Jane Ostler, EVP, Global Thought Leadership at Kantar, and Ashok Kalidas, Global Head of Data Science and Innovation at Kantar, break down what Large Language Models (LLMs) can – and can't yet – do, and what that means for market research.
LLMs have become the centre of attention in recent weeks and months with the likes of ChatGPT and Bard not just bursting onto the scene, but literally rewriting the scene.
One of the most significant technological advancements in recent years, LLMs have huge potential to revolutionise the business world, and are already being used in industries including healthcare, finance and customer service.
For the market research industry there are also many exciting and progressive use cases for the technology. From speeding up processes, enhancing others and creating new opportunities, we can expect powerful benefits to come from the technology now and in the future. But with great power comes great responsibility, and there are important caveats to also consider.
Large Language Models: in a nutshell
LLMs are AI-powered algorithms that can understand and generate human-like language. They are trained on vast amounts of text data and can generate natural language responses to various queries.
Simply put, they are designed to predict the next word or phrase in a sequence. And with lots of exposure to large datasets, these models can learn statistical relationships between words through their co-occurrences.
While it is one of the closest technologies that mirrors a human being, LLMs are far from sentient beings, and we have not yet arrived at Artificial General Intelligence – the point at which a machine will be able to understand or learn intellectual tasks as a human would.
Taking advantage of LLMs for market research
One of the significant advantages of using LLMs in market research is their ability to process vast amounts of data quickly. Traditionally, market research involves collecting data through surveys, focus groups, and interviews, which can take a lot of time and effort. LLMs can be used to analyse large volumes of data in a matter of seconds, providing valuable insights to businesses much faster than traditional methods. This can save time and resources for both market research companies and businesses.
Another advantage of using LLMs in market research is their ability to provide more accurate insights. Traditional market research methods are prone to human errors, such as biases and simple errors in data entry. LLMs, on the other hand, are less susceptible to such mistakes. They are – in theory – designed to analyse data objectively and provide unbiased insights. This can help businesses make more informed decisions based on accurate data.
LLMs can also be used to analyse unstructured data, which is often overlooked in traditional market research methods. Unstructured data includes social media posts, customer reviews and other text-based sources that are not explicitly structured for analysis. LLMs can analyse this type of data and extract valuable insights that might have been missed through traditional methods. This can provide businesses with a more comprehensive understanding of their customers and the market.
Another significant advantage of using LLMs in market research is their ability to provide personalised insights. Traditionally, market research insights are generalised and may not be applicable to all consumers. LLMs can analyse data at an individual level and provide personalised insights based on each consumer's unique preferences and behaviours. This can help businesses tailor their marketing strategies to each consumer, leading to more effective marketing campaigns and increased customer satisfaction.
LLMs can also be used to predict consumer behaviour and market trends. By analysing large volumes of data from various sources, LLMs can identify patterns and trends that might not be apparent through traditional market research methods. This can help businesses anticipate consumer behaviour and market trends, allowing them to make proactive decisions that can give them a competitive advantage.
Be aware of the risks
It’s important to note that LLMs are not a replacement for traditional market research methods. They should be used as a complementary tool to enhance traditional methods. For example, LLMs can be used to analyse large volumes of data quickly and provide valuable insights, but traditional methods and ‘humans in the loop’ are still needed to gather more in-depth information from smaller sample sizes.
LLMs also require large amounts of high-quality data to function effectively. This means businesses need to invest in data collection and management to ensure that LLMs are working with accurate and relevant data. Additionally, LLMs may not be able to analyse certain types of data, such as audio and video data, which may require other AI technologies to process.
In conclusion, large language models like ChatGPT have enormous potential to benefit the market research industry. They can provide faster, more accurate and personalised insights while also identifying patterns and trends that may not be apparent through traditional market research methods.
But market research and data organisations will need to be fully primed and prepped on the potential risks before commissioning large-scale projects, remembering to use LLMs as a complementary tool to traditional methods to enhance and scale market research projects.
In a field where qualitative and quantitative are well defined, it’s clear that humans and LLMs can achieve great things, but need to work in partnership, not in competition.
Jane Ostler, EVP, Global Thought Leadership, Kantar
and
Ashok Kalidas, Global Head of Data Science and Innovation, Kantar
.png)
.png)
.png)
.png){kind=tracking}
(1).jpeg)
.png)
.png)
