From SEO to GEO: How G2 Is Redefining Their Growth Strategy | Belkins Podcast Episode #15

Speaker A: In the next couple of years vibe coding will be easier. Even more. People without technical background can create their own product. Right. Companies like Salesforce or HubSpot or will just, they provide the platform for it.

Speaker B: And so they like OpenAI ChatGPT, they already have a GPT store. But maybe the future it's people sharing their own AI agents.

Speaker A: Yeah.

Speaker B: If in five years software buyers or buyers of any product or service, if all they do is go to ChatGPT, they go nowhere else. Right. That would obviously be very detrimental to our business.

Speaker A: Like the marketing is about how to turn your scalable machine into the authentic machine. Right. You kind of make a mistake these days because the mistake costs you an account. Right now we literally cannot uh, afford and cannot hire junior talent anymore just because of the high quality delivery of what we need to do for clients.

Speaker B: Yeah, no, I think you're describing an interesting shift. Michael.

Speaker A: My CMO mentioned that whenever she posts something like a, like a one feature product on AI or something or the quick like AI fix, it gets traffic and people are kind of going, so whatever you do that has AI part in it, it's got traffic. But whatever you do that is a core problem. It's not, it's less popular these days.

Speaker B: And I do think every business person now, you know, whether you're a solopreneur or you're working for a Fortune 500, I think it's impossible to not be both curious and I think there's also some fear around AI. AI won't replace humans, but humans with AI will replace humans without AI

Speaker A: Goddard Abel is the co founder and CEO at G2, the world's largest software marketplace. Today the platform lists more than 200,000 products and services, hosts over 2.7 million verified peer reviews and attracts 90 million software buyers every year. Investors have bagged that traction with over $257 million in funding and at $1.1 billion unicorn valuation. Before G2, Goddard built and cash out of two category defining CPQ companies like Big Machines sold to Oracle for roughly $400 million in 2013. And SteelBrick, acquired by Salesforce for over $360 million in 2016. Those back to back exits cemented his reputation for spotting market gaps early, scaling fast and delivering enterprise grade outcomes. So I would like to open up with the following. Um, do you agree that uh, for the past 20 years uh, we've been um, living in time of uh, kind of software builders where it has been extremely, extremely difficult to build the software but they're relatively easy to market. 1 and now with kind of AI, we've switched the seeds in a way that now building a software is easier but then marketing and selling it is like 10x more difficult. What do you think about this?

Speaker B: I think it's true. And I've been building enterprise software companies for 25 years and I do remember 25 years ago is super hard where we had to even buy some servers. Oracle database licenses, application servers. You have to really build the whole spec and uh, write all your code in Java. And yeah, uh, it was obviously a very arduous process. And now with AI cloud it's easier than ever to vibe code spin up any app. And we see it on G2. G2 has become the largest SaaS. Ah, software marketplace. There's over 200,000 listings today of all types of software, about 2,000 categories. So now there's purpose built software for every vertical, every department. And so I do think that makes it hard now for entrepreneurs to break through that noise. You know, there's new categories, you know, let's say like AI note takers for example, and there's a great one granola I use or Zoom has them. You know, there's literally hundreds of options. And so I do think now the challenge more for an entrepreneur developer if you're building an app is how do you break through the noise. And uh, I think that's, you know, kind of an interesting new challenge versus you know, building the code building infrastructure. All of that's gotten much easier.

Speaker A: Yeah. Um, how many did you see the spike in new apps registered on a platform in the last couple of years than every year? Is there any dynamic in terms of how many new apps are kind of popping up on the platform comparing to the previous year now with kind of vibe coding and AI being open to everyone?

Speaker B: Yeah, I would say it's been kind of exponential growth in the last year. The number of listings on G2 is growing 23%.

Speaker A: Okay. Yeah.

Speaker B: With new ads.

Speaker A: Right.

Speaker B: You know, that's probably about 40,000 new listings.

Speaker A: Mhm.

Speaker B: And we also have 50 new categories now around AI software. Mhm. Obviously including the AI infrastructure, things like vector databases. Obviously nobody before ChatGPT 3.5 probably cared about. And obviously now some of our newer categories are things like agent builders, AI agent builders, AI orchestration to orchestrate multiple agents. And again those were high categories two years ago. Nobody would have even imagined. And so I think, you know, we're seeing tremendous growth in what we call our GQ taxonomy, both in terms of software categories as well as individual listings. And I think it's obviously wonderful if you're a software buyer, you know, which probably every knowledge worker is, because now you can also not just find a software app, you can probably find an AI assistant to help you do just about any job. And so it's a really exciting time for entrepreneurs. And I think there's just a proliferation now of apps and enterprise AI tools.

Speaker A: What happens with kind of apps that change their profile? Like, they really want it to be kind of an AI application, but we all know them for this traditional sort of like SaaS or software, like, for example, uh, you know, Apollo or let's say, you know, Samsung Sales Loft, per se. Right. So if they're like, now we are, um, like AI software. Uh, right. So they had an AI and then kind of they wanted to push the AI forward. So do they still, you know, do you guys, do you think that it's fair for them to be in the same kind of engagement, you know, platform category, or they should be moved to the AI engagement platform or subcategory of the AI products in a way, if that makes sense.

Speaker B: Yeah, no, uh, and I think you're describing interesting shift, Michael. And you know, you mentioned the category of sales engagement. And actually our CMO at G2, Sydney Sloan, she used to be the CMO of Saleslot. And actually our VP of Product Marketing, Liz Daling, she used to be the VP of product marketing for outreach. And those were obviously the two leaders in the world of sales engagement.

Speaker A: Yeah.

Speaker B: And actually remember watching that category grow over the last 10 years. And I knew both entrepreneurs, you know, Kyle Border, who is building sales block, Manny Medina, building outreach. And I remember for years they're also an arms based on G2. You know, who's going to be number one, who's going to have the most views in sales engagement. And I think you're right. Now both are reinventing themselves. I think there's been a huge wave of now AI agents, uh, being built to drive outreach. And so we also have new categories now around AI Sales assistants.

Speaker A: Right.

Speaker B: And I think both vendors, Sales Loft, Outreach, and probably everyone else in that category has now also launched agents. And you have virtual BDRs to drive your outreach for you. And so I do think that's an interesting shift. Now in software, we're all the big incumbents. And in the category of sales engagement, it was and is outreach from SalesLoft. Now they're all innovating, introducing new AI first apps and agents on top of their platforms. And then of course, you have new startups coming in that are building AI first AI only. And I think what'll be interesting to see and it'll probably be different in the category, there won't be one winner. Right. But who wins? And I think obviously a lot of VCs are betting that in every category of software, these AI first, you know, companies are actually going to emerge and win. And actually speaking of which, you know Manny Medina, the founder of Outreach, I just saw him in London last week. He's now building a new AI first startup called paid AI. And you're doing a whole new startup around AI monetization. And so it's an interesting time where I think there's innovation by all the incumbents. And of course Salesforce and I used to work at Salesforce. Right. They're also reinventing now. Mark Beniaf has an amazing vision around Agentforce and Salesforce customers building all their agents on AgentForce. And so also all the big incumbents, they're going all in on AI and so it's just this massive arms race of startups coming in trying to redefine and all the incumbents also building on their platforms, building on their data and adding the AI layer. Uh, but yeah, obviously it's a great time to be a customer because I think whether you're using an incumbent or use a startup, all this AI automation is being built for you.

Speaker A: So just uh, so I understand it. Um, so instead of kind of moving from a category to a category, let's say I'm still in the sales engagement, um, category. But what I do, I built a product on um, top of my platform. Like a multi product type of company.

Speaker B: Right.

Speaker A: And I lease them under the new category. So now I'm kind of competing into both categories in the same time. My, my core platform is a sales intelligence platform by this AI assistant that I developed as the part of, you know, the multi kind of multi product. I'm just putting in listing into the different category. Is that the case?

Speaker B: Yeah, I mean certainly we're seeing a lot of that in G2. I think that's what the incumbents are doing. Right. If you have a strong sales engagement or strong SC platform, you're still going to sell that. Right. And you have thousands of customers on it and then you're layering a new product, an AI, uh, agent product on top of it and then making both available to your customers. And I think that's also the reality, you know, most of the customers, uh, I think neither and probably both models will coexist. Right. Most companies still have human sellers, BDRs, SVRs at the same time for certain use cases. And we're doing that at G2, you know, maybe for your simpler inbound queries that you know would have been given to your most junior BDR in the past. Now the agent can handle the simple inbound BDR SVR cases and probably handle them better. Right. Because they're available 24, 7 answer within seconds. And then you can on data, uh, obviously also score your leads, route them differently because if you see, wow, it's a best fit ICP enterprise lead, then maybe you hand it off to a more savvy human seller right away. So I think right now the two are coexisting for the customer and I also think most vendors are doing both. Of course, if you're an AI first startup, you know, you started in the last couple years, then you're only going to be in the agents.

