It’s no secret…generative AI offers immense potential in a multitude of ways in the tech world. And we’ve only begun to scratch the surface when it comes to digital commerce.
The internet has been full of buzz about the newest high-profile AI-based tool on the block, ChatGPT. If you believe the hype, it’s the latest technology publicly poised to disrupt content marketing, customer service, creative jobs, digital commerce, and even skilled labor jobs.
However, ChatGPT was not the first product, nor is it the only product, with the potential to disrupt the paid search and knowledge work industries. While ChatGPT may dominate the airwaves, assistive authoring and search technology has been around since 2020.
The future of ecommerce customer services is here but is not a replacement for human operators.
Heather Hershey – Research Director, Worldwide Digital Commerce
GPT-3, or Generative Pretrained Transformer 3 from OpenAI, is the backbone for Jasper.ai and other intelligent cloud-based content writing applications. GPT-3 was first publicly released by OpenAI on June 11, 2020. Since then, it has been widely adopted by chatbot developers, content writers, and machine learning researchers for a variety of natural language processing (NLP) tasks, such as summarization and translation.
GPT-3 is based on the concept of “generative pretraining”, which involves predicting the next token in a context of up to 2,048 tokens. This means that it can learn from a huge amount of data and produce results at scale, making it a powerful tool for businesses that rely on content marketing and sales to drive their digital commerce success.
However, GPT-3 still struggles with long forms of content and is derivative of content that is already published on the public web. With time and larger language models, this problem will likely be ameliorated, but right now, it is still in essence an elevated form of autocomplete.
When it comes to digital commerce and customer experience (CX) the big question is, “can ChatGPT provide customer service, write blog posts and personalize eCommerce shopping for customers?”
In IDC’s 2022 AI Path Survey, 224 respondents who use AI applications for digital commerce indicated that predictive analytics (67.4%), product recommendations (59.4%), and commerce website personalization (58.5%) were poised to bring the most value to their commerce operations. This is an indication of the strength in using these emerging customer experience (CX) technologies.
ChatGPT and other generative AI technologies offer a unique ecommerce experience that may allow customers to benefit from the convenience of personalized shopping and fully automated 24 x 7 customer service on-demand. By leveraging browsing history, purchase records, and other metrics related to a user’s behavior, these chatbots may soon become an integral component of personalization engines for ecommerce sites, enabling more precise product suggestions as well as improving customer service through automated support channels.
While these technologies are designed to disrupt the customer service industry, they are still in beta form, cost prohibitive for smaller companies and can produce incorrect or biased information. It’s only a matter of time before these technologies evolve far enough to improve the overall customer experience of online shopping by streamlining the search process and increasing the chance of successful conversions with automatically sales-optimized conversational commerce. The future of ecommerce customer services is here but is not a replacement for human operators.
If your company wants to leverage generative AI for personalized shopping experiences, proceed with a critical eye on longtail keyword strings shoppers use when searching for specific products. AI needs to overcome this known technological performance constraint of complex queries to be useful as a personal shopping assistant.
- Close hits may not be good enough for modern shoppers to act upon the recommendation.
- Customers demand satisfaction when shopping for specific products online and often use long strings of keywords to formulate complex queries when searching for what they desire.
- Technically speaking, overcoming the known limits of longtail search is much easier said than done. Most keyword strings are only about three to five words in length. Strings that exceed these limits can frustrate customers and push them into the weeds as they search.
- Customers are not trained on how to “talk to the bot” to get the best search results when shopping online. Therefore, the bots need to learn how to recognize various competency levels and meet the customer where they are the most comfortable. This will also ensure optimal user accessibility when these technologies are deployed for digital commerce.
Even with these limitations, the prospects for the future are quite tantalizing. GPT-3 and ChatGPT could represent powerful tools for enhancing personalized product, search, and shopping experiences. By analyzing customer data such as browsing history, purchase history, and other behavioral metrics, AI-driven technology like ChatGPT can provide more accurate recommendations and product suggestions based on an individual’s preferences and needs.
