You have to be crazy to advertise on ChatGPT!

Short answer: yes. Long answer: this episode.

ChatGPT is running ads at $60 per impression. Not per click — per impression. That’s Super Bowl money, for a platform where the majority of users still aren’t seeing any ads, you have almost no visibility into who you’re reaching, and the audience came specifically for an answer — not an ad.

Mordy and Miruna break down why this feels less like a media buy and more like a fundraising mechanism for OpenAI’s investors — and why the brands currently signing up tend to be the AT&Ts of the world, not the Verizons. Second-tier players in their categories spending Super Bowl budgets for approximately none of the Super Bowl upside.

The bigger point: if your brand isn’t showing up in AI results organically, an ad won’t fix that. It’s a brand problem, not a budget problem — and the solution is building the kind of presence that AI wants to mention without being paid to.

Jordan Koene, founder of Previsible and host of the Voices of Search podcast, joins to talk about exactly what that looks like in practice.

Transcript

Mordy Oberstein (00:00)
You hear that? Of course not. That’s the whole point.

Welcome to Say It Anyway, where we say the things about digital marketing that you’re supposed to kind of keep quietly to yourself. And we say it — wait for it — anyway.

Each week, we’re taking an honest, maybe uncomfortable look at something marketers need to know, but might be too afraid to say anyway.

I’m your host, Mordy Oberstein, the Head of SEO Brand at SE Ranking, and I’m joined by she who has a slight cold and might be slightly grumpy — but that works for this show — the CMO of Planable herself, Miruna Dragomir. Damn.

Miruna Dragomir (00:37)
I bring my cranky, critical, judgmental self to today’s episode.

Mordy Oberstein (00:42)
Slightly nasally also. Not nozzle-y. Wait, I didn’t mean that the way it sounded. Like, not nozzle. Nozzle means like a jerk.

Miruna Dragomir (00:49)
With a lot of clogged nasal cavities. Congested.

Mordy Oberstein (00:53)
Yeah, well, you know who’s a nozzle? ChatGPT and their ads.

Because today we’re taking up: do you have to be crazy to advertise on ChatGPT? Short answer: yes.

But first, Say It Anyway is brought to you by SE Ranking and Planable, two sister companies helping brands grow their presence across channels. SE Ranking helps you understand your search and LLM visibility, and Planable helps you create and collaborate on the content that improves it across social, website, PR, and beyond. So you can track, distribute, and validate the impact of your data in an AI world.

I have to read it like — say “AI world” like the movie voice guy. In an AI world.

Miruna Dragomir (01:28)
It’s natural. It just comes up.

Mordy Oberstein (01:30)
That should have been my second career. In a world where AI...

Miruna Dragomir (01:34)
You know, it could have been. You do some other voices too. They’re good.

Mordy Oberstein (01:38)
Yeah, I really feel like I missed my calling in life.

Miruna Dragomir (01:41)
Plan B.

Mordy Oberstein (01:44)
No, we’ll just have an AI voice do it. Are you kidding? That’s a horrible career to go into.

Miruna Dragomir (01:48)
Yeah, that’s a very bad idea. Forget about it. Anyway, advertising.

Mordy Oberstein (01:52)
Okay, advertising. ChatGPT.

To catch you up, in case you’re living under a rock or you’ve been on vacation for the last six months or something, ChatGPT announced that ads are not just coming to ChatGPT. At this point, they are here, and they cost a $60 CPM. Not a click — an impression.

And I’m literally sitting here like, what are we doing here? Have we lost our freaking minds? Because that is quite literally Super Bowl money.

I’ll just share some stats. This is from Anu Adegbola at Search Engine Land. She was awesome. Check out her content on PPC. We had her over on Let’s Chunk, which is another content series we do each Tuesday on the SE Ranking channels. So check that out on LinkedIn, YouTube, whatever.

Okay, anyway, Anu wrote an article and says over $100 million in annualized ad revenue has been generated by less than 20% of eligible Free and Go users seeing ads daily — meaning most people are not even seeing any of these ads.

