ChatGPT Ads Are Here. Now Comes the Hard Part.

0:02Mary Gabrielyan: This episode of Inside the Stack is brought to you by AI Digital. AI Digital is an AI native media consultancy and innovation partner helping agencies and brands navigate today's fragmented media ecosystem.

0:26Mary Gabrielyan: Welcome to Inside the Stack. Inside the Stack is a custom podcast series from Ad Exchanger, where we interview leaders in ad tech.

0:41Mary Gabrielyan: Hello, everyone. I'm Lynn D. Johnson, Senior Director of Content and Community at Ad Exchanger. And today, I'm talking with Mary Gabrillion, Chief Strategy Officer at AI Digital. We'll be chatting about AI and advertising and taking a deep dive into advertising on ChatGPT. Marketers, if you're wondering whether you should take the leap, we have your answer here. We'll get into why advertising on ChatGPT resembles performance search and what's needed for broader marketer adoption. Hi, Mary. Welcome to Ad Exchanges Inside the Stack podcast. Hey, Lynn. Thank you so much for having me today. Yeah.

1:34Mary Gabrielyan: Happy to finally have you and to hear your perspective on chat GPT advertising. I really want to go into this with you because a lot of brands are wondering whether they should enter or not, right? So before we get into all of that first, I'd like to know a little bit more about you, you know, who you are, your path, your journey. I know that you've been on the agency side at Publicis. You also went client side as a CMO. And now you're heading strategy at an AI Native consultancy, usually people pick the lane. What did moving across all three sides of this industry teach you? What do you think the skills it gave you for what you did before to where you are now?

2:31Yeah, that's actually a great question. And to hear the recap like that actually makes me think. however it has definitely been a journey I have always been passionate about media and advertising and marketing and in all the sides agency hold code like client side AI digital side I'm always with my passion around media and how media advertising works and how it allows brands businesses to achieve their goals. Definitely, although these three companies look pretty different, but definitely that they have that common denominator, which is media and my passion for media. Right now, I'm the chief strategy officer at AI Digital, which is an AI native media consultancy and an agency, which is actually specializing again in media and AI and everything around transformation and AI consultancy for media agencies.

3:40Okay.

3:41Mary Gabrielyan: So let's dig into a little bit more about what that means. AI media consultancy for media agencies, but also I know you guys are built around an open garden framework. What is the 30-second pitch on what that all means? Okay, so I'll give you two 15-second pitches, first about open garden and second one about what does AI Native Media Consultancy actually means. So when we say open garden, we are deliberately putting an antonym towards world gardens, which are actually all over the place in the media industry. The walls are getting higher. The Facebooks and Amazons and all the players like that, they are building a higher and higher walls.

4:28They are less and less likely to share. Actually, with the rise of AI advertising, such as GPT ads and whatnot, this industry that has already been sliced and diced is becoming even more tortured. At AI Digital, we are opting for an open garden framework where we are elevated above that whole fragmentation. We sit on top of all tech, including DSPs, SSPs, data providers, and frankly, any technology that you need for a successful media campaign. And we're able to create the best-in-class tool stack for our clients. That's what it means to be following the OpenGarden framework and OpenGarden principles.

5:17In terms of what AI media consultancy means, so every single word there has a purpose. We are an AI native company because we have been called AI digital way before AI was a thing. And when like eight and a half years ago, it was just machine learning and algorithmic optimizations, which our team has adopted from day one, we have been AI digital since day one. So we are truly AI native and not just hopping into a marketing trend. Second, we are a media company. We have services like managed service. We do supply management and ad curation. We definitely do AI consultancy. And that's where the consultancy piece comes in.

6:02So not only we cater to media agencies, amazing media strategies, and run the campaigns for them and have amazing reporting, But also, we are able to hand the hold of our media partners and help them to go through this whole AI disruption without any hurdles. Because we, as an agency, we have walked the walk. And now we can teach our partners to do the same with the help of our AI Labs team. So that's why AI media consultancy, every single word, stands behind a huge development and a huge pillar on which AI digital is based.

6:45Mary Gabrielyan: Okay. I really like the part where you said you're not AI just for the sake of AI. You've been AI since day one before this AI blow up. Like as soon as you said that, I was like, clock it. So let's get into what's happening in the space right now. OpenAI, as we both have discussed, has opened up advertising. They were at Cannes to announce that they're opening up more markets. You know, CPC bidding is live. There's pixel tracking, a conversion API, and the minimum spend is effectively going to zero. I saw something you wrote about this. What is different about this moment? Yeah. So first thing first, OpenAI with ChatGPT, they cannot sell purely subscriptions to make sure they are seeing the whole revenue promises they are anticipating and all the future investors are going to anticipate.

