Meanwhile, ChatGPT is well established and its applications are increasingly diverse. More and more people are starting to use ChatGPT as a search engine as well. Critical writers like Ed Zitron claim this is due to the declining quality of Google search results. Last year, OpenAI (which ChatGPT falls under) already announced the prototype SearchGPT, directly competing with Google. A few weeks ago, there was news again from OpenAI land: ChatGPT is going to integrate options for Shopping, putting ChatGPT at the forefront of the customer journey. Will this be the death knell for Google? Or are the critics wrong and is this mainly an interesting additional option?
How it works
Previously, we wrote about the Integrating Generative AI into Google and its impact on e-commerce. That integration was Google's answer to OpenAI, and this new development of OpenAI seems to be Google's answer. In a demo for WIRED magazine, OpenAI showed how it works. A user asks ChatGPT for a good espresso machine costing less than $200 that makes coffee taste as good as it does in Italy. ChatGPT shows product recommendations in response, including rationale, prices and even a "buy" button. When you want to checkout, you are redirected to the provider's website to complete the purchase. This provider could be Amazon, for example. Isn't this just Google Shopping? Well, no: the results in ChatGPT are organic, not sponsored.
From conversation to conversion
The mechanism behind it is also different and less focused on keywords. Instead, it is the conversation itself that drives the results. If you tell ChatGPT that you don't want to wear synthetic fabrics and prefer cotton clothes, ChatGPT knows not to recommend polyester dresses. But ChatGPT does look for solutions in the biggest online grab bag you can imagine: from cotton HEMA clothing to threads on Reddit full of recommendations, basically everything is a possible foundation for the best possible answer. If you don't trust what a few anonymous Redditors have to say on the subject (who may also be bots, by the way), you can also point that out. Then ChatGPT searches for cotton clothes in the online gigagrabbelton, minus Reddit.
Sponsored content in ChatGPT: preferably not (but perhaps necessary)
Don't think ChatGPT will be full of ads when you soon ask your AI assistant for an acceptable surprise poem on December 4. Shopping answers will only show up for prompts where a Shopping answer actually fits. Until now. While being ad-free is one of ChatGPT's biggest advantages over Google, it is not (yet) popular enough to radically change online marketing. By 2024, Google processed 373 times as many searches as ChatGPT processed. Despite the irritation that ads sometimes provoke in consumers, it is apparently not yet enough to make a massive switch to ChatGPT.
Perhaps this also has to do with Google's familiarity: for many years it has been the default search engine for many people, and ChatGPT is seen as a less credible option because of the hallicinations that occur in AI. Moreover, it is not 100% excluded that ChatGPT will show sponsored ads in the future as well, according to the OpenAI spokesperson. ChatGPT offers paid, better versions to users and attracts many investors, yet the question remains whether this is enough to keep ChatGPT running. To continue funding the data centers and massive amounts of electricity needed for ChatGPT, piles of money are needed - such as ad revenue.
What e-commerce is going to do
Okay, so ChatGPT hasn't knocked Google off its throne. But it's definitely going to play a role in e-commerce. What are the implications of that? How do you capitalize on that as an entrepreneur? Because of the conversational element ChatGPT offers, the messy middle (the chaotic customer journey from first impression to final purchase) is going to look completely different. ChatGPT can actually clean it up: by having conversations with ChatGPT, consumers have an online personal shopper, so to speak, who only shows relevant products and gives advice. Distracting elements, such as sponsored ads or upsells disappear from the customer journey.
ChatGPT calculates what is the best product for the customer based on an unimaginable number of data points. For ChatGPT to show your product, you need to make the metadata clearly visible and well-structured. The more information you provide about a product (color, size, price etcetera), the more likely ChatGPT will "understand" your product and show it to the user. Giving as accurate information as possible is a solution that we also mentioned earlier for SEO for Google Generative AI Integration. In that respect, it does not matter much whether Google or OpenAI/ChatGPT becomes the Shopping Champion: the solution for entrepreneurs is in the same direction.
A new messy middle?
ChatGPT as a possible shopping assistant for consumers presents interesting challenges. Let's stay with fashion as an example. Since ChatGPT's recommendations are going to be based on user preferences, there won't be innovative, surprising recommendations. ChatGPT then plays it completely safe. In practice, this is more advantageous for established brands with lots of metadata and reviews. Smaller, start-up brands have difficulty getting in there. The result: a kind of monoculture, where ChatGPT only allows existing trends to come back and new discoveries lose out. If your own taste in clothes changes, ChatGPT is less useful. All recommendations are then based on your "previous" taste. AI platforms specializing in a particular niche (such as Daydream for fashion) mix personal preferences with knowledge from a particular industry.
Thus, a future may emerge where ChatGPT is a search tool for the general public, and specialized AI solutions are popular among enthusiasts and connoisseurs of an industry. You almost forget that Google also exists. This seems the perfect breeding ground for a new messy middle: the customer journey to a product is streamlined by AI, but the starting point of that customer journey? A new customer journey is emerging for that, with users deciding which platform to use for product research.
