Kevin Kerner: 0:00

Hello everyone. This is Kevin Kerner with Tech Marketing Rewired. Photo market isn't what it used to be. Channels are saturated, the buyers are guarded and AI agents are starting to do the work humans used to handle. In this episode, I sat down with Santos Sharon, founder of Zier AI, to talk about how AI agents aren't just automating outreach they're increasingly becoming the buyer's first move. It's a new model for discovery where agents collect, qualify and surface the best options long before a human ever enters the chat. This is revolutionary stuff from my good friend, santos. I hope you enjoy it. Let's get into it. This is Tech Marketing Rewired. Hey, welcome to Tech Marketing Rewired.Kevin Kerner: 0:47

I'm Kevin Kerner, your host, and today I'm joined by my good friend, santos Sharon. I watched firsthand how he's really had a front row seat and led, but I don't know all the stuff data intent, platforms, all the go-to-market stuff. It's really evolved over the years and, more importantly, I think Santos is at the forefront of how it's changing, especially with AI. He's held roles ranging from I looked at your LinkedIn, saw your Digitas, I forgot about that Lead Genius, rb2b, and we actually go way back to Aberdeen. We were at Aberdeen together and today he's the founder of Zier AI. So, santosh, I'm super happy to have you on the show. How you been.

Santosh Sharan: 1:26

Very good and so thrilled to be here. Look forward to chatting with you, Kevin. Yeah, this is good stuff.

Kevin Kerner: 1:32

Can you imagine if we had the capabilities that we have now, when we were both at Aberdeen? The whole intent?

Santosh Sharan: 1:38

Like 12, 15 years ago.

Kevin Kerner: 1:40

Yeah.

Santosh Sharan: 1:45

I just think we were a little ahead of our time back then. We were, we were, and you know, in some way, the biggest change that has happened since then to now is that earlier playbooks were only accessible to savvy marketers, and now everybody has a playbook with a button.

Kevin Kerner: 2:00

And that's direct marketing with a button and that's the direct marketing. Can you imagine all the rules-based stuff that we were doing? It's just crazy. And you didn't have any access to it. Now you can just go to Claude and just make it happen. It's crazy. Well, so I went a little bit over your history. I wonder if you could take us through a quick intro, your journey of how you got to where you are, and then just dig into a bit about what's happening from your perspective and go to market right now.

Santosh Sharan: 2:30

All right, let's dive in let's do it.

Santosh Sharan: 2:32

So my career started off. I started off my career as a software engineer that's what I studied in college and somewhere seven, eight years into my career I was part of a small startup that got acquired for a lot of money by Johnson Johnson. It was a pharmaceutical startup. So that's when I got this bug of starting companies and I did one, and then I really began to like go-to-market. I realized that more value can be created through go-to-market than through product development and that led me with ZoomInfo.

Santosh Sharan: 3:11

And then last 15, 20 years, increasingly, I've been spending more and more time with sales marketing teams, sometimes operations, but primarily focused on go-to-market. And here's something radical I have to share. I think the last time I felt like this was 15 years ago. So if you break go-to-market into big generational shifts, before 15 years ago I think you would hire a CMO for their gut feeling or CRO for their Rolodex right. People were more intuitive, more relationship-based go-to-market, right.

Santosh Sharan: 3:54

But then, like somewhere around 15 years ago, data and automation started to take over, where relationships didn't matter as much. Right Attribution mattered a lot than intuition. Right Now, it was all good in the beginning. Now, fast forward to today. I feel like everything's broken. Pipelines are down. The sales math doesn't work. All marketing channels suck. Nobody responds. There's almost dignity in sales and marketing that's got lost, because imagine being a sales rep and doing 20 calls and everybody, like screaming at the other end, don't want to talk to you. That's not how it used to be Right and we got to fix that right and we got to fix that.

Kevin Kerner: 4:48

Yeah, do you think it's um? I was just. I just told on the pre-call here. I was telling you about talking to elaine at um tofu. Do you think it's generational to some degree? Like like the part of the change is the technology that we now have has changed, but now the buyers now uh, b2b buyers are in a different generation to some degree. So it's like the tools are have gotten to the point where they've matured and more accessible, but it's almost like the generation now sort of is maybe not used to the same tactics that we used to, or doesn't like the same tactics that used to work back in the day.