Speaker A: Right.

Speaker B: And uh, you know, so I think right now we're seeing tremendous innovation both by the incumbents and new startups coming in.

Speaker A: You know, what I'm trying to understand is like, does it make sense to kind of reposition the entire thing and just move to the new category? Um, or it's like too risky. And you always can do through kind of a multi product or like building up additional layer. Right. I was just thinking about the example of the Jaguar, the automobile company, right. Where they kind of introduce only like electric vehicle and change everything and kind of the model. Uh, and they lost like 95% of their customer base or something. So they're like kind of selling like 10 cars in the last year. So they dropped like 98% of their sales. Right. Because they essentially repositioned the entire thing instead of introducing um, a multiline product. Right. Like okay, we have those. So have you seen kind of successful examples from the folks, you know, in the industry that sort of like repositioned and then moved only to AI? And then that actually helped them to move faster, more agile rather than kind of creating this multi brand thing, which again is kind of difficult I think.

Speaker B: Right. Honestly, I think it depends on the scale of your existing business.

Speaker A: Mhm.

Speaker B: You know, because on the extreme, if you take Salesforce. Yeah, it's impossible. I know very well. But if you have a $40 billion existing business.

Speaker A: Yeah.

Speaker B: And you have 200,000 customers that are in your existing CRM platform, I think obviously both the customers also don't want you to turn that off and go AI only because most customers aren't ready to go AI only either. So I think the reality is if you're a big incumbent, you have a big customer base, big revenue base, then I think you really have to do both. And again this example of Salesforce, we go all on agent force. And then I think it's also interesting they've introduced new flexible pricing.

Speaker A: Mhm.

Speaker B: Where you're still paying for the traditional CRM apps, sales, cloud, marketing, cloud per seat. But then now you can get credits for agent usage. And then I think, you know, I think what Salesforce and other big incumbents, I think you know, same thing with ServiceNow where obviously ServiceNow traditionally the leader in ITSM platform. Um, but now there's also. Yeah they're also building the AI platform to automate all these workflows and for itsm and I think they're also giving flexibility now where you can build agents on top of ServiceNow platform that can handle your ITSM use cases. And I think so I think that's both the opportunity and challenge for the incumbents. Right. They already have the thousands of customers.

Speaker A: Right.

Speaker B: And they can now add agents on top of it, but they have to support both. And I think there are some smaller startups like one where I'm an angel investor, it's called Vivint and happen to know the founder Matt Darrow, he was our sales engineer at big machines like 15 years ago and he started out probably five years ago, let's say 2020 building. Originally it was like a pre sales automation platform obviously built more of the traditional SaaS app and now with AI he's all in on Vivint, uh, on agents and actually more expert sales assistant agents. So kind of the next generation of the sales engineer if you will. And so but they were still small so they've pivoted all into you know, now just selling the AI first biven solution and I think it's working for them. So I think if you're small you could probably still make the full pivot, you know, but if you have a lot of customers then you have to keep serving your customers and on the traditional SaaS platform while you also build the AI agents that your customers want.

Speaker A: Fair there? Yeah, uh, I think like when you are in like upstream enterprise, um, you know, kind of mature market. Right. Like you know it's very difficult to move from one technology to another. So if like I always say I am a very heavy user of HubSpot so in case you provide me an AICRM that is free, I'm not going to probably take it. Right. Because like it's Going to take me so much time to move the entire over thing to a new Systems. But if HubSpot interviews something, an AI addition to what they have, a new product, I would love to use that. Right. So I think to kind of to your point that most of the kind of businesses that are not heavy relying on the core technology, they can probably switch easier like small businesses. Right. But when you go to a bigger company then it's um. Yeah, the kind of the software providers, they expect to provide an AI addition to their core technology. Right. But we are not kind of moving from, from the core product. Right.

Speaker B: So and HubSpot is another great example. I know their CEO Yamini, you know, they just launched their Breeze AI.

Speaker A: Mhm.

Speaker B: And you may have tried it. Right. But where you can use ChatGPT now natively within your HubSpot using all your data and all your apps. And I think that's a great example of an incumbent innovating and sounds like that's also good for you as a customer. You can keep using HubSpot trm and you already have all your customer data there and you know, but now you can layer on their AI and uh, so I think, I think there's a lot of that also happening.

Speaker A: Folks, if you're enjoying the interview so far, I would appreciate if you spend a second like comment or subscribe to Belkin's podcast on the platform where you're watching this. I'm trying to get my show listed in the best podcast and a couple of seconds of your time would really help me a lot to do that. Thank you so much. And let's get back to the interview. You know what? Recently um, I discovered with HubSpot that I never know before. Um, my co founder, he's doing a lot of like wipe coding and he kind of say okay, I wanted to enhance our uh called, you know, the template writing game. So I built a, you know, I wipe code, my own kind of template writing AI kind of product. And what HubSpot allows is to get the app market listing through API very quickly so without a lot of hustle. So he was able to just get that product integrated in our HubSpot to talk to entire system in a way easily. So they kind of allow this uh, you know, easy access to their kind of marketplace where you can, you know, just quickly list your API and get your product integrated in a suite that you already have on a HubSpot without again going through ten hells of uh, again I don't know whether the other products are doing that, but that was like, really cool. I was like, okay, that's smart. So now you can build your own product, add it to the suite and just use it. So if I have a problem, let's say, how do I, let's say ping my sales teams when, when it's time to follow up on the deal.

Speaker B: Right.

Speaker A: Like, I want to have my own notification system. I. I found there's no good product. I Vibe code the product, I created an API, I integrate that API into my HubSpot and now I have the product, the problem solved. Spend a couple of days, some gpu, whatever. It's cool, right? It's like super cool.

Speaker B: So not only sounds like you could Vibe code, solve the problem for yourself, but then Hubs really made it easy to publish the AI, publish that tool to the HubSpot Marketplace, and then the other customers can use it as well.

Speaker A: I don't know whether you can actually publish this for the public access. Probably you can just do it for yourself. Probably. I don't know. I probably like you. If it's possible, then we're going to see thousands of apps on HubSpot and you don't know which app are used. But I think it's more like you can do it for yourself type of thing. Right. Uh, but still, I think like one

Speaker B: package your own innovation.

Speaker A: Exactly. That's right. That's right. Yeah. Because like, I think to your point, in the next couple of years, Vibe coding will be easier, even more. Right. So, um, now people without technical background can create their own product. Right. Um, and you saw lots of product being listed on G2 right now. But imagine a couple of years when it's going to be even more easier. Right. And you can, you can solve your own problem. And the goal is just to have your core technology being easily integrated with whatever you wanted to solve. So it's kind of you build your own product per se and you just kind of people like companies like Salesforce or HubSpot or will just, they provide the platform for it. Right.

Speaker B: So, yeah, now you can definitely. And a lot of people might say you can package as agents. Right. Build your own agents. And so like OpenAI ChatGPT, they already have a GPT store, you know, where you can publish and share your GPTs. And I do imagine, and you mentioned HubSpot, they have a strong partner app marketplace, also Salesforce, you know, they've always had the app Exchange, but maybe the future, it's people sharing their own AI agents.

Speaker A: Yeah.

Speaker B: You know, that they've built for their own CRM use cases and why not share them? And it'll probably be massive long tail. And that's probably true in their app marketplace. A lot of those apps may not get used very much, but even if it can just help another customer or two, you know, why not, right? And then some of those derivative works may even end up being huge hits where all of a sudden you're not even consider yourself software developer, but you build some cool add on to your HubSpot or your Salesforce, you publish it and all of a sudden thousands of the customers can use it too.

Speaker A: Yeah, yeah, that's cool. That's interesting. Um, how much interest is in terms of uh, leads traffic eyes in the AI category comparing to the other categories on the platform? Uh, the reason I'm asking you this is, um, we obviously are all trying to get attention and traffic and generate leads, right? But my CMO mentioned that whenever she posts something, uh, like a one feature product on AI or something, or the quick AI fix, it gets traffic and people are kind of going, so whatever you do that has AI part in it, it's got traffic. But whatever you do that is a core problem. It's less popular these days because it's not maybe as sexy or whatever. So when you kind of look at the more kind of traditional categories and the ones that are now growing with AI, what's the ratio in terms of the interest right now? Is it like 5020 or 5050 or just. It's still. The AI is still not that popular as the traditional one.

