Nearshore Americas

Breakdown: Is OpenAI’s Chatbot CX Tool Any Good?

OpenAI launched GPTs, a tool which allows users to build chatbots of their own. The company hopes that these custom chatbots will be game changers in business, education, entertainment and daily life in general, bringing the long-running fantasies of virtual agents/assistants closer to reality.

GPTs are already being tested by businesses, who are intrigued by the tool’s potential for the deployment of custom-made, branded customer service agents. The results, at least for the moment, are mixed, but still promising.

Easy as pie: Ease and accessibility are two of the main features emphasized by OpenAI when announcing GPTs. Initial reporting underscored GPTs’ potential for SMEs and “mom-and-pop shops”, specifically.

  • “Anyone can easily build their own GPT—no coding is required. You can make them for yourself, just for your company’s internal use, or for everyone,” the company stated in its announcement of GPTs, which are available at the moment only to users of the Plus and Enterprise options of ChatGPT.

The chatbots are built through a conversation with ChatGPT. Users define the chatbot’s capabilities by providing detailed prompts and instructions. These can be fine-tuned in a “Configure” tab. 

  • There’s been a lot of tinkering with this new tool. From product developers and market specialists, to video transcribers and dungeon crawler generators, users are exploring what can be done with GPTs. There are even “prompt factories” which make the creation process even easier. 

The hype: Several small and mid-sized entrepreneurs have expressed their excitement over the prospect of developing AI-powered customer agents which are tailor made to the specifications of their businesses.

  • “We’re eyeing OpenAI’s chatbot-building feature to take our customer service to the next level,” commented Samantha Odo, COO at Toronto-based real estate agency PreCondo. “The way OpenAI works is intriguing […] Our current bot relies on set scripts and rules, which puts a bit of a cap on our flexibility and creativity.”
  • “This could revolutionize how we handle client interaction,” stated lawyer Mike Schmidt, from Dallas-based law firm Schmidt & Clark. “Imagine having a chatbot that’s not just a robotic responder but one that mirrors our brand tone and understands the intricacies of our industry.”

The doubts: Not everyone is entirely sold on GPTs. Some businesses are finding out that the tool’s simplicity might be one of its main detriments. 

  • “This will be great for commodity products, but for high end products that require a more personal touch, I don’t think the chatbot will provide much aid,” commented Stacy Elmore, Co-Founder of pergola producer and seller The Luxury Pergola. “I’ve evaluated the early versions for our particular business and it doesn’t have the same customer care or personal touch we strive for in our customer experience.”
  • “We are still fine tuning everything, but at this point we are not comfortable deploying it in a customer-facing manner,” she added. “Right now we are working to see if it would work as a resource for our customer service team for some of the more common questions”.

The experts: CX industry observers and analysts have yet to jump on the GPTs hype train. The consensus –or the closest thing to it that exists at the moment– seems to be that GPTs, like many generative AI products, require more fine-tuning before they can be successfully deployed in customer-facing situations.

  • “I am confident it will get to that acceptable finish line. The small businesses that are exploring it should seek out advice so they don’t jump in and stand up the wrong chatbot to the public,” said CX strategist and outsourcing/offshoring consultant Stuart Beame. “Perception is reality. If your chatbot messes up those first seven seconds of customer interaction, then you lost a potential sale or positive CSAT [customer satisfaction] experience.”

Emily Wright, Senior Research Analyst at SoftwareReviews Advisory, pointed out that “GPTs definitely have the potential to be a gamechanger” due to their accesibility. Nevertheless, she underscored, their success will depend on “careful planning, meticulous implementation and continual updates to adapt the tool to changing user and customer needs.”

  • “A lot of chatbots developed by companies do not live up to their potential, unintentionally relegating themselves to being glorified search functions,” she warned.

For your consideration: OpenAI Co-Founder and CEO Sam Altman was ousted from his executive position barely days after GPTs was announced and reinstated to the role less than 72 hours later. While OpenAI remains a household name in the world of AI, doubts are arising over the stability of its operations at the moment.

  • “OpenAI’s recent expulsion of Sam Altman and the subsequent threat of team losses, the future of OpenAI’s custom DIY chatbots is a bit uncertain,” stated Emily Wright in a written response previous to Mr. Altman’s reinstatement. “When looking for any tech solution it’s not just about the technology or solution itself but also the vendor you plan on partnering with. The technology and brand are strong, but OpenAI is not in a stable position internally.”

NSAM’s Take: Call centers being completely run by virtual agents remains a sci-fi fantasy. The technology will certainly improve, but it will take many years before chatbots achieve the linguistic skills and intuition that human agents have by virtue of being human. 

GPTs will probably bring a more advanced level of CX automation to SMEs, but we see these businesses using the tool to assist their customer service agents, assuming they are not outsourcing their CX operations already. 

An issue which has been seldom discussed in regards to GPTs is privacy. OpenAI assured that users will have complete control over privacy settings. However, concerns are already arising over how much access other GPTs users have to files uploaded to the data pool used by each chatbot. Google is embroiled in a privacy lawsuit due to alleged lack of transparency on how its AI-powered cloud service is being used by other businesses in customer service calls.

Sign up for our Nearshore Americas newsletter:

It’s difficult not to be carried away by the AI hype train. Business leaders are trapped between their wish to disrupt and the need to be conservative in their spending amid a challenging macro context. GPTs, however, might provide them with a low-risk option to experiment with chatbot technology before committing themselves to a higher level of automation.

Though this article focuses on CX, businesses are already using custom chatbots as “co-pilots” in other areas: from sales and recruiting to data analysis and decision-making. Such will probably be the bulk of use cases for GPTs in business: internal, to improve productivity. It will take a while before the tool can be satisfactorily deployed in customer-facing situations.

Cesar Cantu

Cesar is the Managing Editor of Nearshore Americas. He's a journalist based in Mexico City, with experience covering foreign trade policy, agribusiness and the food industry in Mexico and Latin America.

Add comment