7:57So the advertising and an AI company becoming a media company is the only logical sense. We have amazing, amazingly successful companies like Google, Amazon, Meta that have become media companies. We see how Walmart has become a media company, how Kroger and Coastco are turning into media companies, how financial institutions like Chase is turning into a media company. So any company that has client data is adopting this route or has adopted this route and has proved to be very, very successful in terms of revenue earnings. And I think that is the blueprint that OpenAI is following. So they do need this to unlock more revenue.

8:43They are going pretty quickly, but they are still not a media company. They are a tech company, first thing first. I'm happy to see that their functionality and capabilities are becoming broader, but in comparison, it's still super limited. So, and again, this is going to be more accessible. This is going to be more available to all the countries, companies, brands, regardless of the sizes. Let's think about it as Instagram ads. Like when any Instagram user could like, I still vividly remember those times when an Instagram user started to place an ad on any, frankly, anything with any budget they wanted.

9:27So I think we're monitoring that exact moment with ChatGPT ads. However, it is very specific because in terms of the format and in terms of the exposure to an ad, it is still pretty limited because people are not consuming content there. People are not spending huge amounts of time there. It is going to be basically more similar to Google search and Google's performance tactics versus mimicking like top of the funnel, CDB, programmatic video, programmatic display, or other more road tactics like that that are able to build that top of mind to become that seamless experience that is consumed by a user without the realization of the user.

10:20So in my head, and maybe I'm mistaken, and maybe I'm totally going to regret for these words, the future of ChatGPT ads, I think, is going to be contextual intent-based performance. So depending on what's the context of this user, what are all their searches within ChatGPT, the user is going to be fed these ads. they are probably going to be heavily conversion focused. It's not a coincidence. They have unlocked that API first thing first. So is it going to replace the whole funnel of tactics and channels? Of course not. Is it a beautiful supplement to your lower funnel tactics or like even mid-final tactics like consideration, engagement, extra nudge?

11:15100 % yes. So I'm very excited to see what's going to unfold with this, because honestly, we as an industry have monitored this exact thing happening once a couple of years ago when there was the boom of programmatic and then the social media ads became a thing and how the funnel was getting supplemented by more and more tactics and extra touch points. What's happening right now, ChatGPT is becoming one of the touch points. It's still not used. It still isn't substituting Google. It has the potential, but it's still not substituting it. And I'm very excited to see how the things are going to unfold.

12:04Mary Gabrielyan: Right. It's another tool in a marketer's toolbox, basically. Yeah. And I think you touched on something when you talked about performance, right? Like, yes, there's more access, but the measurement, the measurement is the real piece. And I think what you wrote about, you made a clear distinction between access and measurement. Yeah. And what will it take for ad spend? Yes, people can experiment to see what happens there. But what do you think it will really take for ad spend to make a lot of sense for a marketer? Yeah, just like you mentioned in your question, it is definitely how we can attribute, how correctly we can attribute, how we can attribute it not directly from chat GPT, but how can we see the chat GPT and influence on other channels?

13:06can we tie eventually the online experience and online discovery or engagement or add interaction with the offline events that is for example possible with google however we

13:22Mary Gabrielyan: don't have that answer for chat gps yet how it is being measured in the in the variety of within other tactics. What I'm trying to say is basically you have many, many tactics, you're running simultaneously, you have some kind of omni-channel strategy, you are doing a marketing mix modeling, how is Chagipity entering that marketing mix modeling? How are you sure? How sure can you be that you are attributing it correctly? And does Chagipity ads do Chagipity ads have that tool set right now? Because right now they don't. Will they create it? I think that is the way to unlock more and more attention and more trust and more reliance on this type of ads.

14:17And another thing that is super important to remember, the ads are not available for plus, pro team enterprise or education tiers, which means that a good portion of individuals, especially like the knowledge workers who are day-to-day active users of ChatGPT, aren't fed these ads. So are they going to change their strategy in terms of their ad-free tiers, I don't know, but probably they are going to reconsider it because to my taste, and maybe I'm being delusional here. However, I see ChatGPT, Claude, and all these AI tools to become basically knowledge workers' number one asset, like day-to-day workspace, Just like there is Slack or Microsoft Teams and the whole Microsoft pack.

