ChatGPT and Pricing – What’s Beyond the Hype?

The emergence of ChatGPT initially caught the pricing industry off guard, sparking intense speculation about its possible uses across any number of industry sectors, including its suitability for pricing and its pricing capabilities. The mystery shrouding its inner workings have fueled debates over whether it is a mere mimicry machine or a true artificial intelligence (AI) powerhouse capable of addressing a myriad of pricing problems. The crucial question lingers – where does the real value of such technologies lie? Amidst the initial hype, it becomes imperative to shift from awe to a pragmatic analysis of GenAI (Generative Artificial Intelligence), specifically on ChatGPT and pricing, and its potential impact on the landscape.

With over a decade of pioneering experience in the pricing industry, Pricefx proudly stands as the world’s leading enterprise-grade native-cloud pricing software vendor. Leveraging innovative technologies, we have consistently tackled top-line business challenges, elevating profits and fostering growth for our clients.

Pricefx has been at the forefront of pricing AI, continuously investing as the technology grows and evolves. From machine learning and neural networks to agent-based AI and now GenAI, we continue to pursue AI innovation to drive value performance for our customers. Our focus remains on being Fast to value with our productized use cases, Flexible with our composable Bring-Your-Own Science Optimization capability, and Friendly with our clear-box AI that lets your team see exactly how optimizations are designed and executed. No other pricing optimization technology offers our kind of rapid value impact and long-term scalability and transparency.

Our unique pricing lens allows us to scrutinize ChatGPT, its competitors, and related Generative AI (GenAI) technology within the pricing landscape.

It is crucial to note that while ChatGPT and other Generative AI prompts provide valuable insights, they should only be used as a supplement to—and not a replacement for—human expertise and judgment.

 

Pricing decisions require a comprehensive understanding of the business context, market conditions, and customer needs.

 

When working as a complementary conversational user experience tool, together with human analysis and as a dedicated part of a pricing software solution, Generative AI can serve as a powerful and independent tool to enhance pricing teams’ capabilities, enabling them to make informed decisions, drive revenue growth, and maximize value for their organizations.

 

For example, the conversational user experience supported by GenAI introduced by Pricefx is an example of an extended user experience to democratize access to data and insights and improve the speed of decision-making by facilitating the creation and actionability of curated insights and recommendations. Deploying a tailored and open source Large Language Model (LLM) technology, our GenAI Actionable Insights will enable a natural language-based processing and generation to create and serve prescriptive and predictive insights that users can leverage to quickly identify opportunities and risks that have the biggest volume, revenue, and margin impact for the company. GenAI LLM capabilities are also now part of our product Knowledge Base, enabling natural language chat-based search and discovery for the entire online training and education website for Pricefx products and services, and in our Product Similarity Accelerator to identify substitute products to upsell and cross sell.

 

An exciting future awaits for GenAI in pricing – watch out for our upcoming article, “Dreaming Big: Imagining a Future for GenAI in Pricing” coming soon.

 

*Other company trademarks mentioned in this article are owned by those companies and not by Pricefx. References to them are not intended to suggest affiliation with or endorsement.

 

Nothing in this article is intended to suggest any affiliation with or sponsorship by companies other than Pricefx.

As the dust settles around ChatGPT and its peers more than a year post-launch, it is an opportune moment in time to conduct a comprehensive analysis, devoid of the initial technological euphoria, to discern the true standing of ChatGPT and GenAI systems in the context of the pricing industry.

To do that justice, let’ s spend a few moments to understand the technology in layperson’s terms.

ChatGPT & other GenAI Technology

Peeking behind the technology curtains of GenAI models like ChatGPT, Bing AI Chat, Google Bard and others beginning to sprout up everywhere, we encounter the main player – Large Language Models (LLMs). These are not your run-of-the-mill neural networks; they are powerful entities designed specifically for Natural Language Processing (NLP). Picture them as digital word maestros, orchestrating the understanding of language akin to our brains processing information. What makes them “large” are the parameters, essentially the building blocks of the neural network. In ChatGPT’s latest version, a staggering two trillion parameters drive the predictive prowess of the neural network, predicting the next word in a sentence.

This neural network concept is not new; it has been around since the early 2000s. While not as popular for traditional tasks like pricing, where other newer models excel, in NLP tasks, especially with AI chatbots like ChatGPT, these networks shine. They predict words in a step-by-step, auto-regressive manner, like how you type on your smartphone. However, the catch is they do it word by word, lacking holistic evaluation for logicality or consistency.