Speaker B: I think all the growth right now is in the AI categories. We're probably seeing that in GT where I think traffic to the AI uh categories is up over 100% year over year and obviously also working off a smaller base. But I do think that's where all the growth is. Whereas maybe legacy SaaS apps are kind of, I would say sideways or even slightly declining. Mhm. Because at this point this is known, we were Talking about traditional SaaS, uh, CRM apps and people tend to know there's HubSpot, there's Salesforce, there's Microsoft Dynamics. I think those markets are pretty well known. And I do think every business person now, you know, whether you're a solopreneur or you're working for Fortune 500, I think it's impossible to not be both. Curious. And I think there's also some fear around AI. You know, if I don't adopt AI, my company may get disrupted, I may get disrupted in my job. And I do like the saying And I don't know who originally said it, you know, but I think AI won't replace humans, but humans with AI will replace humans without AI. And so I think it's also incumbent upon really every knowledge worker right now and there's over a billion in the world, right. I think, and I'm trying to do it in my own job as CEO. I'm trying to reinvent myself, right? And obviously like on my browser I have tabs for Gemini, ChatGPT, Claude Perplexity, you know, obviously experimenting with all of them. And it's also part of my job. But I also, if I make me much more productive, new pricing as well, you know, but even mundane tasks like sending my board a follow up email, a summary of my meeting, you know, AI crushes that and you know, where I used to probably spend 30 minutes on it, now I can get it done in three minutes. And there's just so many of those use cases, whatever. Your job is where you can drive a lot of productivity, automation. So we're certainly one. I'm trying to do it myself to learn and get better at my job and also encouraging all our employees at G2 to do it. And I think the whole world's doing that now. And so there's tremendous interest in people wanting to learn about AI. And I think the other thing that people are trying to learn is like, hey, what really works? Because especially I think people outside the Valley are also skeptical, right? There's all this hype around AI and it's magical. But then like, hey, what really works? Right? What works for my company, what works for my industry? Uh, which AI use cases should I do first? So that's where I think there's just a ton of curiosity and people looking. And it sounds like that's also what your marketers found, right? If, if you tie into that AI wave and help educate people on AI and how to use it, then there's a lot of interest. Um, and on the flip side, yeah, if it's not AI, then right now people probably aren't that interested because it's not new.

Speaker A: Yeah. Would you recommend folks, uh, with kind of established brands and established software, uh, to start earning, uh, and branch out like an AI product and start kind of competing in that new category, kind of going back to the roots and uh, sort of like establish themselves in those new categories. Would you think it's going to be,

Speaker B: it's a smart move generally. Yes, of course. It has to make sense for your customers, you know, because also doing AI just for the Marketing, obviously that doesn't make sense. Right. But I also think now with AI tools, it's hard to imagine an application or an industry where AI wouldn't help.

Speaker A: Mhm.

Speaker B: You know, but I think always in entrepreneurship the key is, you know, figure out what do your customers want, right? How can you help your customers more? And usually there's a way. Kind of like what you mentioned earlier. You know, I think even little hacks like oh, make it easy to publish my app to the HubSpot app store. M. You know, I think even if it's, you know, but I think there's always ways to take out administrative effort even maybe it's on the maintenance side of your app, you know, versus on the front end. But we sort of see the G2 and the same thing you mentioned with the HubSpot listing. We're working on this right now. Like why don't we automatically create your G2 listing for you? You know, because AI, we can look at your product description on your website, we can pull information and then we just have a draft on the marketplace for you. And what used to maybe take you a couple hours, you have to set up your D2 listing with your demos, with your screenshots, with your product descriptions. You know, all of that AI can just draft for you. So rather than taking two hours, you know, now it can take you two minutes. And you know, that's just kind of. And I know in our own product and our own G2 Marketplace platform, I think there's literally like hundreds of things we can do to make our customers lives easier, you know, to add an AI assistance to take away some of the tedious work, you know, that nobody likes doing anyway.

Speaker A: Do you remove a uh, longer hyphen when you are um, editing with ChatGPT and other AI, replacing it with like the other thing, you know, like the symbol, like longer hyphen symbol that shows that this was written by AI or edited by AI.

Speaker B: You know, this thing or no, um, I'm not sure I know exactly what you mean.

Speaker A: So when you edit something with like chatgpt or something, they always put like long hyphens on the, you know, on the editing that shows that you never do that. Like most people don't ever use kind of the double hyphen. So when you see something with the double hyphen you're like, oh, it's probably going to be edited with AI. So I was wondering if you noticed that or not.

Speaker B: I haven't personally noticed it and I'd have to actually ask Our product managers, if we're. Yeah. Because we do allow also for reviews. You can use AI writing assistance because it helps you write better. But I'm not sure how we're dealing with double hyphens and.

Speaker A: Yeah.

Speaker B: And generally I'd say it's better to be authentic and transparent.

Speaker A: Right.

Speaker B: And I think using AI assistance is okay.

Speaker A: Yeah.

Speaker B: You know, but, uh, But I'm not sure if we're addressing.

Speaker A: It's like when you go to LinkedIn newsfeed, you see all these dubbed hyphens everywhere. So, like, two years ago, no one was using them, and now you kind of have them everywhere. Just check it out. You notice it.

Speaker B: Okay.

Speaker A: So you're like, look at this. Like, oh, it's all been written or rewritten or at. Which is fine.

Speaker B: It's a telltale sign.

Speaker A: Right, Exactly. So I was just wondering, like, maybe we just reached out to Sam Altman's team and just say, hey, can you just remove the double hyphen because it feels like. Or why you specifically put it there? Because now we start noticing. And this is actually an interesting trend. Do you think we, as people start kind of picking up these small signals, like symbols or way that something is written to tell you whether it was written by human. On AI. And that would actually, um, kind of affect the positioning of a company. So we would feel like, oh, if it's an AI, it means that they outsource this, or like, they try to automate, they try to be more efficient, they try to do less, more with less or something. And if it's going to be more kind of human, creative, we're going to say like, okay, it's kind of handmade, it's craft, it's something like, what do you think about this?

Speaker B: Yeah, no, and I think it's actually getting harder to tell. I think I. Yeah, because I think on ChatGPT first got popular a couple of years ago, it had. It was less good. I think it. Because I'll say I was getting better at assuming personality. You know, you can have it right in your personality in different tones. So honestly, I think it's getting harder to tell. But I do also think the same thing you said about software apps. I think AI is so much proliferating content that I do think the content also won't break through the noise, you know, if it doesn't have some unique human creativity, some unique human touch. And I do think we can somehow intrinsically tell, you know, maybe it's also, like, what's entertaining. And I do think that's true. And even B2B right. Because people, they do want to be entertained, they want to be engaged. Right. And I think that still takes human creativity. Whether we can take out tell automatically, you know, with a double hyphen or. But I do just think as humans we just have a natural sense for what's interesting.

Speaker A: Yeah.

Speaker B: What's engaging, what's entertaining. And I do think that bar is much higher. Right. And I think everyone's seeing it because a traditional strategy. But Also I think HubSpot made popular, you know, with Inbound where you would just write to different business topics and they did a great job of that for two decades and probably had the best marketing blog. But all of a sudden, you know, just kind of putting out how to content.

Speaker A: Yeah.

Speaker B: Probably isn't that effective because now if you have an how to question, you can just go to chat GPT. And I was amazed, right. Even like I love riding bikes on weekends, that's my hobby. Mhm. And now when I have any question on like how to fix my bike, you know, even. And I was kind of amazed. I was just trying to replace the battery on my power meter but I put in I have a specialized Epic and but it just nailed it right away and said oh here's the exact spec of the battery you needed and by the way, here's the instructions of how to replace it.

Speaker A: Yeah.

Speaker B: And uh, but I was just amazed you know, because it's like, seems like kind of like a long tail bike. Not even a lot of people have it.

Speaker A: Right.

Speaker B: And but I think almost any topic now if you have a how to question, right. Like odds are Chat GPT will crush it. So I think so if you're going to write your own content, I think the bar is much higher now and I do think adding human personality, human creativity, uniqueness.

Speaker A: Mhm.

Speaker B: That's what's going to engage us. Right. Because I'm sure it's the same for you if it, if you can kind of tell it's just bland AI content then you don't need to go to someone's website to read that. If you want basic information you just ask ChatGPT yourself. Mhm.

Speaker A: Yeah. Uh, yeah, I was, you know, when I always was telling uh people that the best way to prepare for the AI, uh is to start generating a lot of unique insights, a lot of kind of proprietary information, something authentic, something that only you can have. It could be any insights from conversations with customers, any data you generate, anything whatsoever that would bring something to the table that never Been you cannot just access that in a how to type of, you know, library. Right. Then by packaging that you first gonna create a authentic unique content.

Speaker B: Right.

Speaker A: Like thought leadership, something that only you can have and it will position you very high with like AI and other LLMs because now you are more searchable than other because you can provide the information that no one else have. So now you're unique. And also your customers see that because they kind of just simply go and chatgpt this. Right. This is only you have that. So like the marketing is about how to turn your scalable machine into the authentic machine.

Speaker B: Right.