Around 85% of Free and Go users are eligible to see ads, meaning, again, according to Anu, the current revenue represents a fraction of the platform’s eventual ad capacity. She says also that 600 advertisers are now on the platform.

And also, as part of the headline, ChatGPT is still showing ads to less than 20% — again, less than 20% — of eligible users, meaning the $100 million win and self-serve access is just the beginning. But is it also the end?

And I’m like, this is crazy.

If you look at who’s advertising there now, a lot of them are the second-tier companies in that market, as I like to call them. It’s the Super Bowl for losers. It’s horrible.

It’s like sponsoring the World Series instead of the Super Bowl. And I’m a baseball fan, right? It’s like companies that said, “We really would have liked to create two drunken robots slurping vodka and run that ad for the Super Bowl, but we couldn’t afford that. So we’ll run ChatGPT ads instead.”

It’s companies like AT&T, but not Verizon. It’s companies like Target, but not Amazon.

And what is ridiculous about this is, one, you don’t have a ton of data. You have almost no data on what the user’s saying. And two, it’s a brand channel, which I’m not saying you shouldn’t spend money on a brand channel. Of course you should.

When Doritos advertises during the Super Bowl, great. But how many companies are Doritos? One. There’s only one Doritos. They’re also not a company. They’re owned by whoever owns them. I don’t know who owns Doritos. I should know that.

Miruna Dragomir (04:20)
Yeah, right? Crazy.

It’s just the ratio of that. $100 million, and then you said 600 advertisers are on the platform. Six hundred advertisers, $100 million, and just what, a couple of months? Because this was live in February.

The cost is a bit insane. And you compare it to Google Search — you can’t compare it to Google Search, because it’s a whole different thing in terms of cost. As you say, it’s Super Bowl, not Google Search.

Mordy Oberstein (04:42)
Which is crazy.

Look, I understand if you’re getting tons of visibility and you’re a huge brand and you want to run an ad for a Super Bowl — and having worked at a company where we ran Super Bowl ads, they’re not always a great idea, but sometimes they are.

Okay, you’re at that point where you’re making a decision to run a TV ad, let alone a TV ad of the highest-caliber stage, budget, whatever it is. That’s a handful of companies, and that’s not a decision where you’re like, “You know what? Let’s run a Super Bowl ad today.”

I can decide to run a PPC ad campaign today and see what happens. But this is bonkers. A $60 CPM. For impressions.

Miruna Dragomir (05:20)
And we compare it to Super Bowl, but that’s not the real cost of Super Bowl. In reality, that’s maybe the real cost of live audiences.

Then half the world will rewatch those ads on LinkedIn, on Facebook, on YouTube, on whatever, because that’s kind of the point. You make a statement: “I’m a brand who advertised on the Super Bowl, and I have super creative ads.”

But on ChatGPT, no one’s going to rewatch that ad. No one even screenshotted it. It’s just...

Mordy Oberstein (05:51)
Nothing. You’re literally just speaking into the ether.

That’s what’s crazy. You’re right. First off, I’m a football fan. The Super Bowl is usually the worst game of the year. Do you watch the NFL every week? Nope. You know the Super Bowl. Everyone knows the Super Bowl.

ChatGPT? The money is Super Bowl money, but the asset is not the Super Bowl asset, for the reason that you said and for a million other reasons at the same time.

So what are we even doing?

First off, and you mentioned this when we were prepping this episode, you’re talking about a platform. The Super Bowl has been around for over — I should know this, right? It’s like Super Bowl 60-something or whatever it is. The Roman numerals lose me. I know the Steelers won Super Bowl 40, and that was in 2006, so do the math.

You’re talking about a platform that, A, is relatively new, and B, has a lot of problems right now. And you don’t even know the audience. With the Super Bowl, you know the audience is everyone. Who’s the audience you’re talking to here? You have no clue.

Miruna Dragomir (06:49)
Yeah. And it’s like they say it’s about high intent, by the way. That seems like how they position themselves.

