3 Pillars of ChatGPT's Message on Advertising

The aversion of Silicon Valley tech companies to advertising comes from that ecosystem’s veneration of product. The underlying implication of their resistance is that advertising will corrupt the product, making it uncool, biased, or unenjoyable. Thus, like the founders of Google and Facebook before him, OpenAI’s Sam Altman initially said ads would be a “last resort.”

Inevitably, the company announced last week that it would begin testing ads in free or lower-cost tiers of ChatGPT in the US. OpenAI mainly framed the decision to test advertising as a play for accessibility: “Our mission is to ensure AGI benefits all of humanity; our pursuit of advertising is always in support of that mission and making AI more accessible,” the company wrote. It also noted that ads “can be transformative for small businesses and emerging brands trying to compete.”

These two points, consumer accessibility and small business growth, should be two of OpenAI’s three messaging pillars on advertising. They’re both very true and worth considering in the context of OpenAI’s overall mission, brand, and impact. The third pillar, though, is one they’re currently getting wrong.

OpenAI is still presenting advertising as a threat to its core product, ChatGPT, and its relationship with its users. That’s an error because advertising should be an extension of what makes ChatGPT useful, and OpenAI should gladly position itself as the future’s most effective advertising platform thanks to its underlying dynamics. In other words, ads are not at tension with generative AI; better ads are consistent with it. OpenAI should own that message instead of running away from it. 

What OpenAI got right with its advertising announcement

Product messaging ladders up to and reinforces a company’s overall mission and reputation. A brand narrative is about what’s wrong with the status quo, whom a company champions, how a company endeavors to change that status quo for the people it champions, and what makes it uniquely qualified to do so. Product messaging should nod to that bigger picture.

The announcement got this right — OpenAI explicitly linked the decision to test ads to its core mission: ensuring “AGI benefits all of humanity.” In an ideal world, OpenAI would never have described ads as an undesirable outcome for its business. On the contrary, as a16z co-founder Marc Andreessen has argued, ads are an inevitable subsidy for AI accessibility. You can’t create a product that literally billions of people can afford without advertising.

So, ads are not a boogeyman or even a necessary evil. They are a democratizing force, the business model underlying cutting-edge utilities accessible to all. Even if OpenAI initially resisted this frame, it is now pivoting to it. It would’ve been better if the company hadn’t naively dismissed ads as recently as a year ago, but positioning ads — truthfully! — as the foundation of the accessibility to which OpenAI aspires overall is a wise move.

The second pillar of this “AI for the underdog” or “AI for the masses” narrative that OpenAI should carry over from its overall brand narrative to its product is a message Meta has long espoused about the usefulness of ads for small businesses and emerging brands. Anyone in ecommerce or marketing knows that many businesses were built on the reach and efficacy of Facebook and Instagram ads; ChatGPT will now offer the same opportunity.

But from a messaging perspective, ChatGPT ads are even better than a generally effective ad vehicle because, as a new product, they will likely disproportionately benefit the nimble small businesses and startups (e.g. fledgling ecommerce brands) that are willing to test them. So, for OpenAI, the B2B message also ladders up its accessibility-forward, champion-of-the-masses positioning because it will likely be new entrants that make the most progress through ChatGPT ads. That’s a win for the San Francisco company, its advertisers, and Main Street-minded regulators.

How OpenAI should change its advertising message

While prioritizing accessibility is a wise and natural step, OpenAI is still framing ads as a threat to its product and user relationships: “People trust ChatGPT for many important and personal tasks, so as we introduce ads, it’s crucial we preserve what makes ChatGPT valuable in the first place,” the company wrote. “That means you need to trust that ChatGPT’s responses are driven by what’s objectively useful, never by advertising.”

It’s an understandable sentiment, but it still betrays an anti-advertising bias and indulges the assumption that ads represent a threat to what has made ChatGPT useful to this point. As a ChatGPT daily user and adtech professional, if I were designing OpenAI’s message on this subject, I wouldn’t concede, much less promote, this misguided tension. I’d argue that there’s never been a consumer tech product that lends itself more to advertising than ChatGPT, and the debut of ads, in addition to making the product more accessible, will simply capitalize on that natural alignment.

But, you might be thinking, don’t ads necessarily compromise the objectivity and usefulness of the answer engine? I don’t see why, unless you’re assuming an exceptionally fragile user or a paucity of advertisers (which is unlikely given the massive scale of the product’s user base, even at this relatively early juncture). There’s no reason OpenAI shouldn’t be able to use its extremely granular contextual signals to provide responses that are at the very least supplementary to, if not aligned with, its organic results.

For example, in my personal instance of ChatGPT, I have a “style” chat in which I talk to the LLM about a wide variety of topics including clothes, appliances, and real estate. I recently asked ChatGPT, based on what it knows about me and my style, to recommend cars, and it gave me a handful of extremely well-tailored suggestions factoring in aesthetics, price points, and other variables. The list was far better than what Google would’ve intuited; this granularity of alignment between questioner and respondent is what’s allowed ChatGPT to challenge an incumbent with 20 years of brand equity. 

So, my question is, why wouldn’t ChatGPT successfully be able to leverage this superior understanding of the user and his searches to surface not just organic results but adjacent ads by local car dealerships? Why should usefulness and advertising be in tension? Why aren’t great ChatGPT ads being positioned as entirely consistent with the very nature of the product?

As an adtech person, since I started using ChatGPT daily, I’ve anticipated that ads on the platform would be the best the industry has ever seen. Instagram currently holds the trophy; I regularly find Instagram ads very useful and even regard them as a largely enjoyable component of the user experience. Indeed, Altman called out Instagram as a desirable comp for ChatGPT’s nascent advertising experience. But my conversations with ChatGPT are far more granular and revealing than my usage of Instagram. Accordingly, I see no reason why ChatGPT shouldn’t make incredible ads that are, indeed, enjoyable, relevant, and useful.

The only thing holding OpenAI back from its best potential message on advertising might be its own skepticism about the subject.