Speaker A: And how I can now create more of us that is built on what we do rather than is built towards how we can scale that to more kind of traffic, type of lead, type of, you know.

Speaker B: So yeah, I think you're right. Kind of having your own personal take and I think also related to that, I think building a personal brand.

Speaker A: Mhm.

Speaker B: You know, so then people want to hear from you, you know, some unique perspective and uh, you know, some unique perspective on the world. And if you build a trusted brand, people are like oh, I want to hear from Michael on this topic, you know, your trusted voice on the topic. Then I think people will still engage with your, your content. Right. Because a lot of his angles about who's written, who wrote it, uh, and you also see that with reviews on G2.

Speaker A: Yeah.

Speaker B: You know probably what's even more important in the words that are written is like who's the person and do you view them as an expert on the domain? Right. Do you trust them?

Speaker A: Yeah.

Speaker B: And I think that also is I think very important to break through the noise. Right. Because I think yeah, just generic content without, without a personality behind it, without a trusted brand is probably not going to be engaging.

Speaker A: Yeah, we probably going to go back to the times of uh, like uh, corporate blogs that are very well done, like authentics or like content, something you really wanted to read, follow, kind of subscribe, type of newsletter. You know, something that uh, you know that it's not just some copywriters are pushing it out, but actually kind of people and talking about brand, um, you guys notice that kind of there is a like a brand, um brand value right now in the world. So like if you have a brand authority, especially with like Google, then you are ranked much higher and generally you are more positioning towards like okay, you're like an um, um, authorized or like a expert source. So whatever you do your position higher than the others and if you are not, then it's Very difficult to get ranked position. Have you did your um, marketing team at G2 reported on something like that?

Speaker B: Yeah, no, we have definitely been looking at, you know, kind of what people uh, are also not calling geo Generative Engine optimization, which is an analysis when ChatGPT or Gemini, when they give an answer.

Speaker A: Mhm.

Speaker B: You know, what is the LLM, um, trained on? What's the basic answer on? Obviously sometimes they give you citations, that makes it easier to tell. But there's also an interesting new startup called Profound. I don't know if you've heard of it, but they specialize in geo and I think that's what most brands, most marketers are not trying to figure out. You know, if someone asks ChatGPT a prompt, a question related to your brand, how does it answer and why does it answer that way? And I think back to your original question. We certainly see a G2 and there's an analysis done by the Washington Post on Google AI. You know, what are the most and you could actually look it up online, like what are the most cited sources or what tokens have people trained on the most. And we did see GT was in the top 250 of all websites.

Speaker A: Yeah.

Speaker B: And so for software, business software topics, we do have a lot of influence on DLL manuscripts and I think partly it's related. And we also have good page rank in the traditional SEO world and domain authority. So I do think it's related to how your site ranks in SEO and I think a lot of that is also related to your brand and your domain authority and your trust. And I do think the LLMs are, you know, more they're also, well, uh, just like they did for the purpose of SEO and page ranking, they're kind of trying to figure out, hey, what's your domain? What are the topics you're viewed as an expert on? And then you're more likely, you know, your content is more likely to be weighted in their training. I think one of the interesting about LLMs, and we've all heard about Reddit and Reddit, you know, they did a partnership with Google, strong partnership, part of which was Google also having access and licensed all the Reddit conversations to be able to train their Gemini and their other AI models. And I think we were also talking to Reddit, but I think a lot of what the LLMs want to train on is human conversations and maybe it's back to the authentic voice. Because I think the last thing AI wants to train on is synthetic AI generated content. It wants to train on real human conversations. And so I think certainly sites like Reddit that have a lot of human dialogue are doing well. Um, and I think also G2, you know, it's user generated content, it's real authentic reviews from real users. And so I think that's the type of content that LLMs want to train on. And that's probably something else for marketers. I think the more you can make your content and really LLMs, in some ways it's all back to the Socratic method. Obviously the Greek philosopher Socrates, he taught everyone by Q and A. And really if you look at AI, that's what it is, right? Prompts or questions, you get answers but you have back and forth dialogue with Q and A.

Speaker A: That's right.

Speaker B: And so I think, and I think what the LMS are craving is really human conversation, ideally with real experts. And I think that's, you know, then if you're a brand or marketer, if you have a lot of good authentic human content from experts and it's fresh content, then that's what the are also going to train on. And uh, you know I think one of the things we're actually doing at G2, we did just acquire a startup called On Survey and what they enable, they basically have agents that will interview you. And that's we see as a future of software review. It won't be a static web survey or web form anymore. It'll just be a conversation like you and I are having. And then with AI we can parse that conversation. But now we've launched a G2AI interviewing agent and that's what's going to be asking you about the review. You know, it'll ask you, you mentioned you're a HubSpot user. They'll ask you, hey Michael, how are you using HubSpot? What use cases are you automating? Are you using their new AI tools? What ROI are you getting? What challenges are you having? You know, but it just has a good conversation and it's interactive and traditionally for CM, um, reviews we would have had a survey with 40 structured questions that honestly is not fun for anyone to fill out. Whereas now you can just have a conversation, an interesting conversation about how you're using the HubSpot platform. And then at the end AI, ah, just parses it and you know, says hey Michael, here's a drive for View. You can edit it, approve it, publish it. And so but I think all of that is also going to create more conversational content which we see it'll be better for other software buyers that will learn how about HubSpot, because they can learn from your conversation. And then also frankly it'll be much better content for the train on. And so I think, and this is probably true for any website now, any content publisher or even just as a marketer, how do you run your website? I think having more conversational human expert content. Yes, I think that's, that's going to be the key to make your content stand out.

Speaker A: Yeah, but also like how other authority sources are referring to you. Right. So let's say if you are listed on G2 and how G2 is talking about you, the other ones like so essentially you're building up your own kind of ecosystem of sources that are kind of referring to you that are also our authorized sources. And in this way they kind of lift you up to okay, you're also an authority source.

Speaker B: Right. It's a good way to do your geo, your generative engine optimization. And you're right. Even participating in conversations on Reddit, for example. And Reddit now actually has brand pages.

Speaker A: Mhm.

Speaker B: And Obviously conversations on G2, conversations on LinkedIn. Yeah, but I think the more you're in online communities where people are talking about your products, uh, talking about your brand and then authentically participate in a conversation and spurring that conversation. Yeah, I do think that's going to be the best way to make sure your brand stands out, you know, then through the voice of your customers, through customer conversations, through user generated content, getting that message out there. And so I do think that's kind of new evolution and there's a lot of similarities that was in the SEO era, but also a lot of differences.

Speaker A: That's right.

Speaker B: And I think the main difference is yeah, you need more authentic human conversational content because the traditional, you know, blogs and we did this at G2, we created like a learn blog 10 years ago and honestly it no longer performs very well. You know, just publishing how to content just, you know, it's not, it's not interesting anymore in the age of AI.

Speaker A: You know also what's interesting, um, the people behind it. So um, um, the articles or content, conversational content that is featured with the writers or people in the industry like me and then it's tagged correctly that I wrote it or I posted it and I have a previous experience and backtrack that I write about this actually help to have a higher reputation. So those pieces of content that are wrote consistently by the writers or people in the industry that been around and they have a long kind of tale of the other publication that on this space are ranked much higher than by that are written by someone that is outside this industry or this is not their scope or they never been kind of published. So um, I think it's also kind of we're going to see the future of the guest post right now. So if I am a expert on let's say B2B lead generation and if I write my take or something on G2 blog or something, right? So it means that sort of like we are connect in terms of the authority source, right? Because now you're like, okay, these guys that know what they're doing are writing there and there. So we're going to, I think we're going to see like more gas blocks. We also going to see people more caring about who they write for whom they write and what they write actually. So that. Because now there's like um, a history of everything that you wrote. So the more it's to your point of personal brand, but not just as we think about this personal brand for founders, right? Like oh, I'm the founder, but literally for every person in the organization, you know, what you're out, what you post, where you're at, where you're speaking because that's kind of create that branding score for you. And then the more people in your organization that are involved in this, the overall the brand and the reputation score will grow for the entire organization for you in this industry per se.