When people ask, “What should I buy?” that’s when we’ll advertise. But we won’t put it in the answer, and it will not in any way affect the answer. But somewhere at the end, you’ll have those.

I mean, I know it’s very comparable to Google, and we probably said that — I don’t know, I wasn’t in the working field back then — but people probably felt like, “I can’t trust Google anymore” when they started advertising.

But I feel like for ChatGPT, it might be a bit truer, because people develop this sort of weird relationship and dynamic with ChatGPT. It’s like their confidant, that they can trust, that will objectively suggest the best option to them.

Putting ads into that just feels very weird. How are they going to do it in an actually natural way that doesn’t break the confidence?

Mordy Oberstein (07:43)
Well, that’s the whole thing.

We’re talking about this, and I feel like that’s maybe the underlying theme of our whole episode here. Yeah, it’s an interesting development. It’s an interesting novelty. It’s an interesting, “Okay, where is the market going?” development.

This is not an actual strategy thing you should be worried about, thinking about, doing something about — for the most part — right now.

And one of the reasons is because no one really actually wants it. It’s like ads on Netflix or Disney or whatever. No one actually wants those ads. If I wanted those ads, I would have kept my cable package. I don’t want it.

I go to ChatGPT and AI tools because I don’t want it. I just want the answer immediately. I don’t maybe want an ad.

And you don’t even — we don’t even know that yet. So you might be dealing with an audience who, one, might not notice it; two, might not be exactly interested; three, might not be in the right ecosystem; and four, may not actually want it, because their intent for going to ChatGPT versus, I don’t know, Google, is that they don’t have any interruptions. They just get the answer.

Miruna Dragomir (08:38)
I have two things here.

First, there’s that classic lifecycle of any marketing tactic. By default, at the beginning, it’s super, super efficient, and then it goes down.

I don’t think that applies to ChatGPT advertising and ad placement, because you start by pissing people off. It’s the era of digital tools and streaming tools that figured out a business model, and then investors and the public market are going to press them into finding new revenue streams.

Anyway, that’s a different rant.

But on this point, yes, you’re going into an angry audience. That’s probably not the best idea. Wait for them to get used to it, to where it feels kind of natural.

Anyone who doesn’t work in marketing, if you were to ask them, it sounds pretty reasonable.

Mordy Oberstein (09:23)
All the time. It’s amazing to watch.

Miruna Dragomir (09:25)
But on the comparison to Google, again, it’s not Google. That’s kind of the point.

People are migrating to ChatGPT and AI because it’s not, “I’m looking for this. Give me an exhaustive list of 250 possibilities that I will then research on my own and find the best option.” No.

It is different. It is, “I’m telling you who I am, what I want, what I desire, what are the things that get me going, and you give me that one option that is best for me.”

So the level of personalization is what drives people to ChatGPT. How does the ad fit into that? How? I just don’t see it.

You want chips? You want spicy chips? I don’t know, Lay’s or whatever is the best option for you. But hey, here’s an ad for Doritos.

Mordy Oberstein (10:10)
I’ll one-up you on that, because the way I think people interact with LLMs at this point is that it’s a research phase. It’s not a conversion phase.

If you’re going to Google, you might be at a conversion phase, right? Like, “All right, I’ve decided I’m going to buy the Lay’s potato chips, and I will go to Google to buy Lay’s potato chips and have them shipped to my house because I am so lazy I cannot just walk down to the corner store and buy the Lay’s potato chips.” Welcome to America.

Sorry, I don’t want to upset all the Americans in the audience. Maybe that’s how it is in Canada too. Now we’ve upset two countries.

On LLMs, I’m just in a research phase. I’m asking, “What are the best brands?” And then I’m going to research the best brands.

So all you’ve done, in theory, is — let’s use an actual example — “Which cell phone carrier should I buy?” And AT&T is advertising with ChatGPT. So you see Verizon, you see T-Mobile, and then you see an ad for AT&T.