Santosh Sharan: 5:16

That's a good point, Although I feel like too much of good thing becomes a bad thing. So we got greedy with data and automation, so greedy that we said you know, like it doesn't cost us anything, let's just send 100,000 email when we could send thousand. Right, Because it doesn't cost much. Now think about a CMO of a company that just raised $50 million. So when we were at Appendine 15 years back, this CMO would have gotten, say, two or three emails saying, hey, congratulations, and I think my product could help you scale and let's talk, right. But today the CMO is getting 100 emails because it's become so easy and accessible that with one click, like, every signal provider is providing funding signals right, and every automation is integrated with funding signal outreach, and so that's such a lousy playbook now Like so then, what the buyers are reacting with?

Santosh Sharan: 6:22

This is what Gartner has been saying for like 10 years. Buyers don't want to talk to salespeople increasingly right, and they are responding by being disengaged. They're just stepping away and doing their own research and they want to come in when they have made their final decision to work with, say, two of the 20 vendors right. Of the 20 vendors right. I think part of it is generational, but a lot of this is just buyers responding to democratized or accessible or low-cost data and automation, and a lot of lazy automation, I should say the smart automation still works.

Kevin Kerner: 7:02

Yeah. So the thesis is it's a lot of noise led the collapse of channels because you've got so much spam or poor marketing going on, and it's only going to get worse. It's just going to get bigger and bigger. What do you think the opportunity then is how do you react to that as a marketer? Now, what's?

Santosh Sharan: 7:28

the opportunity. I think everybody has to think about new playbook. So when we were talking about ABM and signal-based marketing, really we were doing noise-based marketing.

Kevin Kerner: 7:40

And even ABM. Now, Sanjay, if you think about it, it's still pretty noisy because everyone has the ABM technologies.

Santosh Sharan: 7:47

There's another vector that is fueling all this, just sitting in the back of the room, and that's AI, a sector that only had, say, 50 vendors. There's going to be 100 in like next 12 months, or 150, because the cost of building tools are going to just go like near zero, right? So what that does is the buyer cost increases because buyer has to search through like a few tools they could count on fingers. Now they can't. They're like 50 more and then everybody begins to look the same. There's like no differentiation between tools. How do you differentiate through the noise?

Santosh Sharan: 8:31

When you're 20, then it's hard to differentiate. Imagine if they're 200. Yeah, right. So this is where a new breed, a new age and new generation of marketers have to emerge, and I feel like a lot of consumer marketing will start to, because consumer marketers are already used to dealing with this noise. So brand, for instance, is more important. Cutting through narrative, radical differentiations, connecting with human emotions all of that becomes more important. In fact, product differentiation will start to take more of a backseat than before, because cost of product building will be same Even if you're different. So what? Somebody can copy you in two weeks. So then there will always be parity.

Kevin Kerner: 9:20

Yeah, that's a super interesting concept. And so, because you're right, I mean I can go into base 44 and replete right now and I'm doing this right now and it's a lot of fun and I'm building apps that have pretty good features. I'm not launching anything, I'm just doing it just to test it. But, like, what is the cost of a feature? Now? It's like it's, it's so easy. I'm not building anything that's production grade, but you know, fast forward, another two or three years, I might be able to build something that's really pretty darn good, and so I think you may be right. I think the big players obviously are seeing this, like Salesforce and others. They're seeing these micro apps, or what does Scott Brinker call it. It's like it sort of democratizes the creation of, you know, app creation. Let's say I think you're right. Um that.

Santosh Sharan: 10:13

But trust and credibility cannot be copied that easily. Yeah right, your sense that this company will stand by me when difficult times come. That cannot be copied right. So I've and in one-on-one conversation like this you can build trust. I think we will start to see more scalable trust building and that will become the primary differentiator for good market.

Kevin Kerner: 10:41

Completely yeah. So brand is the trust part of brand is super important and authenticity and having genuine relationships. So tell me what you tell me what you see the business opportunity for a platform is now. I know you're doing some of some interesting things at Xere. Tell me a little bit about what the vision is for a more authentic go to market.

