Despite Sam Altman, CEO of OpenAI believing ads would be a last resort to his business model, ChatGPT ads were inevitable. Especially when the platform needs to fund infrastructure deals. I believe OpenAI currently needs a measly $1.4 trillion. How Altman was planning to cover these costs without ads is a mystery, I’d love to see his business plan without ads in the mix.
There are some pitfalls to be aware of and OpenAI will do well to learn from the pioneers in this space such as META and Google. Clearly both companies have thrived through advertising funding, but that growth has brought its own challenges with regards to ethics, their impact on society and business demonstrated recently by being kicked out of the iAB in Sweden. I guess this serves as both a warning and an opportunity for which road OpenAI wants to go down. Unlike META and Google they have the opportunity of those that went before them, which will be further enhanced after hiring META ad veteran David Dugan recently, speaking volumes with regards to their ambitions for the significant role advertising will play in ChatGPT’s future.
Even so, the announcement of the first ads on a Large Language Model (LLM) like ChatGPT opens up a new world for advertisers. To those that have lost eyeballs due to search engine AI models pushing down sponsored results, this could bring new hope of revised consumer attention.
But it’s still risky as consumers have more LLM options to source information, so they can jump ship easily. Annoying ads simply will not be tolerated. Anthropic’s Claude is clearly gambling on this outcome by stating they will stay ad free at this year’s Super Bowl.
The pressure to get this right at OpenAI is an existential one in my view. They are already being challenged by the QuitGPT movement which has led to many leaving the app so if the advertising experience isnt managed with care its only going to make their growth ambitions even harder. That said, it’s up to OpenAI and the brands that advertise with them to make it work for the consumer in order to realise this hope for a new advertising utopia.
But how?
Ads must enhance experiences
It’s hard to imagine now but I remember the debate as to whether Facebook would lose users when they started to allow advertising. Would audiences run away towards MySpace or Friendster? I think we know how that turned out.
The danger for ChatGPT is that if the ads distract from or damage the user experience, then audiences could turn to alternative free LLMs. Facebook had the advantage as its experience was better than the competition and users didn’t want to abandon their connections and miss out on their friends’ holiday snaps.
ChatGPT might have a cleaner UI and provide better results, but for the basic prompts people won’t mind moving on as other LLM’s can provide a very similar experience.
It is therefore crucial that ChatGPT sticks to its guns and does things the right way. It must use advertising to enhance the experience and make it feel seamless within the chat. If it can achieve that, then this will be brilliant for advertisers and valuable for users.
Ads must start subtle
There is some debate about embedded ads. Should they be integrated into answers or around the text at the bottom? What would make the user experience better?
It’s clear that using ChatGPT to plan your trip to Barcelona could be paired with relevant ads for accommodation and sightseeing ideas without damaging the experience. It makes sense for food brands to appear alongside recipe inspiration and even for boiler brands to appear when your heating isn’t coming on.
But despite the temptation to gain additional ad revenue by serving ads against every chat, this isn’t Google. And it could damage the experience.
Ads appearing when you are asking ChatGPT to help with a work task or to provide a second opinion on a private matter could cheapen the experience and increase the chances that users move somewhere else.
Maybe in time when ChatGPT becomes the all-dominant LLM, it could get away with being a bit more brash. But right now, it needs to fight for attention.
Ads must consider wider industry impacts
We don’t have much to go on with regards to the plans for OpenAI’s ad platform which has been open to only the deepest pockets with at least $200k to run a test campaign in the US. We can only speculate on what targeting will be possible, whether it can be activated locally, whether we can overlay data, what reporting it can provide, whether we can connect it to our wider media plan, and all importantly, how much this will all cost.
Another consideration is the carbon impact. We’ve all been looking into eco-effects of advertising to be as transparent as we can. As an industry, we’re working towards a net-zero ambition, making positive steps to create a single source of truth via the GMSF. Will ChatGPT ads cast a cloud over these efforts?
All of these questions will be answered in due course while we acclimatise to this new advertising landscape, but marketing leaders should be thinking about it now.
Driving short and long-term benefits
In the short-term, ads on ChatGPT offer an opportunity for clients to offset lost search traffic. In the long-term, if ads are managed in the right way, they could add value to both advertiser and user, through serving helpful ads in contextually relevant, transparent, privacy safe and high attention environments. Just as ideal advertising should be.
We hope that even when advertising revenue starts to flow, OpenAI can stay consumer first rather than being driven too much by advertiser demands. Their priority should be to keep users on the site to stay ahead of the competition but if user numbers don’t grow to rate required to fund future growth and service debts the pressure to increase advertising ad revenue will be very hard to ignore in favour of user experience.
Advertising has a new frontier in ChatGPT ads, but making the wrong move could be the difference between turning the ecosystem into a blooming utopia or an ad filled wasteland.
