AI is Your New Salesperson: Deepak Gupta on Going From SEO to GEO

Your website has to be neat and clean. So good brand, good design, good messaging. Because whenever even LLM find you, customer will gonna look at what you have. Because cybersecurity trust is most important. So if you don't have a good brand, it's not gonna help. Welcome to Cult Products, a podcast by EIA Digital. I'm your host Phil Keeney Bolland and on this show we sit down with the founders and leaders shaping the next wave of cybersecurity innovation. You'll hear the stories behind how they found their first customers, defined what made and built products that gained a loyal following in one of the toughest markets on earth. Let's dive in. Hi everybody. Welcome back to another episode of the Cult Products podcast. Today I'm joined all the way from San Francisco by Deepak Gupta, the technical co founder of GRA AI. How are you doing, Deepak? I'm doing good. Thanks for having me. It's great to have you on. This is something that I am really, really interested in. It's something that a lot of our clients are asking us about all of the time and it's something that I feel like a lot of people still getting to grips with, still learning a lot about. I know it's still an evolving space. It is the complex world of AEO and GEO for cybersecurity businesses. For those that don't know, Graka is a GEO specialist company focused squarely on cybersecurity businesses. But I'm probably not the best person to tell everyone about it. Deepak, tell us a little bit about what you're up to at Graka. So first of all, you're right. This AO GEO space is evolving really fast. As you have seen lot of big companies moving into this space. Adobe did the acquisition of Samras. They just completed the acquisition today. HubSpot recently launched their AEO platform not a week ago. So lot of movement happening in this space and we were quite early. So we started Gracker about a year and a half ago. So my background is cybersecurity. So when I look at cybersecurity companies and they have a lot of challenges with the research and how companies can get found on Google. And now a couple of years ago, like when the AI boom happened, like a lot of businesses, even we as individuals, like we started going to ChatGPT and say, hey, what is the nearby restaurant? So Everybody started using ChatGPT Gemini for search. Like we even need to find a dentist. Hey, what is the good dentist near me? So it's crazy to me and that's where by like about two years ago when I looked at the space I'm like all the businesses start asking the same thing, hey, I'm looking for this specific IAM solution, I'm looking for this specific email security solution, cybersecurity solution, what should I do? And they provide all the context. I'm a company size, I have 100 people company. I'm looking for these challenges. So that's where like we started thinking about and building this solution and we built Gracker specifically for cybersecurity company get found on LLM search. So we have enormous amount of data. So data is one of the key thing we have in cracker for cybersecurity company because we have created our own proprietary data. We have mined every single cyber security resources and that's where how we optimize the prompts and search results for companies so they can found to their perfect buyer so they can speed up their acceleration revenue. I'd love to hear a bit more about that data. I think first this is relatively unusual to have come from software engineering and cybersecurity and then to move into more of a marketing space. That in itself is fascinating. Is that something you've always been interested in? What sort of motivated you to get into that? My always drive factor is solving a pain point, solving a problem. And that's where I started my first company. So my background is software engineer, I did my master's in cybersecurity and forensics. So deep cybersecurity background I have. When I founded my first company I wanted to fix the login and authentication problem for consumers. Like how can they quickly log in to the business applications, websites and then how they can secure their data. So that's where I found LoginRadius. In 2014 I built the whole platform. One of the big challenge I saw you have an amazing product, you have amazing technology. If you don't have a distribution, if you don't have customers, if nobody's using it, it's nothing, right? So that's where I solved the problem of distribution. So I did the whole developer growth and product led growth. I talked to a lot of customers, I closed a lot of deals and that's where I thought when I started thinking about my next venture I'm like hey, why don't I solve the go to market problem for companies? Because I had the same problem 10 years ago. So that's where the whole mindset and then a lot of people, a lot of companies I advise in this cyberspace or B2B space. They are struggling with the same challenge. We are looking for more customers, more revenue, how can we get found to the perfect buyer. And that's where I started thinking about tracker and then when I did the scan and how the SEO is moving to jio, I started thinking and then assembled a team and we started building the solution. I'm curious, are there specific challenges or opportunities for cybersecurity businesses in AEO and geo? There are a lot of GEO solutions nowadays in the market, but most of the solutions are horizontal. Like they focus on B2B, B2C, any sort of companies. But there are specific challenges in cybersecurity because CISO CIOs like they ask for specific use cases when they look for a solution. They don't say that hey, give me a cnapp solution or give me like