Santosh Sharan: 11:05

Yeah. So I was reading somewhere recently that in the last three years surveyed winning by design in Europe surveyed 50 SaaS companies public SaaS companies largest and they found that the CAC rose by about 60% in the last three years. Now let's extrapolate that and I'll answer your question more directly. So with AI and more competitive new entrants coming in the market, cac is going to rise. Right, and what's happening on SEO we'll talk more about that is going to make CAC rise because the inventories are dropping right, the ad inventories.

Santosh Sharan: 11:52

So endpoint solutions. So if you go 15 years ago, best of breed was the way to build a SaaS company. Right, buyers were buying the best of this and best of that and best of that and put it all together and integrate using Zapier and other tools. And put it all together and integrate using Zapier and other tools. Now, if the CAC becomes unsustainable for endpoint solution, where the CAC is so high that they can't afford given the cost of the software, right. So I feel like the high CAC will ensure that all softwares are just platforms. So once your customer is in, you can support all different workflows, not just one endpoint right.

Santosh Sharan: 12:38

So I foresee every software becoming a platform. Either they build a platform of their own or they join a platform. I also think, like someone like Salesforce can build end-to-end platforms, but small companies cannot. But I do think one like Salesforce can build end-to-end platforms, but small companies cannot. But I do think one of the biggest opportunities that's going to come in the next 12, 18 months is there'll be rise of multiple industry stacks where these companies are going to cooperate and compete at the same time and they'll build industry stacks out of box so they'll be priced together. They'll be pre-integrated so you can go just hook it up as if you are working with SAP or Rippling or some of those platforms that are compound, but with different companies. Without that, they won't be able to sustain the high cost of CAC.

Kevin Kerner: 13:35

The platform, ecosystem becomes the channel. That's a great way to put it. Yeah, do you see any industries where that, or can you give me some examples of where you think that might play? First, you mentioned Rippling, which I think is a really interesting example of how you could plug into that ecosystem. Can you think of any others that are top of mind for you?

Santosh Sharan: 14:00

I think the general platforms, like, say, marketing or sales, those are more broad, yeah, also, they have more budgets broad Also, they have more budgets. I would think verticals or smaller platforms where they have less budgets could even be a good place to start to take over, because then the CAC really makes a difference. Also, platforms that are verticals or segments that have, say, premium players, those would be good ones for smaller players to take off, because if there's a premium player, they're charging more but they also have more features with that others cannot replicate, but together five companies can replicate what, say, sap is doing for manufacturing or some vertical. So now, for the first time, I think, there'll be real competition against these premium players.

Kevin Kerner: 15:00

Yeah, and the whole key to all this is the composability between the different platforms, the ease of integration with other systems, and you know what will really activate that is that I have a lot of fun playing around with MCP now. It's so fun too. It's not the easiest thing because if you're, if you're connecting it with something that isn't isn't, doesn't have a really easy integration with, like cloud or something, but it's not that hard. But once that clicks in your brain that you can use MCP to have an LLM dig into something else Right, and I just posted the other day about this GitHub that had 3,000 MCP connections to it. But psychologically, once that clicks into the general public or the general business person, I think what you're saying is right. It could, this connection to everything could really begin to take off.

Santosh Sharan: 15:55

Yeah, and that is going to disrupt. Because if either you are part of a platform or you're out, because then it becomes now there'll be limited slots in these platforms, right. And if buyers are saying, look, it's cheaper for me to just work with a single platform because the car gets subsidized for everybody else, right, Otherwise, the endpoint solution they are to keep spending money into acquiring customers. So, yeah, you're either in or you're out of business. It's going to be amazing.

Kevin Kerner: 16:33

Yeah, that's incredible. So what's the next step from this thesis, Like where does it go from here? As a product owner or a marketer?

Santosh Sharan: 16:45

I think every endpoint SaaS company need to think. Like every SaaS marketers need to consider can they turn their product into a platform or can they go after adjacent workflows easily? Now cost of development has gone down, so maybe they can right. That way they can charge more or give more value to their customer for the same price. That way they can charge more or give more value to their customer for the same price.

