How is the Intersection of AI and Market Research Changing the Landscape of Innovation Research?

The advent of AI, particularly NLP, allows us to process vast amounts of unstructured data in record time. This enables the mining of consumers’ views, stated behaviors, and preferences from social media data in a quantitative manner. Note this is not a qualitative exploration but rather about using quantitative data, coded from conversations, alongside a solid foundation of statistics, marketing, and data science.

The impact of this cannot be understated. This is changing the face of innovation research…

Before only those with the luck or foresight who had been monitoring categories for years could understand the structure, trends, and changes that were taking place. These companies were among the few who could understand the category’s trajectory and foresee its future direction and position themselves for maximum gains.

However, the application of AI to conversation data has enabled an open view of categories. In a matter of weeks or months, one can catch up to the years of data collection efforts and have a holistic and trended view of the categories.

This has some revolutionary changes for how we do segmentations, access the strengths and weaknesses of brands and products as well as the tracking of trends and fads.

Let’s dive into what this means for just three of many areas...

1. Segmentation

Imagine creating a segmentation that is no longer based on old-world ideas such as attitudes, gender, race, and income. Instead, it is built on modern views, behaviors, and preferences, based on language and people’s underlying desires, not just their spoken words. Imagine being able to understand how each segment within your brand’s category has evolved and changed over the last 5 years. Imagine being able to identify which segments are expanding and expected to maintain that growth as well as those certainly shrinking and becoming irrelevant.

Now we can stop imagining and enter the realm of the new market research segmentation enabled by conversation data. These large datasets of millions of data points enable us to create segmentations that have robust base sizes and can be tracked over time.

This approach also allows those looking to innovate into a category to have a trended view of the category without having to have collected data for multiple years. It allows for the preferences, behaviors, and purchase drivers for each segment to be seen, the size of the opportunity determined, and the strengths and weaknesses of the current competitive landscape to be understood.

      

 

2. Strengths and weaknesses

This segmented view allows for a more personalized product and brand experience than trying to create and market a brand for the old-world clusters. For each segment or group of consumers, the strengths and weaknesses of products can be determined intrinsically and extrinsically. These strengths and weaknesses are not limited to what is known.

By using the language of the segments or categories themselves, we can surface and understand what people consider important and necessary. In other words, we can now create the world’s most useful and demand-predicting matrixes of brand-product weaknesses and strengths through a hybrid approach of top-down experts' views and bottom-up topic discovery from the consumer. No brand is safe is the new world where all the strengths and weaknesses are open to see.

Above we can see how powerful understanding the importance and level of satisfaction brands are achieving is for formulating strategy and tactics within a category. We know exactly which brands are challenging the leaders, which have stagnated, which are relegated to niches and if the leaders are under threat or entrenched.

3. Trends and Fads

Now that we understand the segments and what is important or not and how each product is performing, we need to be able to determine how this is changing over time. This is where trend and fad detection become a critical part of the mix.

Will the trend we see continue to grow? Will it break out exponentially becoming a main category driver or erode into irrelevance? Should we invest in a line extension to capitalize on the trend or not? Are these strengths that a brand holds going to continue or weaken? Are we positioned well now and for the future within our category?

All these questions can be answered when we are able to understand what a trend is and what is a fad. When we are able to calculate the sustained driving power a trend has, we can determine if it has the gas to go the distance recruiting more and more consumers or if it is simply a fad generating a lot of fanfare but never becomes the main act.

Now fads are interesting, they tend to happen when a weak trend is marketed and hyped via messaging and products. However, they leave some very tell-tell signals behind which tell us something is wrong - “this is going to collapse” much like in economics where supply exceeds demand… By using conversation data, it becomes evident what is sustainable and what is not. We can quantify the demand and the supply for the trend.

These changes and many more are coming to market research as big data, AI, and language are brought into the mainstream research efforts of companies. Converseon’s Insights and Analytics division is pioneering quantitative market research from conversation data. Contact us to hear more or arrange a demo.