an email security solution. They specifically ask about the specific use case. And that's where we are expert. Like we tell the customer that hey, your icp, you are specialized in this particular category, this particular solution. So we should optimize your prompts and LLM search for this specific use case or multiple use cases, specific region, specific monitoring group. So cyber security is an area where you need to convince the technical buyer to buy a technical product. So you have to match that. So that's where we have expertise. That's where I was talking about earlier about the data. We have mined the data. It took us few months to mine the data and we have to keep crawling the data every single day. Like we have to keep enhancing our data. Like and in a week or so we are actually open like releasing the data to public. We are calling like a quarterly cybersecurity index where every week we're going to be releasing like what are the top cybersecurity solution in specific category. People can crawl through the data. Like they can pretty much navigate across the data and then they can see what kind of data we have built and how it would be useful for companies to say their ranking, what are they missing? How can they use some of the opportunities to get into the searcher in front of their customers. Tell me a little bit more about the data that you've. Yeah, so the data we have, any company, they can put their use case or URL that, hey, this is what we do and then our data matches with their competition, similar companies from that particular customer or that URL and then it matches all the use cases, what that customer has versus other companies have and then it provides the benchmark. So if you are like an email security company, where do you rank, what kind of use cases you are missing or what kind of use cases you should have? Who is your icp? What should be your positioning? Where are you lacking on your positioning on your website, on your third party resources, on G2 or some lot of review sites and things like. So it pretty much scans everything and give them like a summarized way. And that's what we are hoping to launch in a week or two where companies can get a lot of value out of this data and then they can navigate across their category, their competition's category, they can analyze what are the gaps in their messaging positioning and a lot of things that they might be missing. Right now I'm trying to secure these new agents that we're developing, what are the best way to do that? And so they're in more of a kind of build mode. And as part of that interaction and research and actual delivery within AI tools, that's where you want to be front of mind, you want to be going, well, the best way to secure these agents is like this. And you can use this tool to give yourself a head start and avoid having to build everything yourself. And what you're saying is you have the data that the cyber security companies can use to say, well if we're going to connect with those people in that way, these are the things that are relevant for us to be talking about given who we're targeting and what we do as a business. This is where we stand against our competitors and here's the opportunities. So we have created the whole metrics of like, what are the pain points, what are the intents, what are the various different types of intents? So yes, we can connect the dots on the use cases, ICP and then we can bring that particular. So if you are looking for a tool, when you are developing for your example, right, if you are developing an agent, like a company developing an agent, and then they are looking for agent security or agent identity solution, agent registry solution, then we have a lot of crawling in the back end, we have all the data and if like an agent security company comes in, we can tell them that, hey, you should optimize for these particular prompts, these particular use cases, because that's what people are searching, that's what people are looking for and that's where you can be found and then customers can quickly reach out to you and you should be able to make good revenue out of it. I'm interested a lot of the time when we speak to early stage businesses, we will often say it's A good strategy to start off in a niche and establish a monopoly in that space and then build and build from a position of strength. But a lot of people find that quite scary. You've gone in for cybersecurity right from the off. Tell me a bit about your experience of starting with a well defined niche like that. Well, I made the mistake before, so this is my second venture. So the previous company I started we did not had a niche, we just went horizontal on authentication, single sign on. So that was definitely mixed. A lot of learning from my previous mistakes. But yeah, when I started Cracker so the dragger is like growth hacker. So the since the day one we were like hey we will gonna work with the attack cybersecurity companies to optimize their growth. So that's why like the name is also connected to cybersecurity. Like it's a growth hacker. How can we hack your growth? So that's the whole intent and the key thing about niche is you are absolutely right. Like any early stage company, if they don't have a niche, if they don't establish that we are expert in this particular segment, it's going to be very hard to scale. Like we have lot of companies coming from different verticals. Martech, Fintech, like a lot of different industries, lot of different types. And we tell them that hey, we are focused on cyber security. Yeah, we have a solution. You can definitely sign up and try out the solution if it works. But our key focus is