First off, if AT&T is a big brand and doesn’t show up in there anyway, holy crap, you have much bigger problems. And if you do show up, you just spent all that money to show up and get an ad. Good job. Amazing.

And then the user, the person, is just going to do the research now on all three of those brands they just saw. So you’ve just spent that money for a research phase.

You haven’t even emotionally impacted them like you have in the Super Bowl, where they have an affinity with you. “Oh, that was funny. That was interesting.”

It’s a freaking ChatGPT ad.

Why did ChatGPT do it this way? They should have started off with lower cost per impressions, more people in. “Okay, this thing seems to work.” I don’t know if they even know this whole thing is going to work, and they’re trying to get you to adopt it.

Miruna Dragomir (11:47)
Yeah, it’s insane. It’s insane.

Mordy Oberstein (11:50)
Holy crap, you know what it really is? It’s that they all need to figure out how to monetize this thing.

Miruna Dragomir (11:54)
One hundred percent. It has nothing to do with the audience or experience or anything like that. It’s got everything to do with the fact that it will cost the planet and then some to actually get this to a functional, sustainable business model without advertising.

Mordy Oberstein (12:10)
Here’s a conspiracy theory: you might be paying a $60 CPM so that OpenAI can say to its investors, “We’re monetizing. Invest more.”

Miruna Dragomir (12:19)
One hundred percent. It’s like you’re donating to a business. Or worse, you’re investing in a business with none of the shares.

Mordy Oberstein (12:29)
They should have released a Sam Altman GoFundMe page.

Miruna Dragomir (12:32)
That would have been a lot more honest.

Mordy Oberstein (12:34)
That would have been a lot more honest. That would be amazing.

Okay, so now that we’ve saved you all this money by not advertising on ChatGPT — which, by the way, I love that everyone on the planet and their LinkedIn profiles adopted, “I’m a ChatGPT optimizer.”

You’re working with what, like all five brands on there? That’s who you’re working with? You’re working with AT&T and Target? Amazing.

What do you do with all that money we’ve just saved you? We’re like Geico. We just saved you 50% on your marketing budget. More than 15%, probably. A million percent.

So we asked Jordan Koene, the founder of PreVisible and the voice of the Voices of Search podcast: instead of paying a $60 CPM, which can cost you millions of dollars, what should you do with that money?

So here’s Jordan.

Jordan Koene (13:20)
Hey, Mordy. So instead of paying a $60 CPM for ChatGPT ads, what should you do with that money?

I mean, here’s the reality. ChatGPT ads are as expensive as, if not more expensive than, a Super Bowl ad. And the fundamental return on investment in a product that is very much targeted to only the free-user community within ChatGPT is incredibly limiting.

So unless you have a very well-situated ICP who’s using the free version of ChatGPT, you might be missing out on a real chance at securing leads and revenue from that audience.

So I think the best way to invest in generating scale and reach with your audience, your consumers, your buyers, is to fundamentally rethink building authority.

In old SEO terms, this was link building. In new terms, this is brand mentions. This is understanding where LLMs are absorbing content to train and to build better relevance.

So this might be places like Reddit, like LinkedIn. But in reality, the fundamental truth about authority building is that the investment is both in fundamental PR, outreach, connection strategies, partner strategies that put your brand front and center, along with content that is promotable, content that is useful, and content that is relevant to the audiences you are trying to reach.

If you could translate that CPM into activities and campaigns to create more authority for your brand, you will certainly generate a much higher ROI.

Mordy Oberstein (14:56)
Thank you so much, Jordan. Check out PreVisible. Check out the Voices of Search podcast. Look for Jordan on LinkedIn.

When I got that, I was like, dude, I have to share this with the whole team. This is pitch perfect.

Miruna Dragomir (15:07)
Yeah. Literally, if you want to show up on ChatGPT so bad, you know what you should do? Invest in your brand, because that’s going to make you show up.

And a $60 CPM? That will buy you a lot of brand visibility.

Mordy Oberstein (15:20)
It’s amazing.

And I think that’s what’s so hard about this environment that we’re living in. First off, shout-out to doing good social media strategy and using Planable as part of that.

