By John Burns, Ph.D.
Consumers write lots of online reviews and blog posts. Pick a product. Type in a search for information, and you’ll find more than you know what to do with. Food, beverages, toys, tools, vehicles, electronics, clothes: you name it, and consumers write about it.
These posts are often high in quality. Many consumers write soon after using the product. They’re not recalling an event from six months ago. Instead, they’re writing about what happened six hours ago. The writing, therefore, is specific and often contains strong emotions.
These consumer posts are the perfect data source for machine learning.
The value of machine learning, however, has not always matched the quality of the data. This is true for at least two reasons: the limited scope of many machine algorithms and the lack of appreciation for the human element in machine learning.
Applied Marketing Science overcomes these problems.
- We use a cutting-edge algorithm. Created by researchers at MIT, this algorithm identifies real insights. Most machine learning counts mentions of a brand or positive-and-negative comments about a product. But the promise of machine learning is insights, and this is what we deliver.
- We know humans are important to machine learning. Many companies see machine learning as a black box with little need for human input. The machine, therefore, often overlooks important insights or provides useless findings. In contrast, we teach the machine to separate useful information from the rest of the data. Then, we write a report that turns the insights into recommendations you can act upon.
The result is that our machine learning delivers hundreds of insights at a fraction of the time and cost of traditional research.
Consumer products and machine learning from Applied Marketing Science: it’s a perfect match.
Learn more about Machine Learning in our recent webinar, "Machine Learning: Uncovering Transformational Insights"