cyber security and that's what we want to focus on because of the data and intent we have. Yeah, maybe down the line in future once we have more data and intent for other industries, we might start spreading it. But yeah, the main thing is about focus is that our product development makes it easier. Talking to customers makes it easier showing them the data that hey, this is the data we have, this is what we found you with our data. So customers are also convinced that hey, they know what they are talking about. All of our team are cybersecurity people, even me and my co founder and like a lot of team members they have cybersecurity background. So that also makes us a very good solution to build on the space we know. So and cybersecurity is a very growing and with AI it's much needed space. I sometimes wonder if we just really overcomplicating things that do we need to split these off into different categories? Does this just make it something that people are becoming scared of? Because now instead of thinking about one thing they have three things to think about or is it all really the same thing at the end of the day? So it's all basically discoverability and getting people onto your website and into your sales funnel. How do, how do you think about that? The core is always SEO. The even Google's algorithm is going to work same way as they are before. Like they will going to figure it out what the business does, what kind of messaging they have, what problem they are solving and then they will going to find that business. And then if a user is trying to search similar to that business, they might want to connect the dots and say hey, you might be looking for this similar solution. So that's essentially the SEO is like if you are offering a product or solution or service, you want to get connected with someone who's finding similar solution. Now you need to figure it out on your website as well as around your website, all the forums, all the review platforms, a lot of different things. You have the consistent messaging so search engine can understand your product and then they recommend you. Now the complex thing comes in with LLMs is first of all There are multiple LLMs. There was one Google, I mean Bing was there but definitely it wasn't heavily used. But the challenge came with AI is there are multiple LLMs, there is ChatGPT, there is perplexity, there is copilot, there is Claude AI overview and now there are tons of other like LLMs specialized in Leapseek in Europe Mistral like different, different LLMs in different regions. So they have kind of made their own space. So the big challenge or big what we have seen when we do the AI search is each LLM works differently. So some LLM focused on your website, what messaging and positioning you have. Some LLMs focused on third party sources, some LLM focus on purely on YouTube and some video content. So now as a business if you want to get found on ChatGPT or perplexity depends on where your users are hanging out. You need to optimize for specific sources. And before that let me clarify SEO Geo aeo. So technically it's a whole umbrella. So SEO is definitely long standing how you can optimize your website and things. AEO is very close to SEO. So what answering and optimization is what you have on your website, on your brand, how you message and position is better. Geo is your website plus all the other third party resources. So whether it's YouTube, whether it's LinkedIn, whether Reddit, whether Quora, like various G2 different sources. So the collective thing including your brand and third party is called Gu. So SEO is mostly about search engines finding it AU and Gu are two category came in where first you optimize your website like messaging, positioning and then the second thing you optimize your third party resources who are referring you around your brand. So that's where the geo came in. And the previous thing I was talking about different LLMs citing different sources. So that's where you need to figure it out. So let's say for an example you are an enterprise security company and your potential customer might be using Copilot because Microsoft has a good penetration in a lot of regulated industry. So you need to optimize for Copilot. And Copilot might be looking for different indexes, different sources, different citations. So you need to figure it out as a company how you can get found with your users in Copilot. But essentially you need to understand that optimizing your brand, optimizing, making sure that you have a good design, good messaging on your brand and then all the resource or all the other sites are talking about your brand in the same way what you are messaging. And that is the best way for LLM to understand because they need education, they need to understand your brand and that's how they going to recommend it to your potential customers. If you don't have a good design, if you don't have a good messaging, if you don't work with a good design team or good design partner, it's going to be very, very difficult to get into LLM search. I think that is one of the core for the AI visibility. You have to have a good website with good positioning, good designing that conveys the message with good research. What's changed between traditional SEO and how companies needed to deal with that and now the AEO geo space. One of the big thing that has changed and that is still shifting is the context. So even in your messaging you can say that hey, we provide services to this kind of products we have, this is the solution we have. If you don't have good case studies, if you don't have good white papers, if you don't have a good use case that hey, we cater to this