How InMobi Is Reinventing Advertising with AI Agents

Summary

Marketecture: Get Smart. Fast.

Episode Summary: How InMobi Is Reinventing Advertising with AI Agents

Host: Ari Paparo
Guest: Abhay Singhal, Co-founder & CEO of InMobi Advertising
Date: July 15, 2026

---

Episode Overview

This episode features Ari Paparo in conversation with Abhay Singhal, focusing on InMobi’s recent innovations, particularly its integration of AI agents in advertising—including agentic adtech, programmatic platforms, the Glance shopping platform, notable partnerships, and product launches. The discussion highlights InMobi’s vision of transforming how consumers interact with advertisements and e-commerce experiences using AI-powered agents and new user interfaces.

---

Key Discussion Points & Insights

1. Reimagining Advertising with AI and Agents

[01:41]

• InMobi recently launched its Agentix solution, signaling a deeper move into “agentic advertising” and “agentic commerce.”
• Agentic Advertising: The shift involves ad formats and platforms that are increasingly powered by AI agents capable of both decision-making and interacting with supply/demand.
• InMobi’s consumer brand Glance is pioneering inspiration-led shopping experiences built on agentic infrastructure.

“We are squarely in the middle of agentic advertising and agentic commerce… Glance, which is taking a leading position on defining inspiration shopping through the agentic infrastructure.”
—Abhay Singhal ([01:41])

---

2. The Agentix Selling Platform & Infrastructure Evolution

[02:26 - 06:00]

• Agentix Platform: Allows buyers direct integration with InMobi’s exchange, bringing their own algorithms/infrastructure for buying decisions.
• Moving away from “big software” towards environments where buyers can interact with InMobi via cloud providers or custom infrastructure, potentially through natural language chatbots in the future.
• The underlying system leverages the ADCP protocol for universal interoperability.

“The traditional big softwares are no longer needed, where somebody can just do a conversation and it can do the work for you.”
—Abhay Singhal ([02:44])

• Role of AI is shifting: From merely executing buyer-defined instructions to interpreting broader business objectives and acting intelligently.

“How about we imagine a world where we tell AI what we are looking to achieve... and let it figure out what it does, how it does. And that is the world which I am imagining.”
—Abhay Singhal ([04:36])

---

3. The Changing Media Planning Funnel

[03:16 - 04:47]

• Advertising is being driven closer to the point of transaction as agents handle much of the “research” phase autonomously.
• The implication: Advertisers need to focus on presenting their value much closer to the transactional stage since consumers may bypass traditional research steps.

---

4. The Future of Buying Infrastructure/DSPs

[06:16 - 07:00]

• The tech stack is becoming less UI- and button-driven; the next era might be conversational interfaces enabling a more fluid, AI-driven workflow.
• Integration flexibility means buyers could use their DSP, cloud infrastructure, or entirely new interfaces.

“The era of pressing 10 buttons is very clearly over.”
—Ari Paparo ([06:46])

---

5. Glance: Agentic Shopping Platform and Samsung Partnership

[07:00 - 10:45]

• Glance: An AI-powered consumer shopping platform embedded directly into devices (phones, TVs, laptops) that offers tailored, visually-rich product recommendations.
• Users interact with Glance’s agents to refine recommendations and receive personalized suggestions—often leading to unexpected discoveries.
• Samsung TV Partnership: Glance is now integrated into over 50 million Samsung TVs globally, alongside Copilot and Perplexity, shaping the interactive shopping experience via TV.

“Glance is an agentic shopping platform. Think of it like an agent that is trained at an individual level of one… It is learning your buying habits, understanding your style, helping you shop.”
—Abhay Singhal ([07:18])

• Unique in attempting seamless cross-device experiences between TV and mobile.

“If you can think of no other consumer app has been able to link mobile and TV in a seamless way. TV is a lean back experience. Mobile is a lean forward experience. So we are with Glance bridging both TV and Mobile.”
—Abhay Singhal ([09:28])

• Monetization is primarily conversion-based, with performance and “inspiration” as key metrics, making the business model akin to affiliate but with a focus on top-of-funnel engagement.

---

6. The Limits of LLMs in Visual Shopping

[10:45 - 11:17]

• Large language models struggle with contextually vague queries like “what’s a cool shirt?"—highlighting the need for visual inspiration and interactive refinement.

