The potential is there, as a few thousand AI-powered tools are already available to help humans with a variety of tasks, from transcribing videos and podcasts to providing ongoing oversight of the most important business operations.
But how to make an actual use of it? To answer that question and showcase the Large Language Models' growing potential, we combined Chat GPT functionality with the most time-consuming e-commerce processes.
The shopping experience is sometimes confusing. As the shop’s offer changes from season to season (usually by increasing in size, but some products are also being withdrawn), it's becoming increasingly more complex for the consumer to match the visuals, dimensions/sizes, price, and materials with space and budget limitations.
As a result, the customer often needs help in the maze of quite-similar-but-still-not-matching products. This is especially true for global e-stores with hundreds of millions of products ranging from small accessories to full-sized home appliances, furniture, and other more pricey items.
And the more hiccups along the shopper's journey, the less is the chance that the purchase will be finalized. Instead, the frustrated customer might vent their nerve-wracking experience by posting a relevant brand review online. Or make a rushed purchase and get ready-to-be-returned product mismatching their expectations, which only generates unnecessary transportation costs.
We've identified at least several use cases that can benefit both the seller and the buyer:
– Enhance product descriptions using the information available. A typical product page includes many details such as images, price, available configurations, delivery configuration, and a detailed description. We can use Chat GPT to suggest products based on complex criteria and influence purchase decisions by adapting its voice and writing style to match the brand's or the customer's, thus improving inclusion.
– Create promotional materials. Similarly to traditional campaigns, GPT could create social media ads for products and target specific audiences with their content. You can also batch-create products and generate blog posts recommending them.
– Improve search results. You can then feed that data a very specific question so the user can search by simply asking. We can also specify what to do if there is no match – for example, the chat could inform them that there is no exact match but recommend a similar product instead.
– Boost interaction with the website user. The most powerful application of chat GPT stems from its interactive nature, which perfectly replicates neutral conversation. It could be implemented as a smart widget, allowing users to ask questions about the product they are looking at, effectively acting as a shop assistant.
– Summarize. Chat GPT could also be fed links to multiple review websites to showcase the opinions of other clients. It can then summarize them by gathering information about a specific product and determining what they like and dislike about it. This way, users can get a quick product summary and decide more quickly.
– Monitor what user thinks about the product and brand. The database can be asked various questions, such as how our product compares to the competition's, what the most important/disliked features are, and how to solve the listed problems.