particular context or this particular type of companies, these are all good fit icps that actually matters a lot with LLM earlier with Google, if you search, hey email security solution, it's going to give you that these are the 20 links, you can go through it, crawl it, things like that, right? But now when you look at LLM search, if you go to ChatGPT, you're not going to Say email security solution, you're going to say, hey, I'm a hundred people company, we have 10 IT people. We have this kind of challenge with the email security. We are, our emails are going into spam and things. So you describe the problem and then say, okay, can you provide like top five solutions that would be good fit for us. LLM will not going to give you links and things like that. It's going to give you, hey, these are the five solutions you should look for and then you can drill down on that, right? So the context is one of the things that has shifted and is shifting and changing. So if on your website, on your brand, if you don't have that context, if you don't tell the LLM that hey, if you tell the LLM that hey, we are email security solution for everybody, it's not going to solve anything. It's a similar way like Grecker, right? We always tell on our website we are a cyber security geo solution. So a lot of cybersecurity companies come to us and they say, hey, we are looking everywhere. Each and every LLM you are the number one because are looking for the cybersecurity geo solution and there is nobody out there in the space. So same thing context. So if you Talk about previous 5 years ago with SEO on Google, if you search hey, geo solution it's going to give you 100 links. Like it's not going to give you these are the three solutions or these are the top solutions in the market. So the context is the one thing that has changed and people have to understand that they cannot serve the whole industry. They have to find their niche and that's how they can be found very fast and clearly to boost their LLM rankings. It's interesting because I think over the years with SEO there's been a bit of tension between design and user experience and SEO because some of the time some of the things that would get you higher up in search engine rankings were not necessarily the best experience for a user on the site. I think what we're talking about now as far as the AI side of things is it's quite different because actually it is very useful to a website visitor to be able to have more clarity on who is this product for, isn't it for me, more clarity on what the use cases are and how it could be used. My take at least it feels more natural than SEO did some of the time when it, when SEO was like, well this is how game the system to an extent because this is how Google ranks things. So now we're going to build in a bunch of this stuff into the site to get higher search ranking. But then people would land on the site and go, wow, that's a huge table of links. That's great. I don't really see what value that provides. To me I think that can only really be a positive for the end user because you are getting the clear answer. You are not getting like 20, 30 links and you don't want to waste couple of hours doing research. LLM has already done all the research education for them for you and then you just need to make sure that you found the right solution. So that's what I tell all the companies, that if you define your niche, if you define your use case, if you define your ideal customer in a clear way, I think SEO is definitely in SEO also you have to do that. But, but in LLM you have, it is a must. Like you cannot get away with that. You cannot just say that I'm serving every single company, every single industry, every single countries. You have to have a specific niche in mind and specific even location. You have to figure it out whether you are serving us, Canada, UK or Europe or Asia. Like you have to have that particular niche and you have to optimize for that niche. That's how LLM will gonna find you. And I think this is good ammunition for the bigger picture, which is, you know, it really does make sense to have that clarity. What is your target market? What is your value proposition? Get that very, very clearly defined at a strategic level and then roll out consistently across your website, across social media, all those kind of things. Because now consistency more than ever is king. If you don't have consistency, it's not going to have a clear context and you're going to miss out on the clear opportunity that you would have. So definitely even now consistency is more important than before. Even with search engines, less noise, I think they have to create less noise and be more specific and react. I guess that's the key in the new AI world because AI is not going to learn all the things that we are talking about. AI is going to be very direct. AI is not going to beat around the bush and directly this is what I want. Here's a couple of people who do it. This is the one that I recommend that you bit that you do, which is similar in a way to if I was doing DIY and somebody said if you're going to do that you're, you're going to need to drill a hole in the wall. You're going to need a drill. This is the drill that I would use. You go, okay, yes. If I want that shelf to go up, I need to go and buy a drill. Yeah, you had to do hard way like now I can tell you directly you have to drill the hole. But before you have to watch a YouTube video, you have to go read a lot of articles and then okay, now I need to buy a drill. So it's gonna save you a few hours of time I guess with the whole AI and LLM. So it's gonna save you a lot of time. You are not an engineer, but