“The best LLM is still not going to be good at answering a question like what's a cool shirt?... it needs a visual inspiration.”
—Ari Paparo & Abhay Singhal ([10:45-10:58])

---

7. Acquisition of Mobile Action

[11:17 - 13:22]

• InMobi acquired Mobile Action, a key Apple partner for in-app promotional campaigns and keyword bidding in the App Store.
• Brings new capabilities for developers to acquire users and optimize Apple App Store campaigns, enhancing both reach and analytics.
• The team is small (<30 people) but handles a large volume of media and data, making it highly strategic for InMobi’s developer focus.

“Apple has a very mature advertising product where Apple allows the app developers to be advertising within the App store. Mobile Action was one of the very few companies that had partnership with Apple... to allow the bidding on keyword, then to get the intelligence and bring the analytics back.”
—Abhay Singhal ([11:38])

---

8. AI-Driven Ad Experiences: Fossil Case Study and Innovation

[13:22 - 16:02]

• Collaborated with Fossil (their first advertiser) to innovate on CTV ad experiences—making ads interactive and agentic.
• Used first-party data and AI layers to transform standard creative into a conversational experience, qualifying user intent before a landing page.
• Delivered 30% increase in lift and deeper engagement analytics, as consumers could ask questions and interact with the creative in real time.

“We've taken an existing advertiser creative and we have layered the entire agentic infrastructure on top of it...put the conversation on top of it, and then you qualify the intent even way more before you jump the person onto the landing page.”
—Abhay Singhal ([14:36])

“We are actually seeing about 30% overall lift increase in the lift. We're seeing much more deep engagement data on when the consumers are dropping off.”
—Abhay Singhal ([15:31])

---

9. Lightning Round: Company Culture and Challenges

[16:02 - 16:32]

• Strength: Deep understanding of mobile.
• Challenge: Keeping up with the pace of growth.
• If InMobi were an animal: “Tiger.”

---

Memorable Quotes

• “The era of pressing 10 buttons is very clearly over.”
  —Ari Paparo ([06:46])

• “How about we imagine a world where we tell AI what we are looking to achieve and let it figure out what it does, how it does.”
  —Abhay Singhal ([04:36])

• “Glance referred me this shirt… but I saw it myself on Glance and Glance recommended and I ought to buy it. There are many categories in the world that are very visual in nature and you should try Glance.”
  —Abhay Singhal ([07:44])

• “Apple has a very mature advertising product… Mobile Action was one of the very few companies that had partnership with Apple to bring that product to the market.”
  —Abhay Singhal ([11:38])

• “We are actually seeing about 30% overall lift increase in the lift.”
  —Abhay Singhal ([15:31])

---

Notable Segments & Timestamps

• Agentix Platform & Agentic Advertising: [01:41] – [06:00]
• Shift in Media Planning & Funnel: [03:16] – [04:47]
• Future of DSPs & Infrastructure: [06:16] – [07:00]
• Glance & Samsung TV Partnership: [07:00] – [10:45]
• Limits of LLMs in Visual Shopping: [10:45] – [11:17]
• Acquisition of Mobile Action: [11:17] – [13:22]
• AI-Driven Ad Experiences / Fossil Case Study: [13:22] – [16:02]
• Lightning Round: [16:02] – [16:32]

---

Final Thoughts

This episode is a comprehensive look inside InMobi’s aggressive push toward AI-empowered adtech, commerce, and consumer experiences. By embedding agents into every consumer touchpoint—mobile, TV, and more—they’re enabling seamless, conversational, and context-sensitive advertising that blends content, interaction, and commerce. Their ongoing product launches, acquisitions, and high-profile partnerships indicate a fast-evolving sector and a company intent on leading—perhaps even redefining—the next wave of advertising.

Transcript

[A]

This episode is brought to you by Google Chrome. You think you know a browser, but Gemini and Chrome, that's new. It can help you with practically anything on the web like restoring a vintage motorcycle from a 50 page restoration block. Or finally break down that long article you've had open for weeks. Gemini and Chrome is here for it, ready to make anything online make sense. There's no place like Chrome. Check responses set up required compatibility and availability various 18 foreign.

[B]

Texture where you can get smart fast with in depth interviews of leading technology executives. I'm Ari Caparo. I'm joined today by Abhay Singhal, the co founder of InMobi and the CEO of InMobi Advertising. Thanks for being here.

[C]

Nice to be here Ari. How are you doing?

[B]

I'm a bit little, I should have said we're at the UTA beach at Cannes and the air conditioning has been working hard doing its duty but by two in the afternoon and maybe needs a little restart.

[C]

Oh, we need all little bit of cooling in this hot attic world.