Santosh Sharan: 17:12

The real battleground is going to be once you have a customer, how do we retain that customer? Because the new acquisition will just either be too expensive or start to dry up, because everybody will be holding on to the customers they have, right, yeah, and if you're a large company and you enjoy brand premiums and that, I think premium will kind of there'll be downward pricing pressure. So I think there will be issues with enjoying those premiums invested, hoping that there'll always be this winner-take-all model where excess capital will turn some of these companies into enjoying these brand premiums and charging far more than their competitors can. I think with AI and lower cost of development, those premiums are going to start to give way to smaller companies or start to get distributed.

Kevin Kerner: 18:11

Yeah, I completely agree, and I know you posted a lot on different pricing models and the sort of more credit-based or results-based pricing models. I think it's going to put incredible pressure on the HubSpots and the sales forces of the world, the big enterprise platforms. It's just even as a buyer, as a business owner, it's really hard to swallow some of those bigger fees. Now, now that I see the other pricing models, if I can pay for usage or credits, it's so much easier than paying per seat. It just doesn't work for me as well as it has.

Santosh Sharan: 18:46

Right, right yeah. In some ways, those are market inefficiencies, right yeah, Marketing is the science of exploiting market inefficiencies to your favor. And now we are getting more efficient and the companies that are benefiting from it are going to give back.

Kevin Kerner: 19:04

Yeah, okay, I want to go back to the channel thing that you brought up earlier. Search is difficult because AIOs paid is tricky, although I have a perspective that I think paid search is maybe a bit more important now because you can't get in the top, you can't get past the AIOs, so pay, you know, spending some on paid search actually is is not a bad idea if you're trying to get your brand noticed. But ads are tough. Email is almost impossible.

Santosh Sharan: 19:37

Right Phone is impossible. Even email is getting ranked.

Kevin Kerner: 19:42

I have an agent now that goes through my email. I don't know why Gmail doesn't do this, but it basically goes through my email and pulls it. It's the Relay app Pulls out what it thinks are cold emails, identifies them, builds a nice reply to them, puts them in my draft folder and I just scan them every day and I just hit send. I don't even have to look at them anymore. It's pretty incredible.

Santosh Sharan: 20:06

What you just did with email. What you just described, I think, is the core or the crux of the future of go-to-market. It's really important to understand what you just did, because you were looking at emails yourself and now the agent looks at it, first decides for whatever reason, and then you look at the remaining emails. Now apply that model, exact same model, to search. We were looking at search results. Now we're not going to be looking at search results. Some agent will look at it and give us only the top or top two. Apply that to sales calls.

Kevin Kerner: 20:46

Browsers.

Santosh Sharan: 20:49

Buyers are not going to talk to so many sales calls. They'll only talk to the top two right, and that's just going to be the norm. So those comprehensively like going through every candidate in the. Those days are gone and there'll be a lot of agent to agent communication. Now the race will shift to how do I impress the agent? So if I am sending email to your inbox I should be thinking what should I write? So your filters or your agent will escalate it to you Completely. I think that would be the new market 100%.

Kevin Kerner: 21:22

And same thing with your website, like do websites become really just agent tools, like the agent to agent? Yeah, so that's super interesting, did you see? I just saw this morning that Apple's new beta has a browser function. So if you're looking at your phone, you can hit I think it must be in Safari. So you hit the button in Safari and you can click on any part of the screen. That's a distraction and it will remove those containers from the browser. Wow, so if you're reading an article, like in wired or something or you know, you just maybe you want to read an article and there are a ton of different ads that show up you can just like you literally just hit this button and then you just start clicking off the ads and it makes them just like poof and they just, they just completely disappear.

Santosh Sharan: 22:08

Just went up a little bit. Somebody spent money to reach out to you.

Kevin Kerner: 22:14

So can you imagine, yeah, and I don't know, maybe the click action is actually a click, so you're spending the money as a brand, but your ad is disappearing. There's no reason that that can't happen with an agent instead of a person having to, like, put their finger on their phone. So it's just a matter of time before that stuff is tricky and there really is no system right now to scale authenticity, so authentic brands. It's tough because you know I do the podcast, because I'm super interested in it, but and I'd learned from all these great people but it's also you know we as a brand need to be out there talking about this stuff, because you know we're a tech marketing organization, so we need to be at the forefront of these discussions. But it's not easy, you know there's no, it's very manual.