now anybody can build the solution. If you have an end goal in mind, you can tell these cursor Claude, hey, I'm looking for solution like this, can you help me? And then it's gonna navigate and build a solution for you also. And that's why I think being very clear on that value is important because people aren't going to be just searching for the tool, they're going to be searching for the end result. That's the use case like they are trying to solve this particular thing. And what should be the bridge between these things? That's the main thing. AI is going to help you. If you were sat in front of a cyber security business owner or a marketer and they said to you right, if I don't do anything else in this space, what are the, what are the top three things that you would recommend that I can go away and do right now? Biggest bang for your buck, biggest impact, absolutely essential stuff. The first thing I think boiled down to is like the door of your product. So your website has to be neat and clean. So good brand, good design, good messaging because whenever even LLM find you, customer will gonna look at what you have. So if you don't have a good UX UI design messaging, they're not gonna because cyber security trust is most important. So if you don't have a good brand, it's not gonna help. So the first thing I would say is focus on your website that the front door has to be very nice and clean, organized with the good value prop, good messaging and things like that. So that's one thing. The second thing is you should have use cases, case studies, things like that that you have done before with other customers or other type of companies that always help to get into search as well as when even customer comes. The third thing is around your website, you should be present in a lot of these review forums and different, different types of sites like Reddit, LinkedIn and things like that social media. So it can have a little bit value prop that hey, you are quite active on these communities and those communities will gonna if someone is looking for a solution like yours, it can found you. So your website definitely having a blog or case studies and white paper is important. And the third thing is participating in those communities with the specific niche on a specific solution that you have. So those are the three things very important with the community stuff you see that with brand accounts, with individuals who are strongly linked to the companies, how does that work? So it depends on the company. But yes, if you are affiliated with the brand account you have to mention that hey, this is what I'm saying and this is my brand and I'm biased towards that brand. So definitely you can do that or you can have a personal account where you can also promote yourself, not like explicitly, but solving the pain point again going back to the same thing, right, so if a user is looking for some specific problem, you can solve the problem, you can help them with your solution and then you can recommend they say hey, you can look at this our product if you like, but this is the solution of your problem. So it's communities. I think participating in communities are more around education, like how you are spreading your education and knowledge, your solution, the problem you are solving to your potential customer. And then if customer wants that, hey let's go and look at this company, maybe we can look at this solution, then they can come to you. But you should not self promote in communities. That's what I recommend. You should not self, you should focus on education, you should focus on the value first. So communities are more around value and that's why all the LLMs, they value Reddit, LinkedIn, Quora and things like that more because they have quite a bit education and value rather than promoting the businesses. Interesting. Do you see? I mean you mentioned LinkedIn and Reddit. A lot of the time we're getting asked about which are the key platforms, which are the best channels to be on. Is there a hierarchy in your mind? Is it best just to pick one and go right, I'm really going to focus on this one and do really well on it. Or spread yourself across all of them. I think you should spread in a couple of them because you should not just rely on one because you don't know whether your potential customer might not be looking at that particular community. So you might at least look for. And that's what our solution does. Also like it look for that particular use case and customer Persona and then tells you that hey, you need to focus on these communities, these five communities or these five sources. Is there a benefit to the more sources? The more different individual sources that cite you, the more sources you have your AI visibility definitely will going to go up. LinkedIn is more around like B2B community. So if your solution is heavily on B2B 100% LinkedIn should be there. Reddit could be both B2B and B2C because it's a wider community. Quora Hackers News There are like tons of communities out there. If you are targeting indie hacker, there is indie hacker community. So there are specific communities also for a specific niche. But some of the big ones, I think the brand should be present in those big communities. I'd be interested within the cybersecurity space. How much can you quantify the impact of good NGO strategy? What is it that people could really achieve by getting this proper attention? There are a couple of indexes, share of voice. So share of voices, like if you are like 10 competition in the same space, like how much voice you share in all the communities as well as on your brand. Right. So we have an index where we tell