[B]

Wait, aren't you based in India?

[C]

I am, but I also spent enough time in San Francisco to know what the difference is.

[B]

Okay, right. So if someone from India says the architecture is not good enough, I'm like that's a different level.

[C]

Let's just say I'm spoiled. Okay, global citizen that I am.

[B]

But we're doing our best to bring you the ad tech news you want. So thanks for joining me. So I guess last time you talked to marketexture was it possible and you were launching your agentix solution. So tell us about that, what's been going on in the three or four months since and what you're seeing?

[C]

Well, it's been an amazing launch so far. We are squarely in the middle of agentic advertising and agentic commerce both as you know, InMobi as a brand consumer brand called as Glance, which is taking a leading position on defining inspiration shopping through the agentic infrastructure.

[B]

Okay.

[C]

And we've also taken a very lead position on coming up with new ad formats that are very AI enabled and trying to use the existing rails of RDB to bring in the entire agentic advertising. So couldn't be more excited to be talking about these things.

[B]

Yeah, so tell me a little bit more about it. So I think you, I don't know if you just launched or are going to launch a agentix selling platform. So tell me about what that is.

[C]

Yes, Agentix selling platform is a platform where the buyers can directly integrate with immobie exchange and they can bring their own Algorithms, they can bring in their own infrastructure to be able to make decisions on what to buy and what not to buy. As you know, the whole SaaS industry is going through a massive, massive shift.

[B]

Yeah.

[C]

The traditional big softwares are no longer needed, where somebody can just do a conversation and it can do the work for you. So as agencies and brands are looking to changing their own infrastructure.

[B]

Yeah.

[C]

We ourselves have evolved on completely on the ADCP protocol to allow us to be accessed through whatever the infrastructure that the advertiser chooses to have.

[B]

Right.

[C]

But also very important is the viewpoint that we are taking on the agentic advertising. Devi. That's what you want to talk next.

[B]

Sure.

[C]

So interesting part on the agentic advertising that I look at in the world is how the advertising is actually going to move more down the funnel.

[B]

Okay.

[C]

With agents doing the conversation with the agents. I actually feel that the whole component of advertising that used to come in during the process of research is going to get eliminated. What does that mean? Is that means that the advertisers have to be really careful about presenting themselves in a way which is much closer to the transaction.

[B]

Right.

[C]

Than to the conversation stage. And that is a viewpoint that we are taking on the advertising, which is sort of prompting us to launch a bunch of different ad formats in the, in the advertising space.

[B]

So. So like in the old way of doing things, you did a media plan and a research plan and then that determined what you wanted to buy. And you can use AI to do that still. And probably everyone is. But I think what you're saying is that if your AI is really just telling the live system whether you want impressions or not, then you're on a tighter line and you have to be more accurate and precise about what you want. Yeah.

[C]

And my view is that you're still not using AI to the fullest. If you are saying that, look, I need to tell the AI what to do. Right. How about we imagine a world where we tell AI what we are looking to achieve.

[B]

Right, of course.

[C]

And then let it figure out what is, what it does, how it does. And that is the world which I am imagining. Yeah. Going into the future.

[B]

And tell me about this ad CP integration. So a buyer can use cloud or they could use another vendor to interact with your system.

[C]

I mean, they can use a Scope three. Yeah, that's a beautiful. Scope three is a beautiful launch. They can come up with their own.

[B]

Yeah.

[C]

Interfaces as much as they would develop out there. And they can directly call my agent and then integrate with our agents to access the supply now the agents are in the process of getting fully matured. Of course they're going to keep maturing as we go along, but the journey is probably as much fun as an outcast in this scenario.

[B]

What's. What is the, what is created when the buy side agent uses your MCP server? Is it an order, like an insertion order, effectively that's created in your system that then delivers ads, or is it just. There's more like RTB filtering or curation?

[C]

It's more curation and RDB filtering because we're still building it on top of the RDB protocol. So it's still, it's a whole process of deal creation, audience segmentation, audience appending. All of those things are taken care of in this process.

[B]

Brilliant. So, so let's say I'm a, I'm an active buyer using rtb. I have my DSP of choice and then I integrate with you. I'm now selling my AI. You know, get the Inmobi inventory that meets my media plan, create some deals. Here's the prices I'm willing to pay. But then it's still running ultimately through my DSP and then potentially my ad server as well.

[C]

You could choose to run it through that or through any other infrastructure.

[B]

Right.