Santosh Sharan: 23:02

It's the toughest problem of the next decade. I can tell you how we are trying to do that. Yeah, I can't say we have completely cracked the code, but yeah, that is. I'm sure there'll be multiple initiatives to do it right.

Kevin Kerner: 23:16

Yeah, Things that we don't even expect to see yet and then and then it'll get all spammed out. So what we'll be after on authenticity, who knows what that will be. But yeah, I would like for you to talk about what you see the opportunity to scale. It sounds terrible to scale authentic relationships because you don't really want to do that, because they're authentic and real. But how do you keep it real and do it in a way that allows you to at least scale, multiply yourself in some way, scale multiply yourself in some way.

Santosh Sharan: 23:50

Yeah, I think a big part of that is to just acknowledge what's going on and not fight it, ride the wave right and then the other is. So, for instance, this noise is just going to stay Right. And this sales-led marketing where sellers raise the hand and said look, my ICP has 10,000 people, I'm just going to go round robin on all 10,000 or a bulk of them and see who's in market to buy, that just generated a lot of revenue for every vendor in that platform, every channel. So everyone recommended that approach, but then it generated a lot of noise and a lot of mistrust and it basically wrecked those channels in the long run. Right, I think?

Santosh Sharan: 24:39

Think about the buyers. They just want to buy something and it's difficult to buy. They're not even paid or recognized for what they're doing, right? So, companies, you're talking about how to get you know how to build trust at scale. One way to build trust if I am a vendor um, even enterprise larger vendor um, and I recognize that buying is difficult in my space, I would create a website. I I would create buying tool. This is exactly what we are enabling at Xero. That will help my buyers buy better. So it reduces all the you know, hurdles and friction and that generates trust and credibility in ways that otherwise you can't. That's one way.

Santosh Sharan: 25:27

The second way if buyers can be encouraged to share their problems every day and the problems could range from like I have churn, I need to hire a controller, I need to do a campaign, I need playbooks and best practice for getting some lead gen. So buyers are constantly having these problems every day in business, in the course of business. If they could come to a site and like ask for those problems, and then agencies and vendors could respond just to those problems, right, not start pitching and saying, hey, I have this tool I think you should buy, and here are the five features my tool does, which is how they do today. Right, and if that can be done in a targeted manner, at scale, through agents, agents are actually going to get pretty sophisticated. Agents won't be like just chat gpt giving you generic answers. Agents will get. It will become an extension. It'll begin to.

Santosh Sharan: 26:33

If I have an agent, um, over time, if there's a mathematical certainty, then it starts to sound like me. Now, in the beginning it's clunky and it doesn't sound like me and it's barely reactive, uh, but uh, it's just like learning If it has all the content, it learns pretty quickly and it starts to. So we could respond in a very targeted manner to buyer pain and buyer needs On demand in real time within the context at scale. And now think about the buyer's mindset If they're getting answers and responses from the same brand or same agency again and again, they like to work with this agency, maybe prioritize and just go to the agency next time, right? And so this is how business is going to be done, where the agents create credibility in the beginning and these agents don't get annoyed.

Santosh Sharan: 27:30

If a seller agent asks the buyer agent 10 times a day, is there any demand for me today? But buyers get annoyed, like, stop chasing me, but agents are just API calls. So in our platform, agents are asking each other all like 10 times a day if needed, and that's fair game, right. And as soon as there's a demand, as soon as there's a question, as soon as there's a problem, as long as they're in the right ICP, these agents figure it out and then start helping each other.

Kevin Kerner: 27:58

Yeah, so agents can filter out the noise between agents and serve up the stuff that's valuable for the buyer right.

Santosh Sharan: 28:06

Yeah, so one more point there, since you mentioned the noise part. The biggest problem in marketing and sales today is we have created too many jobs and chores for ourselves, a lot of which are not useful right, since people are sitting on calls. If you look at their calendar, like a lot of the calls don't yield into revenue, but then there's no way to get around. You have to do like 40 calls a week to find those four that are valuable, but you don't know which four. Otherwise you could have done just four right. So when agents do those 40 calls between other agents, the sellers just have more time right and they can focus and they can do something more valuable with their time.