ChatGPT does something like the below table, except that it does not look at literal text; it looks for things that in a certain sense “match in meaning.” However, the end result is that it produces a ranked list of words that might follow, together with the possibilities.

LLMs in AI chatbots can generally process around 3000 previous words in a conversation, allowing a fluid chat experience. Dive deeper into ChatGPT, and you encounter the GPT algorithm, not just an LLM but built on the transformative Transformer architecture, giving it superstar status in NLP. This finely tuned orchestra of technology involves feature engineering and a secret sauce: reinforcement learning, where ChatGPT learns from interactions, becoming a better conversationalist over time.

While the technology is indeed marvelous, the magic is incomplete without human interaction. OpenAI constantly refines ChatGPT based on user feedback, ensuring it is both helpful and safe. While some details remain secret, one thing is certain – in the AI-powered chatbot realm, change is the only constant.

Does ChatGPT ‘Do’ Pricing?

The short answer to that question is ‘no,’ but let’s jump in a little deeper and think about why ChatGPT cannot mimic the outputs of pricing software.

As we discussed above, ChatGPT is all about words, it’s not about numbers.

Using GenAI for numbers problems is like using a dictionary to solve a math problem when you should be using a calculator.

 

GenAI is like a ‘language calculator.’

 

It’s not about crunching numbers; it is about crafting the perfect words. That’s it.

 

GenAI is NOT a replacement for pricing software, but simply another component in pricing software’s AI suite of tools. GenAI (and all other AIs) are inter-dependent and should be “composed” (meaning that the resulting AI solution is made of multiple inter-dependent specialized algorithms originating from multiple AI domains, like Lego bricks of different shapes and colors being assembled on a board).

 

To be successful at GenAI, companies should establish roadmaps today and prepare their platforms, data, and predictive and prescriptive AIs as they are the foundation of GenAI.

 

Check out the article below if you would like to learn more about the prescriptive and predictive AI models used in pricing software:

 

 

Embodied by LLMs, ChatGPT and other chatbots do not seek to replace established pricing processes or pricing software itself; rather, it serves as a supplementary tool, fostering collaboration between humans and machines in the realm of pricing. This collaboration extends beyond the confines traditionally enjoyed by predictive and prescriptive AI, reaching various sectors and stakeholders.

The standout quality of ChatGPT lies in its adaptability. It proves versatile, seamlessly transitioning between tasks such as crafting poetry, generating stories, and formulating chatbot responses. Heck, it can even possibly augment the way your salespeople work with your  CPQ (Configure Price Quote) software:

Crucially, GenAI avoids confinement to a singular niche; it is a generalist, capable of providing support across diverse domains. For businesses, this translates to leveraging its creative capabilities for content marketing, chatbots, product descriptions, and more.

In contrast, predictive AI within pricing software excels in specific tasks—determining optimal prices, identifying market trends, and adjusting strategies accordingly. While commendable in its defined role, predictive AI operates within a structured framework. The GenAI of ChatGPT on the other hand, flourishes in the realm of human language, where creativity and diversity intertwine.

Despite the GenAI and ChatGPT buzz, pricing managers may prefer maintaining control over automation levels rather than letting AI take over completely.

As GenAI and ChatGPT primarily deal with unstructured data and lack analytical and logical capabilities, they are not intended as a replacement for predictive AI in pricing but serve as a valuable complement. For instance, at Pricefx, an LLM is utilized to identify substitute products based on unstructured textual descriptions, highlighting GenAI’s effectiveness in tasks of this nature.

In scenarios where companies have limited structured data but access various sources of unstructured data challenging to import and process, GenAI or AI-powered chatbots prove instrumental in converting such data into more easily analyzable formats within a pricing system.

So, is ChatGPT a silver bullet for pricing managers and teams? No, by design, but is it a handy tool? Undoubtedly.

Is ChatGPT Ready to Take Over Pricing?

In brief, is ChatGPT on the verge of revolutionizing pricing?

Not yet, and it probably never will, but that is not what it does, nor was it designed to do.

But are we already using GenAI in our products at Pricefx?

You bet we are.

The future role and uses of GenAI in pricing remains uncertain, and it is still too early to make definitive predictions. But rest assured that Priecfx will be leading the way in its practical application as evidenced by our vision.

There are promising use cases for GenAI which Pricefx is incorporating into our platform and we will be leading the way.

Are you curious to delve deeper into the use cases of all types of AI technology in pricing software?

To explore a great overview for 5 really cool things that AI can do for your pricing, check out the handy article below:

Or if you would like to dive deeper into what GenAI is and what really makes it tick, please check out my eBook by clicking on the image below:

Meanwhile, Happy Pricing!