[C]

That you are more suited for. And which brings to an important question that what the future of the DSP world may look like.

[B]

Sure.

[C]

Which is another topic that we can spend hours talking about. But yeah, I mean, whatever the infrastructure is that the buyer chooses to do, to interact and maybe in the future those, those infrastructures are going to be purely a natural language chat bots and not really any more. Click here and go there and press 10 buttons.

[B]

Yeah, the questions, the era of pressing 10 buttons is very clearly over.

[C]

Absolutely. And that's why when you say that, which infrastructure they will use, I'm like, it could be just a simple chatbot or it could be much more complicated than that.

[B]

Yeah, absolutely. So you've been announcing a lot of things at can. So the first thing I want to bring up, you had mentioned Glance, which is a product that you offer and now there's a Samsung partnership. Can you walk us through what that's about?

[C]

Oh, absolutely. I mean, look, Glance has a very unique distribution strategy.

[B]

Okay, why don't you explain what Glance is for people who don't really know?

[C]

Absolutely. So Glance is an agentic shopping platform. Think of it like an agent that is trained at an individual level of one.

[B]

It's a consumer product.

[C]

It's a consumer product and it is helping you Shop.

[B]

Yeah.

[C]

It is learning your buying habits, it is understanding what your style is and it is helping you shop. Right. But it is helping you shop in the categories that are very visual in nature. Categories that are very inspirational in nature. For example, look at this shirt like

[B]

it's a nice shirt for those, those just listening on the podcast. Yeah, this, he's wearing a nice blue shirt.

[C]

This Glance, Glance referred me this shirt and I won't, I won't, I won't buy it usually like I don't, I don't wear these sort of wavy patterns as you can imagine. But I saw it myself on glance and glance recommended and I ought to buy it. So there are, there are very many categories in the world that are very visual in nature and you should try

[B]

glance and I'll, I'll do an AI enabled shirt.

[C]

Give us, give us a feedback on that. And so that's what we've done. But we have is a very unique distribution strategy. Our distribution strategy because it's an agent, we wanted it to embed in various different consumer surfaces that the consumers use on a day to day basis. So for example, consumers use their phone, so we've embedded that into the devices itself. Consumer uses TV so we've embedded that into the TV itself. Consumer uses laptop, so we are integrating them into the laptop on the edge. So we're integrating them onto different surfaces where Glance as an agent is going to appear and help you shop in a best fashion. So what we announced with Samsung is a partnership with Samsung TV. Now glance is going to be on 50 million plus Samsung TVs across the world and it will, it's one of the three partners that Samsung is integrating. The other two partners are Copilot and Perplexity and we are going live along with them and we're going to redefine how shopping happens within the Samsung TV universe.

[B]

I have a Samsung TV at home so I'm looking forward to buying your shirt using my remote control. If we could do this. You're going to be that, that'll be a miracle.

[C]

Oh, love that. You know the interesting part Ari, is the holy grail of TV is when you can link TV with mobile.

[B]

Yeah.

[C]

And there are very few apps. I mean YouTube has been able to do so very conveniently and a bunch of our people, but if you can think of no other consumer app has been able to link mobile and TV in a seamless way. TV is a lean back experience. Mobile is a lean forward experience. So we are with glance are bridging both TV and Mobile. So when it goes live in your TV environment, you'll be able to see a very deep seamless integration between the TV and your mobile device.

[B]

Advertising supportive.

[C]

It's primarily monetization model is depending on how much conversion we drive for the merchants.

[B]

Right.

[C]

That makes sense because

[B]

akin to affiliate.

[C]

I think you may say so. But my view is also on the inspiration part because if you look at in a world of agentic advertising, how would merchants be able to get the consumers to get entered into the top of the funnel? Like it's a very pertinent question. As the search is kind of going away and consumers are interacting through the chat interfaces, one has to ask a question that how does somebody enter the top of the funnel? So maybe the pricing model is going to be an affiliate, maybe the pricing model is going to be top of the funnel. But at the end of the day we know we'll be measured on the performance. Yeah, we will deliver.

[B]

I mean the best LLM is still not going to be good at answering a question like what's a cool shirt? Right. Because like it needs so much context. It needs a context and that's, that's

[C]

too vague and it needs, and it needs a visual inspiration.

[B]

Yeah.

[C]

Look, chats are training us to be talking.

[B]

Yes.

[C]

If you, if you'd experience glance you'll realize that we actually show you an image and you feel good about it and then you can interact with the agent and keep guiding it, keep training it, keep nudging it in the directions that you want it to be nudged and you'll realize you'll be surprised with the answers that you're getting.