15:21I see the future there and I see them earning money there a lot. So are they going to make it ad-free or no? I don't know. But money definitely lies there.

15:35Mary Gabrielyan: That's interesting that it's more of an enterprise workflow kind of tool than say something that supplants Google, right? As a search tool. Yeah, that is a very interesting. And you know, you touched on something because I was wondering, people kept telling me, I'm seeing ads in the wild on ChatGPT. And I'm like, well, I have not seen one, but now I know why. I didn't know that if you paid for Plus or Pro Education, you didn't see it. Yeah. I think also what's interesting is that You talked about the lower funnel before. It's like, I feel like there's like this compression of the funnel when you're using ChatGPT because you can make the decision immediately, right?

16:24Mary Gabrielyan: You're looking for the thing. You're making the decision. You're, yeah, I think there's a lot of things that marketers will have to factor in as they make these choices. but I'm curious though is there you being an open garden company built on the premise that you know walled gardens can be a disservice to advertisers at times right ChattyPT is trying to become another walled garden how do you like How do you help marketers align on this idea of, you know, trying this new channel, but yet it's still a closed ecosystem, right? That's a great question. And actually, you are absolutely right. Chajupi not only is trying to become another world garden, it is literally another world garden.

17:32Mary Gabrielyan: World garden. With its own data. Can you imagine a world where Chai GPT or OpenAI shares its data with Amazon? Like, probably that's not happening. Right, right. Just like Amazon doesn't share his own data with anyone. So we as an industry already cannot avoid vault gardens. What we can do is we can make sure we have access to every single one of it. And we have an independent intelligence layer, measurement layer, that is able to collect all this data from all these world gardens and to report back to the marketer in one single analytics. That's exactly what we are doing at AI Digital with our Elevate platform.

18:27And it's not a coincidence. It's called elevate. As you can see, we love the workplace. So it is elevated above the industry. And it is that intelligence layer that allows you to collect all the data, to do math modeling of everything, and to report to you, okay, this is what's really happening, considering all your channels and considering the whole ad spend that you are having. this is your result, not independently, not per channel, but combined result. So that's what we are working on actively. And I think that is the best justification of our OpenGarden framework. We are not saying we're avoiding the WorldGardens.

19:12We cannot avoid placing ads on Amazon or OpenAI ChatGPT ads. We are all about it. At AI Digital, we are running multiple tests there and we're basically preparing every single answer for every single possible question that might arise amongst our clients so that we're able to tackle them as we place ads scaled. So we're not going to avoid ChatGPTS because it's another walled garden. The industry is full of them. We just need to be better and one step elevated above them and to see the whole picture combined. So that is like, that's why being independent, being not tied to any vendor, any DSP, any SSP contractually, being able to become the neutral player that is going to be able to choose the best in class technology or platform or ad format for your client.

20:25And that makes you unique, actually. That is our core differentiator. And that's why we are achieving just, frankly, better results on behalf of our clients. Right.

20:41Mary Gabrielyan: That's very interesting that having that layer that you talked about, because I know transparency is a big problem for marketers. and being able to provide them with some level of transparency is always a good thing. I'm curious, though, how will you be able to help them see visibility into LLM, like LLM visibility as well? So do you mean how they are appearing in different models? Yes. So actually, like, there are like big SEO players like Ahrefs or SEMrush that are actually like tackling that problem very well. But also our AI Labs has built a tool, for example, that is able to scarp lots of LLMs, including like Anthropics, including OpenAIs and a couple more, and show you how you appear depending on different contextual queries that users are putting there.

21:52So plus, not only that, but this tool also allows you to understand, okay, what are the improvements that I can do so that I appear there for sure, or I appear there for additional queries, or I don't want to appear for these topics, what is my to-do list for that? So there are tools, our AI Labs has created one as well for our clients. And actually, our OpenGarden framework applies not only to DSPs and SSPs and that neutrality, but also neutrality to different LLMs, different large language models. So when there is a new model out, when there is a new version out, our team explores it, learns it, creates a special newsletter for our clients so that everyone is kept in the loop.

22:50And we definitely share that open garden and like neutral position in every single aspect of work, not only the media stuff, but also AI side of things. so ai digital is switzerland that's that's the best way to look at it i'm afraid you know we're a global company but i'm not mistaken we don't have anyone from switzerland which is no but i

23:21Mary Gabrielyan: mean like being neutral right being neutral switzerland is always neutral they don't get into any of the drama yeah so you guys don't take sides you're just like here here are the players here's what's available to you here's what yeah yeah if there is one side that we are taking is our client side that's the only side we care about okay and that's an important distinction for sure. Let's talk about, I think it's an elephant in the room. You know, ChatGPT was getting all this buzz for a long time, OpenAI, ChatGPT, but Google was like, boom, here's Gemini, and it's baked into a tool we already have all these hundreds of millions of users on, right?

