AI NEWS
Your Customer is Trying to Tell You Something, But You Need to Read Between the Lines
We live in an era of customer-centricity. No one knows this better than customer support professionals. Even before customer service was the new marketing, support professionals had been saying for years that customer support was your secret weapon to uncovering product insight and innovation.
While the business world is shifting emphasis to the insights that can be gained from a deeper understanding of the customer feedback that lives in the service center, the resources haven’t always followed. This is primarily due to the fact that there are three main challenges in support:
Volume of data - the thousands and hundreds of thousands of support requests that are generated
Manual processes - typically a lack of integration of support tools leads to manual processes
Siloed departments and individuals - the lack of integration can lead to lack of communication between both departments and from one support individual to another
Support always seems to be playing catch up and due to the real-time requirements of supporting customers, companies don’t get the opportunity to shut down support in order to catch their breath. If ever there was a use case for machine learning, analyzing the avalanche of data that comes forth directly from customers, this is it.
Equally important is the fact that internal feedback such as support requests and external data such as product reviews are rich with emotion. Yet support teams rarely track this most human form of feedback. We tag top issues or note the level of severity but nuanced emotion goes completely unnoticed by support teams other than to note occasionally that a customer was angry.
Big Data is a game changer for customer support
Not only can we finally catch up (whew!) simply tracking not just the top 10 but the top 100 or the top 1000 issues in support, but we can also finally do a true evaluation of what is most important to our customer base.
Breakthroughs in natural language processing now allow us to bring data science and computational psychology together to expose the nuance in what the customer is telling us. Customers have much stronger feelings about certain parts of their experience than other parts. Their feedback likewise gives us insight into which parts they actually care more about. With these breakthroughs, we now have the opportunity to uncover emotional insights behind the words that a customer uses to explain their problem, or ask their question or request a refund.
Extract quantitative insights from qualitative data
We all know that it’s not only what you say but how you say it. Computational psychology goes far further than simply noting sentiment (sad, happy, mad). Breakthroughs in NLP now allow computational psychologists to identify implied feedback and truly prioritize and identify customer emotions. No longer will the customer who shouts on the phone or types in all caps be given all the attention. We also care about the customer who whispers, for these customers may have far more important things to tell us.
For more information about gaining deeper customer insights, read our Whitepaper.
[Blog by Kate Hobbie]
Stop Arguing About Roadmap Priority: Discover the Right Product Problems to Solve Using Emotion Detection
Product teams use lots of data to make decisions on launching new products and experiences, and managing them throughout their life cycle. These decisions range from tactical to strategic and require insights into customer needs whether they are consumer or enterprise. A standard process for discovering customer needs is the combination of qualitative user research and quantitative verification.
For example, a product manager may interview people to uncover pain points and hypothesize what they want. With this direction, the hypothesis can then be tested quantitatively through surveys or coding feedback in public reviews or customer service tickets. However, finding the right hypothesis relies heavily on Product Manager intuition in recognizing problems in customer feedback.
Three key problems exist in this type of discovery:
This process is highly labor intensive with an uncertainty of discovering compelling problems with enough frequency and sample size significance.
Not all feedback is useful.
Uncovering a level of intensity of pain points across all the feedback is manual and therefore subjective.
Typical Feedback on Amazon
We believe that understanding what makes people emotional is the key to uncovering the most compelling pain points that lead to amazing new products. If discovery of emotion in feedback can be paired with relevancy and quantitative discipline, product teams can uncover why people buy and create the next big thing.
Summary:
In short, we believe that product teams expend huge efforts to understand the people that buy their products. There are countless internal cycles burned negotiating and arguing over priority and impact. Often either the loudest (or highest paid) voice wins, and even the most advanced teams still rely on vague and lagging indicators to build their case. While these efforts were necessary in the past, product teams can find the most important problems if they use emotion contained within feedback. By pairing emotion with quantitative methods now available with NLP AI, product teams can now bridge the gap between past qualitative and quantitative user research practices to build better products, dramatically faster.
For more information about improving product and R&D success, read our whitepaper: The Future of Product Development here.
[by Michael Choi]
In the News: The Stitched Insights Data Science Team
In the News: The Stitched Insights Data Science Team
A Fortune magazine article in May 2018 called data science “the sexiest job” this century, and Glassdoor’s 2018 rankings put data scientist at the top of the list for Best Jobs in America. Why are data scientists so valuable these days, and why are we feeling lucky at Stitched Insights to have the best of the best on our team?
The answer, of course, lies in the importance of data to businesses today. The data being created as part of the pervasive digital ecosystem holds a ton of insights as to consumer behavior, operational processes and much, much more. But the insights are only as valuable as the algorithms and know-how of the team in place (or the engine they’ve created) to turn all that raw data into easily understandable analytics that management teams can use to make decisions.
The intersection of big data and psychology is changing the way enterprises interpret consumer data and interact with their customers. Today’s business leaders need to go beyond the obvious metrics (website traffic, convergence rates, and churn) and understand the human reasons behind those metrics. The most exciting developments in data science are helping business leaders understand the mind and how people think. As EQ is critical to successful management, so too is bringing psychology and the human element to data science in order to help enterprises understand customers and the market at large.
Two individuals driving immense innovation at this intersection of psychology and data science are Johannes Eichstaedt and Andy Schwartz of Stitched Insights. This world-renowned data science team has recently been all over the news (see links below) for its ground-breaking work with the World Well Being project. Here is a sampling of the recent articles:
Wired https://www.wired.com/story/your-facebook-posts-can-reveal-if-youre-depressed/
NBC News: https://www.nbcnews.com/health/health-news/facebook-posts-may-point-depression-study-finds-n920356
US News: https://health.usnews.com/health-care/articles/2018-10-15/facebook-posts-may-hint-at-depression
These articles discuss a study that interpreted language used in Facebook posts to help predict clinical depression, which could help with early detection. As quoted in the IFL science article: "There's a perception that using social media is not good for one's mental health," Schwartz. “But it may turn out to be an important tool for diagnosing, monitoring, and eventually treating it. Here, we've shown that it can be used with clinical records, a step toward improving mental health with social media."
While critics raise some privacy concerns in this era of increased scrutiny on the practices of the major social networks, the underlying technology (from a data science perspective) may hold the key to unlocking billions in enterprise value.
Inspired by the work of Eichstaedt and his team, Dmitriy Pavlov, Stitched Insights CEO, had an interesting thought: How can we take the ability to draw psychological insights from language and deliver this in the language of business to companies struggling with the time and effort it takes to understand consumers and their own internal data?
Pavlov saw that the underlying technology held tremendous promise for bringing unprecedented consumer and market insights to enterprises. By evolving the algorithms created by Eichstaedt and Schwartz and delivering an enterprise-ready engine to enterprises that could quickly parse external product review data and unused internal customer support data, the Stitched Insights team could help accelerate the costly R&D process for enterprises, help them create better products and even reduce warranty claims.
The result is Stitched Insights today. With the help of our famous data science team, Pavlov is helping Fortune 100 companies understand their customers in a way they never thought possible – faster and more affordably than traditional market research.
Interested in learning more? Contact heidi@stitichedinsights.com today.