
Amazon Ads announced a partnership with ChatGPT to deliver ads. There are many reasons ChatGPT ads are going to be very successful as the numbers from OpenAI already show in terms of growth that it is hitting a $1bn run rate already. There are many naysayers arguing about the lack of sophistication of ChatGPT ads manager but they are missing the point. ChatGPT ads are not like ads on Meta or Google ads, they are much simpler to execute but very powerful as they tie into the context behind consumer conversations.
For example, someone speaking about a skin condition and asking ChatGPT about a skin cream that can help could receive an ad from L’Oreal for a product that can help. This is markedly different from a search ad where the targeting had to be keyword based or a Meta ad which had to be audience based.
What is going to be challenging for brands is to figure out how to specify “Context” for their products. For example if a user asks ChatGPT to help plan a dinner party, with a prompt like “I am planning a dinner party for some friends next week and would love to bake a casserole, what do I need to get”. An ad for a baking dish would not reach that user since the catalog does not say anything about it being suitable for cooking dinner or a casserole. It instead describes its technical specs like materials, color, size etc.
This is the same problem that exists with product discoverability on LLMs. The next generation of agentic discoverability products like DaVinci Commerce solve for exactly this problem by enriching catalogs with use case information that matches better with conversations. This could also be used to power ChatGPT ads by providing valuable product context for each product.
Traditional ad platforms like The Trade Desk and Amazon DSP were built for targeting on audiences so its not clear how effective they will be at doing this at scale. The announcements have been made but it remains to be seen how well they will work. Also advertising on ChatGPT does not require all of the complex layers that adtech has and charged for – DSPs, Ad Servers, SSPs, measurement tools, data tools all added to costs, taking away from media spend so the question brands will/should be asking themselves is whether any of that is necessary, especially if they don’t address the “Context” issue.
Will be interesting to see what role, if any, ad tech will play in the growth of ChatGPT ads and what the ad tech world will look like in a few years.