you that you are 10%, you are at 15%, you are at 6%. So it depends on how much share of voice you have across your competition. There is AI visibility score where you have 100% score at the top, but where are you? You are like 30%, 20% and how your AI visibility score and share of voice is going up and up. So if you talk about specific how these companies can quantify, the main thing is whether their AI visibility score is going up and up. So are they getting more and more indexed based on the use cases and problems they are solving? So that's a good indicator if they are getting more and more customers. Like if like a lot of our customers, they always come to us and say hey Deepak, we got customers. Ask us on LLM and then they reached out to us. So we got really good customers. So that's really good to hear that at least the solution is valuable, right? Like if you are optimizing for AI search and then potential customers are finding them. Win win. For customers as well as the company who is looking for the solution, there are some indexes but ultimately boils down to how many potential customers you are getting through AI and how many citations and how many searches you are getting into with these LLMs. I think it's interesting because it's such a strong intent signal. This is different to just doing general research about tools because it's linked to a thing that that person is trying to do, really wants to do, and it comes up in that environment and says, this is a good way to achieve that objective. I don't know if it's really as relevant to just look at the kind of raw numbers in the same way that it might have been more so with, you know, SEO or even Patreon. Because what you're getting is somebody who's been kind of already pitched to by an LLM out your product. They're arriving with a lot more context and they already have ident or have had it identified by a trusted referral, essentially that it can do a good job and it can do the job that we want it to do. So maybe it's not a case of using the same old metrics, but actually thinking about this as it's actually much lower down the funnel. You are right. I never thought about it that way. Like, AI is your salesperson. They are pitching this. I never thought about it that way. That's a really good point that you made. Right? AI is your salesperson. AI is your affiliate manager. They are like linking you to the customer. So that's a good analogy that, yeah, AI is becoming your salesperson and selling for you. Yeah, you just need to train them properly. You just need to train the AI that, hey, this is what I'm selling, this is what I do, this is the problem I'm solving and AI is going to go and sell it for you. This has been super interesting. I feel like I've really learned a lot in the time that we've been chatting. I'm very, very grateful to you for coming on the show today. I think it's always a nice way to end. And I always like to ask everybody this kind of question, especially given we're living in a world that's just changing all of the time and we're all just desperately trying to. Trying to keep up, really. At the end, I'm wondering if you now had the opportunity to go back to the founding of Gracher and give yourself one piece of advice. What would that piece of advice? Very slightly younger self. In last one and a half year, things have changed drastically and it's gonna keep changing as the LLM goes. Like earlier we thought that, hey, let's just optimize for ChatGPT. That should be good enough. Like, that's one of the core widespread audience. But as we get into, deeper into that later, we found out that a lot of enterprise buyers, they are not using ChatGPT they are using Copilot or they are using Google for there because they have the solution in their companies for Bing and Google. Right. Microsoft is quite heavily penetrated. So I think it should be like one of the advice or maybe thought process is like think a little bit wider, not just like what the mainstream industry and media is going after. So that's one of the things that now whenever we develop something or look for something, we always think about the wider scope of the problem, not just like one narrow way. So that's one of the biggest learning we had with these LLMs, especially these new LLMs coming in like deep seq Mr. And things like that. We thought that they are not going to be that big, but now they are specialized LLMs for specific regions, specific countries. So yeah, that's another good learning. On the same front that things are keep changing. So you should be open about adding all these things to your toolbox to add value to our customers. Certainly going to be interesting TV thank you so much for your time. Where can people find you if they're interested after this? My handle is Dipak. I am on LinkedIn. If you have any questions or if you want to know about cybersecurity, a GEO solution, you can reach me at deeppacketcracker AI. But yeah, I'm quite active on social media. I have my personal blog, Gupta Deepak.com, where I post a lot of my thoughts around cybersecurity, marketing, geo and a lot of research that I do day in, day out. Brilliant. Thank you everybody for listening. Thank you Deepak. And we will see you next time for another episode of Cult Products. Awesome. Thank you so much, Phil. Yeah, great talking to you. Cult Products is brought to you by Yaya. Helping cybersecurity companies define who they're for, what they do and how they're different. To learn more, visit Yaya Co. And don't forget to search for Cult Products in Apple podcasts, Spotify or wherever you listen. Follow the show so you never miss an episode. On behalf of the team at Yaya, thanks for listening.