you in the new year.