Speaker B: And I totally agree that like who's publishing the content and then every one of your employees, you know, building their personal domain expertise or personal brand on certain topics, obviously it's gotta be authentic. So it's gotta be topics uh, that almost everyone in tech, we all have topics, areas we're passionate about, where we're experts and then yes, you know, also online establishing, sharing our expertise and building that brand. And I think you're right. And also I think there's data on LinkedIn where you know, nobody really engages with brand content. And we see this@ uh, G2, you know, G2 we do have a lot of followers on our G2 corporate page. But honestly if we page some, if we post something from just G2 corporate, it probably gets you know, 25% of the engagement versus if I posted our CMO, Sydney, posted our CRO, Eric, you know, so those of us that have personal brands in the industry, if we share it, you know, we get much more impact. And obviously we also put our authentic verse personal voice on the industry, our personal context, then it gets you know, probably 10 times the engagement, right. And because people don't want to hear from brands, right? I think we want to hear, especially in the age of AI, we want to hear from real humans, we want to hear from real experts. And so I think you're exactly right. And I think probably even for anyone's career now, you know, I do think it's critical that you know, somehow you establish your own fine brain. And I've also told like my sons, they're uh, you know, we have twin boys, they're at the University of Colorado, just finished their sophomore year. But I also told them early on, like, hey, you know, at A M minimum, build your LinkedIn profile. That's right, build your X profile and start sharing on topics, start building on your networks. Because there's really never, it's never really too early, right? And it's a compounding effect 100%. And I've seen this now. I've been doing it in some way. I've been entrepreneur 25 years. I've been LinkedIn for 20 years. And you know, if you continuously build your network, continuously share your content, years of expertise and over time you develop a following, you develop some domain cred in your areas of expertise. And I do think it's essential now otherwise you're not going to break through that noise. In the age of AI, we've even

Speaker A: noticed that um, if we engage um, um, with the content writer to write on certain topic and if it's a known writer who has been creating in this space for a longer time, when we post a publish something of that person, then they're going to go rank very high within like two, four weeks if that person, uh, doesn't have the authority name in the space. The ranking of this article went down. So uh, just a tip for listeners who can even kind of negotiate with the writer, the byline. So if you give them the byline, you sort of like you pay less for them because they are all wanted to get that byline, that they wrote it, that they published it. Because that's just helped them a lot to kind of establish authority. And the more of their pieces of content are ranked in very high position, the more business they can generate, right? Because now they're like authority source, like whatever. What I post as Joffrey or Samana is actually getting a lot of attraction even in the ranking. So it's like, it's literally like Google and Geminis and LLMs. They also kind of track about it and think about this as well. So just a tip.

Speaker B: Yeah, no, I think that's great. Advice?

Speaker A: Yeah. Um, I asked Chad GPT, you know, why you have, why G2 has a strong brand and what's it it's based on? Um, do you want to Hear what the ChatGPT says?

Speaker B: Sure.

Speaker A: So it says, um, the answer is that it's you guys are consistent, um, you provide consistent value. Um, and um, also you have, you guys do unbiased feedback and then you have a community around the brand. So it's a community, it's unbiased and it's kind of verified by, you know, by buyers and the vendors and the sellers. So that kind of live conversation. So these three things actually build JITU in a great brand. And I think to just, to kind of follow up to our conversation just now, probably that's applicable to all other companies who wants to build brands in this space. Right. So do you agree with ChatGPT or there was a secret sauce to all of that that you guys establish a good brand?

Speaker B: No, that really resonated with me because I think it said one, we're consistent, two, we have a community and three, it's unbiased. Yeah, it's unbiased, trusted, authentic content. And I will say no, those are all, I mean it's all intentional. And I will say though it took time, especially consistency because we've been building company for over a decade and I do remember and that's a challenge in building a community, especially with user generated content. And G2 is also a marketplace for matching software buyers and sellers. And I'd say if you're an entrepreneur, it does take time, you know, because honestly for the first five years we didn't have a very good business so we had to be patient. And uh, you know, so at the beginning we didn't take shortcuts because we wanted to stay on authentic content. And, and obviously some, you know, there's like some software analyst firms that will just kind of write paid content and which might accelerate your revenue curve. But I think we've also always believed authenticity is one of our core values and we have our peak values. But we always said, hey, let's make our content authentic. And even if a vendor is willing to pay you a lot, like hey, we're not going to just move automatically to the top of the rankings. And so we had to be more patient in terms of then getting enough brand buyer traffic so we could actually start to monetize. Uh, but it's obviously nice to see ChatGPT recognize that. And of course we're in no way done right. It's very much day one at V2. And also we're very much reimagining now for AI, you know, because we are probably when we started we kind of built maybe the web 2.0 era. Uh, at the beginning we called it Yelp for software. And now we're also realizing, wow, you know, there's massive opportunity and threat, you know, as AI comes around. Because also our challenge could be if in five years software buyers or buyers of any product or service, if all they do is go to ChatGPT, they go nowhere else. Um, and there's no attribution, right. That would obviously be very detrimental to our business. Um, and so that's also something. But on the flip side, as we also were talking about all those LLMs they do need data to train on. And I think also I think perplexity started a good trend where they're doing more and more citation and being more transparent around the sources. And so I think anyone that's running an online community, online site that provides content like G2, there's I think both threat and opportunity now in AI. Uh, but I think I'm glad we have that good foundation that we do have a good community. We have trust, a reputation for dust with unbiased content. But obviously we have to scale that now in the age of AI and as we were talking about throughout this conversation, we have to make our content richer, deeper, better for the age of AI and more authentic and more by human experts because that's, I think that's what's going to stand out, especially in this AI age.

Speaker A: Uh, my co founder who is a big, big fan of your work, um, asked uh, to ask you the following, um, how do you guys going to compete with kind of LLM search and kind of retaining the business at G2 where like um, you know, people can again go and search as you mentioned, like for the software in there instead of paying X subscription 20, $30,000 to you to, instead of kind of pay nothing. Like what's the, you know, what's your um, generally talk track to kind of, you know, to kind of battle that challenge?

Speaker B: Yeah, no, and I think uh, overall I'd also say we don't want to compete hm. You know, with AI search and LLMs because honestly I think anyone trying to compete with ChatGPT, I don't think it's a winning strategy. And I think they also realize that you know, we're one and we talked about these themes earlier one day need the most trusted content to train on mhm. And so, and I think Right now and as I mentioned we're kind of a top 250 data source, that citation source. So one we want to make sure we keep providing content that Chat GPT and the LLMs do train on and I think over time it will provide more and more citation and that obviously still creates a lot of value for our marketing customers. You know if we can influence buyers through and that's I think what I call geo, you know, generative engine optimization. If marketers realize their reviews on G2 are the number one training set for AI search for the LLMs and obviously that actually increases the importance of a G2 review because now it's also the number one trainer for the LLM answers for the ChatGPT answers. And secondly I think what all the LLMs are also seeing like ChatGPT, they're also going to call other agents and you uh, know G2 we've already put on app on the chat GPT GPT store and now I think with this whole MCP infrastructure we're also adding an McP layer to G2 and maybe the future, you know I think a lot of people are saying the future is agent to agent. Mhm. And, and I think even the LLMs, you know, potentially they're going to be getting their training data through MCP instead of through web crawling. So we're also adding an MCP layer so that you know, as it goes more agent driven other agents, whether it's ChatGPT Gemini, they can learn from us. And then I think also people are envisioning a future as are we, that agents are going to be called to do specific jobs.

Speaker A: Mhm.

Speaker B: And I think if we can have deeper data on software stacks, software pricing, then you know the OpenAI agent may call our agent and you know exactly how this works. I don't think anyone's sure but I think our vision now is to be the trusted source of data in the age of AI. And uh, I think most people are envisioning agent agent future where there's certain expert agents for to do certain jobs and we're positioning G2 to be that agent. If you really need deep expertise on software and ratings and grids then you know other agents will call our agents. AI search will reference G2 and so you know I think like I said we're, we're aiming to be complimentary, the most trusted data source that then complements the lms, the trends to AI search as well as providing agents and then helping our marketers also create selling agents that can answer all the Buyer questions. And I, uh, think what's also interesting, you know, I think all of these AI LLMs like perplexity, we're also having actually direct partnerships discussion with them because they're also trying to create differentiated experiences. And it's actually something Perplexity is already done with TripAdvisor or hotel search. If you go to TripAdvisor and ask it like, hey, what are the best hotels in Boulder? And that's where I'm living now. If somebody visits me, you can just ask Perplexity, give you the best hotel rooms in Boulder for under $200 a night. And what's interesting, it'll give you their LLM driven answer. But it also actually does direct attribution of TripAdvisor and will actually embed the TripAdvisor reviews. And so I think that's also something we're discussing with Perplexity and ultimately we do with all the LLMs where we have a unique joint experience, you know, where our content can be surfaced through the LLM, both direct attribution and with richer information then you get for a generic LLM answer. And so I think this is something the whole industry is probably going to figure out, you know, where the. I think the core LLMs are also going to be referencing expert sites, expert agents, to actually provide better, more engaging answers to people prompting on their AI. And I think that's ultimately arms race. You know, it's interesting and it is a bigger arms race. Right. ChatGPT is trying to, and I think right now they're leading where they have like 800 million users.

Speaker A: Yeah.

