Consumer Intelligence offers great opportunities to better understand your consumers, your markets, and your competitors. It also helps you to discover what you don’t know, which empowers you to spot new business opportunities and achieve true customer -centricity.
When we say “discover what you don’t know” we’re talking about unknown-unknowns. Sometimes as a marketer you may have gaps in your knowledge about a particular issue, but you’re equipped to get the information you need to fill those gaps. But with unknown-unknowns, we’re talking about gaps in your knowledge that you don’t even know exist, so you can’t find answers because you don’t know what the questions are.
You can use the AI-powered features of our Radarly consumer intelligence platform to identify these gaps.
Many of us would dream of having access to a powerful technology that effortlessly gives us a full understanding of our market, our audience, our brand, and our competitors. This dream is about to become a reality thanks to AI. Indeed, the power of AI, combined with a strong market research methodology now allows us to identify emerging trends and patterns, new weak signals that would have been impossible to detect before.
Explore Weak Signals in Your Data
In market research and consumer intelligence “weak signals” are the very first indicators of changes that could become significant in the future.
Whether your data set has been built to explore a topic, monitor what is being said about your brand, understand your target audience to optimize your content, or even monitor new trends, you should be able to detect emerging changes by exploring the data more deeply.
Let’s have a look at our key exploratory features powered by AI.
The first benefit of AI is that it enables you to explore an important social data set without looking at every single post one by one. This means you can analyze social data at scale, which would be impractical to do manually, with the AI highlighting important data points for you automatically.
Also, something that many people are unaware of, AI enables us to provide you with a high quality data stream by clearing out all of the junk and spam that can pollute social data, using powerful spam filtering algorithms and automated content tagging.
In addition to this, Radarly offers AI-powered features that can be used to discover new insights that you were not previously aware of. They will allow you to highlight interesting data points at several levels.
Emerging and Distinctive Word Clouds
We recently released this powerful feature that allows you to highlight the most specific terms within a defined period of time (emerging word clouds), or the most significant change in volume measured between a specific context (i.e. data with filters applied) and the full, unfiltered data (distinctive word clouds).
This provides a good first step to explore your data set and see if you can spot interesting weak signals to monitor, whether you focus on a specific period of time or select some filters to see what might emerge.
Content Classification
The content classification feature enables you to gain depth and accuracy in your exploration. By classifying the content into a 3-level ontology of categories that are widely recognized in the digital marketing industry, it helps you explore a topic or a sub-topic in its entirety without missing an important detail. It is a great way to identify unexpected emerging topic categories.
You can use this feature to replace a complex boolean query while creating your request, or also use it to filter your data set and explore a specific topic. With this approach, you can create a very relevant data set on a domain that, combined with our other exploration and discovery AI features, will allow you to identify weak signals and emerging trends in a vertical.
Content Landscape
Content Landscape is the exploratory tool par excellence. It offers a monthly macro view of the conversations around your brand, or whatever topic you’re monitoring. This AI-powered feature automatically categorizes the most important topics and entities, showing the associated sentiment and context.
Using this map you can easily see the relationships between a topic, a brand, a person, a company, or any other named entity. As you click into the map, you can find deeper insights on the emotional expressions used to qualify and discuss a specific entity.
To go further, you can even use the tool in a project on a corpus of data that represents your target audience, or a new community in an emerging market, to explore how they express their needs, their topics of interest, or which brands and influencers they talk about.
Topic Clusters
This feature automatically clusters conversations by topic, so that you can easily explore a group of conversations and uncover interesting insights. What is especially useful about this feature is that you can follow the volume of conversations of a specific cluster and see how it evolves over time and through different spaces on the internet.
For example, a blog post could spread on social media channels, and then get covered in online news media, which might be shared on Reddit, prompting a second wave of social media distribution. The tool also highlights sentiment distribution of the clusters.
Computer Vision, Complementing Text Analysis
All the above ones are text-analysis features. To complement this approach it’s also interesting to analyze images. The social web is a highly visual medium, with millions of images posted every day, so when hunting for insights you can’t ignore this type of content.
Computer vision is an AI-powered feature allows you to detect the following in an image:
- Logo
- Scenes
- Number of people and their ages
- People gender
- Objects and animals
- Celebrities
It allows you to identify new moments of consumption, and new behaviors associated with your brand or products, so you can understand your audience even more deeply.
Our Powerful Trio: AI + Methodology + Human Expertise
To unlock the power of these exploratory features powered by AI, it is essential to couple these software capabilities with human expertise and a robust analysis methodology.
Let’s take a concrete example: a while ago, a pharmaceutical company asked us to explore obesity and excess weight in Europe (specifically, France) to improve its knowledge on this topic in order to build a communication strategy for the launch of a new treatment to fight excess weight and obesity.
For any analysis, the best way to start is to take a macro view. In this specific case, we began by analyzing the content classification feature to understand associated topics and sub-topics to our area of interest.
Exploring the content classification, we noticed interesting data points:
- Of course, the majority of conversations around obesity and excess weight were related to health, but not all of them. We also identified other topics, like Public Safety, Social Issues & Advocacy (Fatphobia), Cooking & Recipes, Sports.
- On the health aspect, a multitude of sub-topics are covered: the link between Obesity and Mental Health, Diabetes, Medical procedures, Public Health, Nutrition, Reproductive Health, and so on.
- Digging into the sub-topics, we found interesting links between obesity and Alzheimer's disease or cardiometabolic complications.
Another way to dig into obesity is to look at the evolution over time and see if we can spot spikes. This way, you can analyze in depth the reasons for those spikes.
During this period, we can easily see two spikes of conversation around obesity and excess weight.
If we go into details in the first spike, we can see the preponderance of the hashtag #PlusDe100kgEtSereine (which, in French, means “over 100kg and calm”) that emerged on Twitter in December 2020.
Focusing on the sentiment analysis, we understood that this spike of conversations was more due to people who denigrated overweight people accepting their weight on social media rather than overweight people themselves.
Next, let’s have a look at the spike of April 2021.
If we focus on this period, and look at the frequency word cloud here is what we find:
Many topics are covered, from Covid 19 to fatphobia, weight loss, diet, etc.
Looking at the new emerging word cloud, we can see that the spike of conversation at this specific time was due to the fact that vaccination against SARS-cov-2 in France has been made available to obese people:
Using exploratory features in the right way helps to deep dive into a topic in a very detailed way and allow us to derive unknown insights. To go further you can even create a dedicated Insight Page to disseminate these valuable insights to the entire organization.
Request a Radarly demo to go deeper and gain a competitive advantage on your competitors!
If you need help to explore a topic of interest, through our expert Insight Services team, Linkfluence can complement your in-house capabilities and bring you expert assistance or provide hands-on social media research training, so you can become more autonomous and get more value out of the Linkfluence platform.