A lot of that exists on building up your positioning, building up your narrative, building up the momentum. It’s all interconnected. It’s all one ecosystem.

If I’m putting out really great content on social media, someone is going to mention that. Someone could link to that. Someone could talk about that. Someone can invite me on a podcast where I can now talk about it again to another audience, whatever it is.

Good content spreads. And social is the great connector where your brand meets your audience.

So a lot of your strategy around how you get visible in LLMs doesn’t just start with, “Let me go on Reddit just so I can get the LLM mention or the LLM citation.” It’s really going back a step, using content, using social, using everything — online, offline events — to build up the momentum of your brand so that you just have a wider digital footprint.

I feel like we’ve gone back in time. We’ve talked about this. We’ve gone back in time to where marketing is like, wait a second, I think you might need to do some marketing.

Miruna Dragomir (16:23)
Yeah, the good old “build your brand and your credibility and brand awareness” and all of those words that we were taught in college but forgot existed as soon as we heard about PPC and performance.

And you know what? I have a very good segue, actually. Here it goes.

Again, we’re so desperate to appear in these AIs, and we’re so desperate to find the hacks to do that. But if you do what you should actually be doing, which is brand visibility and awareness, then your results will span far across the AI-visibility chase that you’re in right now.

It will actually build your brand and your results for the future.

Mordy Oberstein (17:06)
I’m shaking my head, and I know we’re going to keep this short because it’s a short bite.

But when people ask me, “What do I do here?” I’m like, “You should do the thing that might work for brand visibility, but if it doesn’t work for your LLM visibility, it’s still a good tactic that brings value to your brand anyway.” That’s what you should decide to do.

But I want to say for our SEO audience — I know it’s a mixed digital audience, you may already know this — but if you’re not on the SEO side, Google recently said to The Verge, “We kind of know about all of these hacks people are doing, and we’re trying to stop them.”

And Google, if you know on the algorithmic side, in the traditional results, has gone after a lot of these kinds of hacks that we’ve seen in LLMs before and gotten rid of a ton of them. It’s coming. It’s a matter of time.

Miruna Dragomir (17:50)
That’s always going to be the case.

That’s why focusing short term and trying to hack your way into the results might help you today, and tomorrow might get you some immediate results. But at some point, you’re going to face the reality and build the things that need to be built — the solid foundation for...

Mordy Oberstein (18:11)
I am out. Because I smell another podcast episode.

I think that’s where we’re going. Maybe not the next episode, but maybe the episode after that, or the episode after that. You’ll just have to keep listening.

But as we sum up and move into what we’re clearly saying, what we’re clearly saying is companies are pretty stupid for spending a $60 CPM, even if they have a lot of money.

Send your mother a birthday card. Buy your neighbor a beer. Whatever it is, probably better money spent.

Miruna Dragomir (18:33)
Yeah, donate it to charity.

Mordy Oberstein (18:41)
I’m sure there are a couple of companies where it does make sense. I’m not saying it’s impossible, but yeah.

As we close off the episode: say it on social.

Go out there. Go on your social media platforms. Tag us, tag Planable, tag your next-door neighbor who you bought the beer for.

How would you spend that money instead of paying a $60 CPM? We’d love to hear your ideas. Tag us again on social, and let’s chat about it out there in the ether.

Which means I think we’ve said quite enough for this week.

Join us next week as we talk about not dumping your visibility, or your LLM visibility, on your SEO team, or on your social team, or on your performance team, or on any team.

Don’t dump things on other people in general. Good life hack.

Miruna Dragomir (19:23)
That’s the episode.

Mordy Oberstein (19:24)
Yeah, that’s the episode. Don’t dump on other people.

Look for a new episode of Say It Anyway on the Planable and SE Ranking websites, and wherever you choose to listen to podcasts.

Please don’t be shy about leaving us a review and a rating on iTunes and Spotify and so forth.

Look for the podcast on the Planable and SE Ranking social media channels.

And until next time, don’t keep quiet.