Speaker B: With growing engagement. But obviously, Gemini, Perplexity, Claude, they're all trying to win. And I think ultimately they're going to compete on providing the best experience to their users. And I think they're also all recognizing, wow, they likely need some third party partners in specific domains that have agents and data better trained for particular use cases. Yeah. Because the world, there's millions of use cases. And so, uh, that's kind of what I see as a future. And that's the future we see G2 fitting into. You know, if we can have the ultimate software expert agents, then core LLMs, other agents will call our agents in that particular domain. And so ultimately I think it'll be a complimentary world. And we're aiming to, like I said, be the AI expert agent and the AI expert data source that can be accessed through mcps so that as the world moves to AI first, you know, we're still the most relevant source of insight for software.

Speaker A: So if we are living in the world where there's a buying, um, AI buying agent, then I would probably wanted to buy that agent from someone like G2 who knows how to kind of buy the software instead of trying to train my own agent or using a third party agent. Right. So I'd probably go to someone that knows everything about how to buy a software and get an agent, buying agent who will help me to this. Right. So, um, so you launched the agent and this is something that I also wanted to ask you. So who would you use an agent like that and what could be the buying process and what is the timeline for this? If you think when this kind of actually becomes popular, what do you predict?

Speaker B: Yeah, and we do have a buying agent available now on G2.com um, and we also launched it, we've launched in parallel at G2AI. You can actually go to that domain, you know, if you want an AI for shopping experience for software. And I think, you know, on G2 both are available now. And I think the reality is the majority of users are still using more the traditional web search and the traditional G2 interface. But we do see rapidly growing use of the G2AI interface and it's growing much faster. You know and I think at what point that tips, I'm personally not sure. You know, it's like with the whole Internet. I think if you look at it, Google SEO search is still overall much more popular and obviously now it's also being blended with the AI summary in search. Um, but I think it's an interesting kind of question. At what point does a tip, you know, at what point do Most buyers at G2 engage with the agent first, then the conversation versus a traditional web interface? Um, and if I were to guess, I'd probably say like I said, the AI conversational shopping method on G2 is growing much more popular. Um, but I think the web interface still probably has 90% share. Um, but if I were to guess, I don't know, in two or three years it probably flips and probably in five or ten it's truly AI first where everyone's just starting with an AI conversation and it might still be the human doing the prompting and then when it goes to an agent is calling our G2 agent and there's no more human in the loop that I don't know. You know, it's probably, probably a long ways off.

Speaker A: Yeah. So what do you think would be the first use cases like I am, uh, there's like hundreds of note takers and I just wanted to say, okay, it's not a very big deal for me. It's a $50 subscription per month. I just set it up as like, okay, here's what I'm looking for. You know, me, just, um, go and get me the note taker. So I'm gonna start with this, you know, um, with the software that is kind of, uh, there's so many options out there, I just don't have time to go and talk to them. And it's gonna be like a, like a low ticket size for me as well. So I guess this could be the first use case or.

Speaker B: Yeah, and you could try that on G2AI right now. But you just ask it, hey, what's the best AI note taker? And then we'll actually ask you to sign up. Because if we have more context, you know, like the size of your company, your job function, your role, then we can, our AI can give you better recommendation the more context it has on you. And that's probably true with all AI. And I think we'd probably ask you to sign up, hey, sign up with your LinkedIn.

Speaker A: Yeah.

Speaker B: And then, um, you know, then we know a lot about your professional context and we can make better and better recommendations. But yes, and I think traditionally it would have been on G2. Well, the traditional process pre AI, you would have gone to Google, you know, typed in best note taking, so where. And then just go from there, right? We would have taken you to a category page on G2, uh, you know, where it listed by ranking the top 10 and you would have read about it. We had filters, you know, but I think obviously it's much easier with an AI query where you just say, hey, what's the best AI note taker for me? And if you've already signed up with LinkedIn and we know enough about you, then we can very quickly make, you know, very good bespoke recommendation. And, and obviously also give you the pricing. Hey, yeah, it's only 50 bucks a month. This one sound good to you? You know, you guys ask me some more questions. Hey, tell me more. What are users like me saying? You know, and what's, you know, granola versus zoom native notetaker. What are the pros and cons? Um, I have an idea.

Speaker A: Have you thought of it? Because when you talk about like buying for me, it's like, oh, should I just give my credit card? They will buy, they will charge it. I don't know, what if it's like a procurement or evaluation, um, agent. So it's like uh, for example I set up my stack like all the tools that I'm using and it's always searches and sort of looks at what's option and it always kind of send me different options of what's available, what is coming up right now in the market, what is a good alternative, what is trending right now as well as sort of like helping me to evaluate what is next to help me to set up the best sort of like you know, so it's like oh, I have these three tools that uh, I'm paying X. But you know, here's this new trending tool that is combining the features from all three of there and at the same cost. You should check it out. So it's always kind of going and searching and evaluating and procuring and kind of. But not making the purchase but essentially recommending me. You know, where is that? I would love to use something like that.

Speaker B: And one thing you try on G2AI is we have a new canvas.

Speaker A: Mhm.

Speaker B: Which is, it's probably a lot like the OpenAI canvas we've seen. You know, where now when you prompt and you're let's say creating a document, it pulls up a separate canvas where you know, which is kind of updating. And you could do that especially because we also see that use case in buying software it tends to be collaborative, it tends to be a buying committee. It's not just one person. And now with G2AI you kind of start that canvas like hey, I'm looking at note takers or I'm looking at CRM software for my enterprise. And the G2AI will obviously start the canvas but then you as a human can edit it, you can share it, you can refresh it with the AI. And I think that would be a great way to kind of do what you're describing, where you're just kind of keeping notes on three apps, you're getting updates on them, um, you're editing it, you're adding your own notes, you're pulling in your colleagues so you can all collaborate on this canvas, you know, around apps or around your tech stack and then you can use the AI to automatically refresh it. Um, and I think that, you know, I think for considered purchases of enterprise software, which oftentimes, like I said, require many people on a buying committee, like most business software apps, one person doesn't just. It's not like a control buy right, where you just kind of decide, hey, I'm going to buy this shirt. It's more considered. You have to collaborate with colleagues on it and it might take several months. And during that time, especially in AI, we see it with like the LLM rankings, right. They change every week. And there's also cool sites now like Ella Marina or the Scale AI leaderboards. Um, and I think it's the same thing with now with AI software. It does tend to change week to week. And so I think you make a good point like how do you then refresh it? But that's our vision with our G2 canvas where you can create your own notes on products considering shared with colleagues and then automatically refresh it with AI. Uh, and I think like I said, the canvas concept we didn't have to invent. You know, all the LLMs have it now where you can collaboratively create documents, edit them, collaborate with the AI and collaborate with other humans. And so that's uh, a use case. We're really excited about that. We've already launched in uh G2AI.

Speaker A: Um, do you remember those kind of early days of when kind of ChatGPT first came out 3.5 and then it kind of shaked the world. Uh what were the kind of early concerns that you had, uh either about the business or generally about the AI that uh, has never come to be that you were like okay, we're not, you know, we're not in that world. And then are there any concerns that actually um, came true and now you're like oh shit, uh, something like that. You remember anything?

Speaker B: Um, well I think the concern that has had an adverse impact on our business is you know there's less click throughs through Google.

Speaker A: Mhm.

Speaker B: And especially I think a lot of that is you know, with the AI summaries now that Google puts it at the top of the page and the reality is most buyers, um.

Speaker A: Do I what do, do you hate it as well? Do you hate this AI summary at the top?

Speaker B: Well, I don't know. I think it's useful. You know most of the time it's useful. I think one thing I wish Google did at least last time I looked at it, they don't make it easy to interact with it. You know unlike ChatGPT like to ask a follow up prompt on the AI summary. And I think Google's doing that intentionally. My theory is they still want you to click on all the surrounding links. You know, they don't want you to just engage with the AI. But uh, I think the concern that's valid there for us and for all websites. I think there's about you know, 20% fewer clicks now. On Google search to any of the surrounding links.

Speaker A: Yes.