Santosh Sharan: 28:52

Plus, I don't think we'll need this large teams We'll probably undergo, you know, higher quota per rep. But also, I think there's going to be the way agencies work will change. I think a lot more budget will shift to agencies than before. But yeah, this is a massive time of change and transition. A lot of the noise that we created in the last two decades agents will cut through that noise.

Kevin Kerner: 29:24

You know, what's super interesting about what you're saying is if I'm going to ask a question and I want to go to a community, I'm probably going to go to G2, or I'm going to go to Reddit or somewhere. I'm going to look through Reddit throughout. I'm going to ask that question, I'm probably going to start in Google or something or maybe ChatGPT. So I'm going to the community in that way. But if I'm if I'm I don't do this proactively, but if I'm following a brand on LinkedIn and they have good content, I will interact with that content. So my brand affinity is happening on the LinkedIn side, let's say, because it's through content and stuff and so and it's a lot of it's happening through personalities with that content.

Kevin Kerner: 29:59

And then the the part of me that wants to buy something is, or ask a question about how to solve something, is going to the community sites. There's no bridge between those two things and if a brand shows up in the Reddit side, I'm going to not, I'm going to be like not, I'm probably not interested, and LinkedIn really isn't the place to ask. Ask, you can, but it's kind of an odd place to ask a question that doesn't have the form for that. So it's a super interesting concept of something that links that same I guess need together for a-.

Santosh Sharan: 30:32

Spot on. When you ask a question on Google, you're really engaging with content that somebody posted some of the time that was not designed to answer your question. You're trying to barely match to like a demand and a supply that were not made for each other, right? But agents change that. With agents, when you ask a question, like in our platform, it will go first do ai research, um, and then it will go to brands that it thinks should be able to answer this question. It will position it as as a lead to say a vendor or agency, and say, hey, do you think this is relevant?

Santosh Sharan: 31:15

Because there's a $100 million company, there's a VP, marketing of this company that is having this problem. Are you interested in writing a note? Right? So the good thing about agents is humans just have to respond to questions, like one time. This is like a long tail of Q&As that was missing in search. Search was built for the dominant use case, right? The same question that were asked the top 20% of the use case where search handled very well because it was doing keyword match, but there was this long, 80% of long tail that were not handled, which I think agents and with support of humans, because these will turn into real leads, could handle.

Kevin Kerner: 31:59

And AI mode in Google doesn't quite get there because it's using….

Santosh Sharan: 32:05

The same 20% content.

Kevin Kerner: 32:07

The old regurgitated content yeah.

Santosh Sharan: 32:10

If Google could add a feature that if somebody asked a question at that time, they could go to the content author that they thought was the best expert in those content and say, hey, there's a question that your content doesn't really answer, but I think you have the answer because I see what you wrote and can you please give me a new answer? Right, and if there was incentive for this author and they could give that answer, then Google could do this right. But then the use case of Google is instant real time. In our case, if it's lead gen and business problem, it doesn't have to be real time. It could be like one or two days or a week later. If you get some churn playbook from some VP marketing of Zendesk or somebody credible like you would probably, it's still as valuable right?

Kevin Kerner: 33:01

Oh yeah, oh yeah, for sure. And that's after it's vetted, after the agents have talked to each other and got out what they needed. Yeah, it's like you have a character AI sort of scenario where you can talk to the. You know I can talk to, uh, you know one of my kids characters that they like, but then, uh, then that's vetted. You know that is being vetted between the agents, but then I can actually talk to the character. Like, I can actually reach out and get a and there's something in it for both parties. I get the answer but the character gets the. You know they get incented somehow. The opportunity, let's say, is the incentive Super. Is that mean there's a massive amount of training that needs to go on on the brand side or the agency side to be able to have these conversations in a platform with these agents? Like, how do you get the data to make sure the conversations are?