[B]

It's fascinating how you've diversified the company in Moby. You know, it's never, never quitting on adding new stuff. Speaking of which, another announcement that you've made at can is the acquisition of a company called Mobile Action. This is a San Francisco based in app promotional company. Tell me about it because I'm not that familiar.

[C]

Yeah. So I think look, Apple is a very, very vibrant ecosystem and Apple has, most of people won't know about it but Apple has a very mature advertising product where Apple allows the app developers to be advertising within the App store.

[B]

Sure it's a keyword based and it's

[C]

a keyword based and it's a very, very, it's developer's first choice and pretty much every developer uses that product to basically increase their reach. Mobile Action was one of the very few companies that had partnership with Apple to bring that product to the market. They were probably one of the three companies in the ecosystem to allow the bidding on keywords, to allow the bidding on keyword, then to get the intelligence and bring the analytics back and so on, so forth. And so that's why we got the company with us. And it fits into our thesis. We've been working with developers for the last 15 years through our DSP product and we thought that diversifying it from just a third party in app supply to bringing in more premium Apple App Store is not only going to give us more opportunity for our developers to experience the Apple products, but also give us, give them the intelligence that is needed to make overall advertising more effective.

[B]

Incredible data too.

[C]

I'm sure it's an interesting, it's an interesting insight.

[B]

Any, any details you give about the size of the company or anything.

[C]

The company is pretty large and very small. It's, that's probably, it's probably less than, less than 30 people in total. But in terms of the media that they handle is incredible because it's like search. Like, it's like search. They do a huge amount of dogs, they do a large dollars and they are, they do well for themselves.

[B]

30 people. All right, so mobile action. And then lastly you also announced a, a partnership with I guess Fossil for ad experiences on TV or Fossil was your first advertiser. Tell me about this Ad experiences product on ctv.

[C]

So look, we're working on bringing ads in AI and AI in ads for the last I think six years towards that. We did this with Fossil. You know, we're going to come out with a bunch of other case studies. By the time this podcast will go live, I think there are going to be a bunch of other case studies that we can talk about. Yeah, but the Fossil case studies were very specifically trying to link Glance tv, which is our marquee consumer experience, and linking that onto the third party and trying to figure out how much lift can we generate when we use the first party data to bring that into the third party ecosystem. So as much it was about the ad experience where we were actually our biggest innovation was taking in standard creative of an advertiser and making it completely agentic conversation. The insight came from Ari that ads and you and I have been in the ad industry for very long. As a marketer, we've always wondered that ads should generate some sort of a curiosity, right?

[B]

Sure.

[C]

But typically that curiosity was always measured by a click. Now when a consumer is seeing an ad before it can generate a click, a consumer may have 10 questions about the product itself, which they're not Asking here, which they are either going and searching on a Reddit or somewhere else. What we've taken is we've taken an existing advertiser creative and we have layered the entire agentic infrastructure on top of it to take creative and convert that into a conversational creative itself. Which means you take an advertiser's creative, put the conversation on top of it, and then you qualify the intent even way more before you jump the person onto the landing page.

[B]

So this is sort of a top of fun or conversation with the consumer that before they click in, before they get.

[C]

Absolutely. And it's just. And there is, there's an incredible amount of data that we are seeing. We are actually seeing about 30% overall lift increase in the lift.

[B]

Right.

[C]

We're seeing much more deep engagement data on when the consumers are dropping off. Remember, earlier advertisers used to run their survey campaigns outside the advertising and then they would try to understand what's going on in the advertising. Here you're actually running it in a very real time, integrated fashion and you're able to see much more data from the advertisers. So that was the, not only the fossil case study, but case study with tens of advertisers that we are bringing and probably the most credible example of ads in AI and the AI in the ads that we are seeing in the universe.

[B]

Fascinating. All right, so let's do a quick lightning round to finish up. Quick questions, quick answers. Thinking about in MOBI as a whole, what's your greatest strength?

[C]

Understanding of mobile.

[B]

All right. And what is the biggest challenge?

[C]

Keeping up with the pace of growth.

[B]

Yeah, always. And lastly, if in MOBI was an animal, what animal would it be?

[C]

Tiger.

[B]

Tiger. All right, well, thank you so much for being here. Abhay Singh Hal, the co founder of InMobi and the CEO of InMobi Advertisement.

[C]

Thank you, Ari. It was nice to be here.