Speaker B: You know, paid or free. So obviously that's also been adverse impact for V2 where we're getting less clicks from Google, which was our number one and still is our number one source for new users. So I think that's the fear and concern that we still have. And of course on the flip side, my co founders and I, the opportunity we saw was, wow, we can build like what we were talking about earlier. We can build G2AI, we can build buying agents, buying assistants. I think what was frustrating about G2 for ourselves, we tried to build software recommenders but they were never very good. Mhm. So you still had to just take people to the CRM category. You show them the rankings and then yes, you could filter them. Hey, show me the rankings for SMB versus enterprise versus mid market versus vertical. But it was sort of a clunky software discovery experience. And obviously now if you start an AI like we were talking about, hey, just tell me the best AI notetaker, tell me a little about yourself. You can have a much better shopping experience for software and you no longer have to know which of the 2000 categories to shop in. And our industry has been kind of wonkish, you know, and like we all remember like account based marketing, that's been categories have been created over the last decade, you know, but it gets super confusing because now there's like account based execution, account based marketing, account based selling, right? All of this which is all. If you're like a marketing or sales software nerd, you sort of love it, you love nerding out about it. But if you're just like, yeah, I'm just a cmo, right? Especially I'm not in tech. Like I'm just trying to grow my business, right. I'm just trying to improve my marketing. Like why are you throwing all this category jargon at me? Um, and I think that's the exciting opportunity. Now if you're that cmo, you just go in and say that, right? Like, hey, I'm trying to improve my marketing effectiveness, my marketing roi. How do I do that? And then you have a conversation obviously to ask you, well, please sign up. What industry are you in? Oh, you sell B2B. Oh, it's a very considered purchase. You have to sell to a buying committee. Great. And then ultimately they might surface. Oh wow. Did you know there's these account based solutions now, you know, like $0.06 demand based that make it much easier to surround that buying committee and we'll make it much More efficient. But you know, but the category comes much later once you have the business problem, the business context that that software buyer is trying to solve. And so I think that's what we're really excited about. I think it'll be uh, just a much better shopping experience for software. And you no longer have to know the categories, you no longer have to scroll through top lists. You know, you just get bespoke recommendations through an AI agent based on the same data we had in kind of a web format. So you know, so we're excited about the opportunities and still you know, aware of the threats. And we talked a lot earlier about GEO and I guess the other solution to SEO declining and that's also been exciting for us and now in our sales pitch we have a kind of a side by side slide to be like hey we're number one in SEO but that's declining. But more importantly now we're also number one in GEO and LLM influence in all these software categories. And that's the future. And so and I think that's probably true with AI most companies like there's some adverse impact for us it's SEO declining but then there's an opportunity, wow, look at the GEO influence we have. Look at the great agents we can create. And so I guess entrepreneur it's also, I tend to be more glass half full but obviously you also have to be aware of the you know, glass half empty threats. And I think AI for G2 presents both and I think for most people our industry presents both.

Speaker A: Yeah, I mean like we're so align with you seen the same thing like the SEO traffic or like views have been growing, click through rates have been down because of the AI assistant. My team really hates it. So like why do we need like the entire page like you can't find anything, all the good links are below and then you cannot run ads now. So it also impact that strategy. Right. Because you know your ads are not as visible. Um, but then yeah the LLMs our traffic are growing so the um which is exciting and I love that you mentioned that definitely the future would be some kind of partnership citation, some other things. So it's not going to be as it is but to you know shout out to Google like you guys definitely need to move faster because you're going to lose a lot of good business from folks.

Speaker B: Right.

Speaker A: And true.

Speaker B: Um, as a user like if you compare it to perplexity, you know like a pure AI search experience and I think and Google understand it right. They have ah, I don't know what $200 billion quote unquote legacy business. Like best money printing engine ever invented, you know, Google SEM ads.

Speaker A: Mhm.

Speaker B: Uh, so I understand why they need to protect it. But you're right, it's sort of a weird hybrid now, right? Like a big AI, uh, summary that's not very interactive and then you still have all the traditional links surrounding it, but it's kind of like, hey, which is it? Right? Because it's not optimized for either. You know, you don't get that traditional search experience that maybe many people still like. And you also don't get the best of an AI conversation or an AI search. So, uh, but obviously Google has a lot of smart people, so who knows, they'll figure it out.

Speaker A: Um, I think the thing that they've done that I would probably don't do it is they're trying to optimize both the back end and the front end, right? So when you're optimizing the ranking and every marketer is already shaking like, okay, our pages are not ranking as high, what is the keywords, how they got. So they still wanted to figure that out first, but then they also introduced the front end with like AI assistant and summary. And then everyone's like, oh shit, so what's happening here? So then now they're like, everyone is like, okay, how do we optimize this? What do we do? And then actually my SEO lead, Eugene, who is like amazing talent, I was so happy to hire him back in the day. Uh, so now because of, and thanks to him, we're killing it in the, in the ranking space. But he said like, Michael, I'm probably going to take like a vacation for a year or two years until this thing settled down. Because like, there's so many changes in the, like the hundred pages that I'll just literally like, don't need to you know, just kind of chill, go somewhere and go back, understand how it works and then we're going to figure this out. It's like, I know, Eugene, you cannot take the vacation. No, let's just kind of crack this, right?

Speaker B: So, yeah, no, it's interesting. Maybe we should connect them to Ross Briggs. He's our VP of Buyer experience and he also owns our SEO now, our GEO. And now, yeah, I think it's for traditional SEOs, it's kind of wild because now it's this whole world where they also have to figure out the LLMs and it will take a while to settle and they're all experimenting, right? Every day it's different. And do they add citations? Don't they add citations? Expectations? And then obviously, as they train their models, as they launch a new version of their model, you know, the training sets are different. And so, yeah, it's kind of an exciting, dynamic world. But I think as an entrepreneur, I think for most entrepreneurs, it's kind of fun. You know, change is fun and it's scary. Like, there's all this disruption. Wow, I might go out of business. On the flip side, it's like, oh, wow, look at all this cool stuff we can do with the ad.

Speaker A: Yeah, but, but the budgets, you know, like, like, it's one thing, like when we are experimenting and the other thing to deploy a budget on a paid ad or content, and you're like, do I, do we deploy the same amount? Or we optimize or we change something? So we said, oh, let's wait out. And the worst thing is, like, when a lot of businesses are waiting out these days, like, oh, we don't know what's going to happen. It's like uncertainty.

Speaker B: Let's not do that.

Speaker A: Exactly. And we like. I mean, marketing. I hate when people are facing the spend like, you still need to build a business.

Speaker B: So not clear yet what the paid models are with the LLMs.

Speaker A: Right.

Speaker B: You know, how are they going to embed ads? How are they going to have sponsored content? And somehow they will. And I think, like, people are like, oh, will AI disrupt ads? And then I've kind of thought about it. I'm like, nothing has ever disrupted ads. You know what I mean? Obviously, I mean, ads have never disappeared. They always change format with new media, with new interfaces. Right. But the concept of advertising, no tech revolution has killed it. You know, it's just kind of reinvented it. And it'll be interesting to see how ads are reinvented and sponsored content is reinvented in this age of LLMs.

Speaker A: Uh, that's a challenge. Yeah. So talking, um, about AI transformation, um, have you regrouped and rethought your kind of team structure, team hiring, any kind of, especially on the leadership side, to meet the new AI challenges? If, yes, who you've been looking for, who did you bring on on board and generally, what is your now vision or, uh, position on the team structure now with the kind of AI in play and where the G2 is going to.

Speaker B: Yeah, and I think what I'm most excited about in talend we just acquired Y Combinator startup on survey, and I might have mentioned that they also became a new product squad MHM and their CEO, Alex David, is now our GM of our AI Solutions squad. And I've also heard Parker Conrad talk about this. The founder CEO of Rippling and Rippling also. They do great on G2. Yeah. Because they ship good products. They have probably 10,000 reviews. And I think most people in the blogosphere and tech have heard, you know, Parker strategy is to build compound apps. And you know, you started with payroll as kind of a core HR app, and now they probably have 10 HR apps. Now they're doing finance apps, IT management apps. But the idea compound startup build, you know, it's almost like an SAP for a small business. Like build every. Every app a business needs to run. And what's also interesting about Parker's talent strategy, he talks about, you know, Almost all his PMs are what he calls failed entrepreneurs. And I, uh, do think that can be a brilliant talent strategy right now where you take. And it's like this unsurvey team we just picked up, you know, and frankly, they tend to be younger teams coming out of y Combinator, maybe 20s, early 30s, that are truly AI first. And I'm an older entrepreneur. I just turned 54, so I probably will never be AI first, you know, like, I was, uh, I remember as a young entrepreneur, like when dot com happened, and I was like a cloud first, dot com first, you know. But now I'm not AI first. So I do think that is a great way to pick up product talent now is to acquire acqui hire m, you know, Y Combinator, other startups that are already building and as we talked about, there is so much noise. So I think a lot of those AI first startups, they actually won't succeed commercially, you know, because there's hundreds in every category that these things are funding. But I think a lot of the quote unquote failed teams are actually great teams. And they didn't really fail in terms of tech, you know, they just failed. They couldn't build distribution fast enough to get enough revenue traction to get their next round. That's a talent strategy I'm really excited about right now. Um, and if I were to pitch, if you're a quote unquote, failed AI entrepreneur, uh, looking for a new home, like ping me, because I think that's a great way.

Speaker A: Yes.

Speaker B: And you also add a whole product squad, you know, that's very nimble, very AI first. So that's a new talent strategy. And also my co founder, Tim, he's, uh, you know, he's leading product. We've gone into product mode but, but he and I are both very excited to do more of that talent acquisition through you know, small acquisitions, equi hires of AI first startups and teams.