24:14Mary Gabrielyan: but yet they're not throwing ads in that chatbot yet what do you think about that so a wonderful question and I think Gemini is in terms of the fame it is lagging a bit behind like Anthropic and Claude and ChatGPT they are like they have created more buzz. But I think Gemini is going to be that one like smooth, great cardinal that comes into our laptops, phones and our, frankly, ad platforms, BD360 and whatnot, and starts reshaping interaction in every aspect of work with Google stack. I think Google is, Gemini is probably like more interesting for marketers and for media people because it has that media backbone.

25:18So Anthropic and OpenAI, they don't know anything about media. They are hiring very talented people. Don't get me wrong. For example, OpenAI hired the chief strategy officer from StackAdapt. They hired the BPF ads and monetization from Meta. So they are building that infrastructure, but Google has all of that.

25:44Mary Gabrielyan: Has an infrastructure. Not only that, but Google has millions and millions of advertisers that are using Google's ad platforms daily, Google's SSP daily. So I think they have that advantage. And we only like, I don't think we will have very direct Gemini ads. But I think Gemini will become a big part of user interaction with ad platforms, with optimization strategies that Google is proposing, with research and with like, probably search ads are going to be influenced by Gemini and by this AI overview answers that are given. So I'm very excited to see how they manifested into life. But I think they are positioned to beat the competition in terms of AI ads.

26:42Mary Gabrielyan: Right, just because of how much data they have on users already. anyway. It's interesting. I asked Gemini recently. I was like, hey, are you remembering things I'm telling you? And it's like, it said, no, I'm not actually. This is using it in the search, though, not using it separately as just Gemini. It claims it is not retaining my conversation. so I was curious about that because I'm like well if it's not then how is it going to note a personal give me more of a personalized experience so I'll be interested to see where the future of that goes as well yeah so so to wrap things up

27:32Mary Gabrielyan: I like to ask people for three words about their particular space within the advertising ecosystem. And so for you, it's AI and advertising. What are three words and why these three words that you would use to describe the future of AI and advertising? Okay, I love the question.

28:03If you allow me, I will answer to a slightly different question, which is definitely like, it's not about describing the future of AI and advertising, but rather like how I perceive that future. and my three words would be some things are better human it is more than three words I know although although I work at AI digital I I am a firm believer and like we're in a front line of AI usage AI in media in marketing in creative I believe that some things are better human and some things are better with the enhancement of that human intelligence. It's an unpopular opinion. I know it's a hot take, but I cannot deceive myself.

29:00Mary Gabrielyan: Okay, so usually I end here, but because you went in that direction, I cannot end here. And I want you to just leave people with, especially when we're talking about marketers or agencies right where is it that it's important to keep a human in the loop in the lead however you want to phrase that what what do you see there yeah so definitely we that we need okay ai is going to replace lots of things a lot of things like probably most of the things. However, I am a firm believer that human course correction is important, especially in creatives, especially in, yeah, probably especially everything around creatives and how creatives are perceived by real people who are fed those creative ads.

30:02Other than that, it's probably if you are delegating, let's say the buy side and sell side completely to AI. And these agents or these AI tools can be set up in advantage to one side or the other, making sure that there is this super human supervision that the client at the end is not getting ripped off. So that's what I mean when I say some things are better human is, yes, we are talking about delegation, we're talking about the product activity and how much time it unlocks and how effective people are getting. But when we are doing those things without supervision and without truly like thinking, okay, but like if you can set up everything to work advantage of a small group of people or a small group of companies, how we can make sure that client is not getting hurt and how we can make sure that we're not just blindly inflating things because we accidentally wanted to be more productive.

31:11Okay.

31:12Mary Gabrielyan: All right. So I think we can leave it there that we still need the human and creative and we need human course correction. There needs to be guardrails, as some might say. Yes, exactly. All right. Mary, thank you for stopping by Inside the Stack. I've really enjoyed this conversation with you. Thank you, Lee, so much. This episode of Inside the Stack was brought to you by AI Digital. AI Digital is an AI native media consultancy and innovation partner, helping agencies and brands navigate today's fragmented media ecosystem.