Santosh Sharan: 33:55

relevant. So in the beginning we thought there would be massive training required, but now we are realizing it's actually much simpler. So we could start with whatever data is out there and train based on public data. But then if a brand signs up on a platform and says, hey, this is what I do, Say it's an agency or a vendor and they have a software that solves churn and retention so they set it up. Anytime there's a conversation or question around those topics, we put it in the demand tab for there and said, look, you told us you are interested in these topics. We think this is early stage, but these are the questions that came up in the last 24 hours that agents have already answered and we share with them what conversation stage we are in. By the way, this is going to be important for Google as well. I think the value will shift from just keyword match to what stage of the conversation are you in? Are you just like very high level, or this buyer is serious buyer and they're like going back and forth. We think they have to make a purchase or the problem is serious and we can tell, based on the queries and repetitive, how long they've been working on the same problem and so on. So we share and then the vendor chooses to. So to your question.

Santosh Sharan: 35:15

Here is how the training happens. The agents will respond to these questions and the response will say it's AI generated. But either the buyer kicks back or challenges the response or the agent themselves know this was not a high score or high confidence response. So then the response gets escalated to a human on the brand side it could be a seller and then when the human responds the agent learns for the future questions, right?

Santosh Sharan: 35:44

So we find that this one-time response in the first four weeks, six weeks, this seller would be seeing, say, 10 questions a day, but two weeks later there'll only be six questions a day. Four weeks later there'll be four or six. Eight weeks there's like one question a day or no question. So learning happens where upfront there's a lot of missing sort of concepts, and then it's better to train based on what people are asking, because the questions itself changes over time, rather than pre-trained somehow and expect it'll all work perfectly. I believe the human curation aspect is a key part of agents functioning, which I guess Google Answers or others are not really thinking about it this way.

Kevin Kerner: 36:36

No, because now you have a two-way conversation that can train the agent on the next conversation. That's similar. It's super interesting, it's very interesting. Where are you at with just your product? What's the launch been like? What's the status? I haven't got an update from you in a while on things.

Santosh Sharan: 36:59

What we are building is when people sign up to our platform, everybody gets an agent of their own and then they can come to our platform and write about whatever challenges they have. If you're a sales rep, you say I want leads. If you're a buyer, you could say, hey, I need to change my CRM vendor, just give me some options. And if you are just a CS rep, maybe you talk about churn, right? So everybody's got some problem. That is top of mind. They share that. And then we have this AI research that we do, and then we bring experts and peers that are having the same problem, or experts in those problems. Our agents go out and proactively keep thinking about this problem throughout the week. So it's not a one-time search like Google does. This is a continuous search and as soon as the agent finds some interesting information, it pipes it in your Slack. So this is going on and meanwhile there are seller agents and buyer agents, because eventually we want this to turn to leads and sellers to engage with content.

Santosh Sharan: 38:11

So what we have done is we have built just the seller agent and we want some sellers to start adopting standalone. It'll take us maybe another three months to build this whole research platform, but once it does, the seller agent and the buyer agent bolt on to this platform. We also have built a buyer agent where buyers can send out the agent and then this agent goes to sellers and says pitch to me, I'm working on behalf of Santosh. So instead of bothering Santosh with your pitch, santosh is in market to buy a CRM. You can just pitch and then it comes back to me with a grade saying I spoke to 10 CRM vendors, but these are the two you should really care about, because their responses were best scored against the five qualifying question you gave me Amazing.

Kevin Kerner: 39:00

Wow, and they're always. The thing that really gets me is, unlike search, when you have a question they're always working to answer the question.

Santosh Sharan: 39:09

They're always analyzing looking for the next person. Yeah, they're always going to LinkedIn, google searching and also, as the capabilities increase, they'll be scheduling calls and showing up on calls on behalf of me.

Kevin Kerner: 39:21

Crazy. I mean, that is just a whole new dynamic and I can't help thinking about it. A combination of like where I go to get information, Google and LinkedIn where I go, and Reddit and G2 and other places. It's really. It's really a unique connection between, for a buyer, like somebody who's buying me, a small business owner, let's say, seeking some sort of answer to something. It's really because there's a lot of utility. Your agent has a lot of utility for me because I could go have it work on my behalf and find the answer and I don't have to go through a bunch of Reddit threads.

Santosh Sharan: 39:57

Right or search Google.

Kevin Kerner: 39:58

It's amazing.

Santosh Sharan: 39:59

Yeah, right now those agents are just like search engine or they're just returning information to your questions. But we are already seeing the advent of agent and we certainly have on our roadmap where I could give it tasks to like execute and come back to me after like a few days or week, or it doesn't have to be done right now, at this instant.