Speaker A: You know, I would just add on this like failed entrepreneur who at least made a million dollar. So it's like. So because like I've uh, I spin off a couple of like brands and startups in the past couple of years and about 70% of them kind of failed. And the reason it failed is that I sort of like branched them out with the people who has been doing something but they hadn't tasted the real hustle. When you kind of are generating revenue and it kind of helps you to polish you as the entrepreneur in a way. So you start being good at the HR and recruitment and sales and the marketing and the finance and you're great at time management and you don't have the uh, you know, the other things that you don't. So you literally know how to deliver, how to deploy, how to get shit done. And you, you cannot do that if uh, you never earn revenue. Right. And pay.

Speaker B: I think Michael, you're saying you don't want them pre revenue, right? Yeah, want and I agree, right. They've been grinding to generate some revenue because until you have some revenue and people paying you, it's just not real yet. Right?

Speaker A: Yeah.

Speaker B: And uh, and even, and I have a lot of empathy for failed entrepreneurs. I feel like I've almost failed many times in our ventures, you know, and just kind of maybe gotten lucky and escaped. So I think there's no. Yeah, but, but I also, I, I agree with your point though. Like you want the team to have actually done the hard work to actually start generating some revenue because then, then you have kind of full empathy for what it really takes to build a product in a business. Right. That actually produces reven versus just a science project if you've never generated any revenue. Mhm.

Speaker A: And then again like failing a company when you've been to a play to a, to to a, you know, in, you've generated revenue, right. Or you've had 50k MRR, 100k, then you failed, something happened, that's fine. But you have been there, right? Instead of like, you know, being inspired or doing something or building something like a, um, um, like a phantom startup and then you know, never kind of deploying it. So um, what is your position about sort of like there are two ways of thinking about the talent strategy right now. Um, optimizing your junior talent and really focusing on bringing on the seniors and kind of um, Just enhancing them with AI agent and really being effective and moving forward quickly. Or um, sort of deploying um, a junior strategy where you kind of are getting more junior talent but then growing them and kind of enhancing them with AI. And in this way they're kind of turning into kind of the senior talent. Because like uh, for example, for my business, right, we see that right now we literally cannot uh, afford and cannot hire junior talent anymore just because of the high quality delivery of what we need to do for clients. So you cannot make a mistake these days because the mistake costs you an account client. So you're literally looking for the best people at the budget that you can get and then try to obviously optimize what they're looking for, but trying to go for senior, which I understand it's not the best strategy because then I won't be able to nurture the pool of junior talent then turn into the senior talent. And so I'm kind of limiting myself into my kind of capabilities. So, um, obviously with your resources and with your brand, like uh, you know, everyone who wants to work for G2, you have a large team. So with this AI agentic future and your current talent strategy, are you leaning towards more the senior pool? The junior pool. Devout to senior. Like, curious to hear your thoughts on this.

Speaker B: Yeah. And I think at our scale and we're over 100 million revenue run rate, you know, we use some of both. Like I just hired a new cfo, Alex Bradley, and he obviously comes on the very experienced side. Um, so I think kind of at the C level, senior leadership level, I think you still need the experience. And as I mentioned on the other side, I would say, you know, I'd say some experience. Like I said, I love the Y Combinator startup team. Unsurvey that we just picked up. Right. And to me they're probably in the sweet spot, you know, where they all, let's say more like 10 years experience trying to apply AI to a similar problem that we want to solve. Um, that to me is probably the sweet spot. And I think the other trend now in industry, G2, we're also trying to keep up with. I think everyone's saying now the mantra is increase your revenue per employee, right?

Speaker A: Yes.

Speaker B: Whereas in 2020, 2021 it was all about raise and higher. Mhm. Because kind of capital was free. And 2021 the flex was like, oh, we grew from 500 to a thousand people. Now the flex is like we stated, 500 people and we tripled our revenue. And so I think we're Trying to do more the latter. Um, and like I said in senior roles, if you're managing a lot of people, I think you still need experience. Um, and, and then I think we're also having some success. Like we're still have a class of interns at G2 in our Chicago office this summer. You know, we have a class of 15 interns and I do think we still want to feed and maybe it's also because we're, you know, we're around 550, 600 people globally. Like obviously we have some scale. Right. So we, we do also want to continue every year still get a batch of new hires. And I do think what's amazing with college students, I see my sons just finished with sophomore year at the University of Colorado. They actually are much more AI first And I think also an interesting fact, I read about ChatGPT, their usage actually drops in June and you're like, what do you mean? Like they're on such a long term growth trend but that's when college students are out of semester and because college students use it so intensely.

Speaker A: Yeah.

Speaker B: And frankly also somewhat nefariously and somewhat upsets my boy. I think they're good students but they're like, wow, all my classmates are cheating.

Speaker A: Yeah.

Speaker B: You know and even now I think post chatgpt they try to do take home or like at home digital exams and then there's like a camera watching you and watching your laptop. But I think what they're doing now below the screen they have their mobile phone and they take a picture of the screen of like the test problem prompt chatgpt on their mobile below the camera to give them an answer. And so there's also all kinds of, you know, but anyway college students are figuring out how to use AI in every single way and they're growing up with it. So I also think there's some advantage to some junior talent that you know and I think they are AI first now anyone graduating from college because they added throughout college. Right. They, they started college when ChatGPT 3.5 existed. So it's just a different way of working. So I think yeah, so I think a mix, you know what I mean? Like I uh, think senior talent used to want very experience like I said probably the best to me to building product is like the mid level like let's say 10 years post college some experience ideally have built some AI first startup have gotten to some revenue as you pointed out and then I think still adding you know, kind of a fresh batch of college grads and maybe that's like a non answer. But I do see in the age of AI, there's still value, you know, kind of across those levels of seniority. And the one challenge though probably for us and most companies, right, the reality is we're going to hire fewer junior engineers. Yes, right. We're going to hire fewer junior BDRs. And to me that is a valid concern for our industry. Right? Like what's the future pipeline? And I think most great CROs I know, even like Brian Millam who just retired as a president of Salesforce, I've known him for a long time, but he was like the first BDR at Salesforce, you know, like 26 years ago, you know, and like how are we going to create that next generation of leaders, whether it's in sales? Yeah, uh, you know, like T shape,

Speaker A: you know, to your point. And your son's like, like. So I think like, um, yeah, you can cheat, that's fine. But if you're a generalist, if you look a well rounded person, so it's not about just the uh, kind of the T shape, right. So you can be good at all of the things and you can go dive deeper into one specific section. So let's say yes, you are cheating on something, but as a generalist and this problem solver, you can be better. So I think like the future of the BDRs are like generalist BDRs GTMs. Someone that can both do the technology, the content, some design, do actual outreach, understand all the things and instead of just saying, oh, I am a, uh, junior sales exact, I won't have to pick the phone and call and sell. Right? Like now it's like selling is like okay, yeah, it's one of the skills. Tell me 10, show me 10.

Speaker B: Right. So and frankly in business entrepreneurship, like that's not cheating, you know what I mean? Like business, like the opposite. We're like, wow, that's a clever use of an AI. You know what I mean? Yeah, of course. Yeah. And uh, so, but I think you're right. I think yeah, people can go more full stack more Google or maybe as a bdr, even show a mock up because they're like talked about the product. You're like, I don't quite like it. Well, how about, you know, for our next release we make it look like this, you know what I mean? Like so I think there's just so many exciting use cases and I think you're right, we all become more full stack because it used to be like it was hard enough to learn Just was specific to the sales domain. But now that you can have strategy, anything, you know, like, yeah, you should go more full stack. And I think those are like the young people that are going to succeed because they have more power. They're going to use AI full power and they're going to be more creative and it's uh, going to take them outside their domain much more quickly.

Speaker A: So if you're not full stack, why we're not doing right? So it's like I'm talking to my bears. Like, okay, did you did the clay? Did you do the personalization? Did you do this? Did you research? Did you outreach? Did you content? Can you do the social? Can you do design? Can you just create the banner? It's like a no brainer. Like you have all of the tools available for you.

Speaker B: Yeah, don't wait for marketing. Right. To create.

Speaker A: Exactly.

Speaker B: Like why didn't you want to just do it, you know?

Speaker A: Yeah, exactly. So let's get the juices. So that's why we're hopeful for the kind of the new generation. Uh, it's been a pleasure having you on the show. I appreciate you, sir.

Speaker B: Yeah, thank you, Michael. Fun conversation. Thanks for having me guys.

Speaker A: Hope you enjoyed the interview. If you haven't like comment or subscribe to Belkin's podcast, I would appreciate if you do that. It's just going to help me a lot to get even better, um, guests and interview them on your behalf.

Speaker B: Have.

Speaker A: Thank you so much and I'll see you all in the next, um, interview.