Kevin Kerner: 40:23

That's where it gets really powerful man, santa, just like well, we could go on forever, because I remember the old days when we would go on forever. I know we both you have limited time. I know this is somehow linked. At some point on another podcast, you and I are going to talk about your ability to. This sounds like something that's sort of genuine in you, because on LinkedIn you're answering questions. You're giving a lot of value, like you're a brand of your own on LinkedIn. I can see how the kernel of some of what you're doing there actually has affected this part of the product idea. But I don't think we're going to have time to get to that, but we'll do it at some point in the future.

Kevin Kerner: 41:04

I definitely wanted to get to my AI roulette question. So this is a question I do with everyone and I load just your profile into Perplexity and I hit a button and then Perplexity gives me a question and I've instructed Perplexity to be kind of sassy and, you know, maybe controversial in some way. So let me go ahead and do this. I just literally just put in your. I'm now interviewing Santosh. Here you go. Okay, here's your question. It's a little bit long, santosh, given your deep experience with both data-driven sales and now AI agents. Let's play with a unique scenario. It's 2030 and 80% of B2B buyer interactions are handled by AI agents, but research finds that most buyers can now spot AI. Speak the sanitized, overly polite, jargon-filled responses. In fact, a group of meme marketers M-E-M-E meme marketers armed with humor and human quirks, begin outperforming even the most advanced AI-driven campaigns, because buyers crave authenticity and a good laugh. So here's the roulette question If AI drives efficiency but meme marketers drive trust and engagement, which side would you bet on to win over the next generation of B2B buyers?

Santosh Sharan: 42:22

That's just silly. Would the memes win?

Kevin Kerner: 42:26

Does meme win or I don't know what it's asking?

Santosh Sharan: 42:28

Yeah, no, I think there's a profound lesson in here. I'm amazed by how it came so eventually differentiation wins. So what it's saying is look, authenticity and scale and trust can be scaled through AI, let's say in the next five years. Let's say in the next five years, and then through memes which is silly you can tell which creatives are done by humans or which campaign. I think the memes will win hands down and there's a constant battle for humans to seek trust and credibility. Even if it comes by I mean that's kind of even if it comes through like some gimmicks.

Kevin Kerner: 43:16

I agree. I was kind of thinking too the memes would win, that meme marketing could become a whole thing. It already is a thing, but maybe it's the thing that takes down the AI stuff.

Santosh Sharan: 43:27

Yeah, there's going to be AI noise after like 2030. You know what?

Kevin Kerner: 43:32

It's really weird too to see all the AI video stuff, because a lot of the AI video stuff that's on MVO or Kling or whatever is kind of torching the AI space. So it's using AI to say this is ridiculous. Why is this happening? It's just funny to see. Well, it's so great to talk to you. I don't get to talk to you enough, so I had to. I had to build a podcast so he could get on. Get on a call together.

Santosh Sharan: 43:58

A lot of conversations. Kevin, let's do it again.

Kevin Kerner: 44:01

Definitely. So how does I know you're active on LinkedIn? So I would encourage I'll put the LinkedIn your LinkedIn profile in the show notes, but how do people learn more about Xero?

Santosh Sharan: 44:13

So we are intentionally kind of keeping it under wraps. One of the things I learned while at RB2B is the benefit of working in public, but also the counterintuitive reasons not to do some of that, because we created like 20 other competitors in the process. So and then building these tools are so easy these days and what we are building is complex in idea, but it's taking time just because we have less resource. So the best way to learn about what Zier is doing is reach out to me. I love talking, I love sharing and more often than not, I love to share what I'm doing and get your perspective. I don't think I'll be mentioning it on a website just yet. If we do, we'll just talk about the sales agent, what we are launching, but I'm more excited about the overall vision Sounds great.

Kevin Kerner: 45:13

Yeah, and I can say that you're super active on LinkedIn and you're always willing to share perspective and comment and stuff. You're just a great friend and thought leader to have out there. You're my in part, you're my AI thought partner, Santosh. I still need Claude and chat, but you're not that good, but you are pretty good, All right. Thank you so much, santosh. We'll talk soon. Cheers, cheers.

‍