Empowering human-centric organizations using AI driven emotion analytics — The Influential Series Issue #4
For this issue of The Influential Series my esteemed guests are Lana Novikova and Hollie Gordon from Heartbeat Ai. Lana is the founder & CEO and Hollie is the Director of Business Development & Social Impact. Heartbeat Ai was founded two years ago and is an award winning software-as-a-service that reveals human emotions so you can truly understand customers, shoppers, employees, patients, etc. Heartbeat Ai transforms text input from any source into ten primary and one hundred secondary emotion categories. These emotion categories offer revealing insight into the “deep why” behind behaviour and are accessible through a user-friendly dashboard.
JP: How can organizations benefit from using AI to measure and predict emotion (emotion analytics)?
HG: Currently we have organizations and technologies that fundamentally lack emotional intelligence. Companies are dealing at a scale beyond their capacity to be emotionally intelligent with individual employees or customers. At Heartbeat Ai we are teaching emotional intelligence and the language of emotions to machines so they can be more empathetic to large groups of people. This is empowering organizations to be become more human-centric. In terms of design thinking, it starts with the user experience and how customers feel. Our technology allows organizations to prototype on a large scale. They are not limited to one-to-one persons which can be quite subjective. We now have a consistent way of interrupting a large number of responses in the thousands and tune into how they are feeling.
LN: Freeform natural language collected through focus group environments, surveys, etc. has been greatly under-utilized because there are no tools to mine the text. At Heartbeat Ai, we are now building the tools from the ground up. Machines do the low level processing and humans connect the dots at a high level. This has the potential to revolutionize fields such as market research, marketing, customer experience, HR and any field where a human is at the centre.
JP: Can you think of an example where AI driven emotion analytics had a profound social, political or health impact?
LN: I wish there were more examples right now, emotion analytics is still in its early stage. At Heartbeat Ai we conducted a profound experiment where we not only predicted who would win the US election but the percent accuracy per state. We ran an analytical model using the responses to the closed ended question, who would you vote for and the open ended question, how do you feel about each candidate. We uncovered that trust was very low with Hilary Clinton in comparison with Donald Trump and the anger towards Hilary was higher and far more intense.
To learn more about the US election case study, click here.
HG: We recently joined the Paragon Partnership led by Market Researchers from Unilever and other leading companies. We are working with them to provide insights that will help to tackle the 17-point plan of the UN Global Goals. For example, those that went from eating one meal a day to now eating three meals a day; we can analyze how they are feeling and how this has changed their life. This is a layer of understanding we can provide to go beyond the numbers and show a fuller picture of the humans that organizations are working to support.
JP: Are emotions universal or do organizations need to consider cultural or geographical differences?
LN: Absolutely, when designing an app for a rehab centre, it’s a very different cultural framework to live in. Your behaviour is different, defined by nature, nurture, cultural background, trauma in your life, etc. When we design a solution, we keep in mind our audience’s philosophy, spirituality and their culture.
We all have emotions, they are driving our behaviour and our decisions.
For example, it can be geographic or at an individual level as to whether or not to use drugs. Heartbeat Ai is built on the universality of emotions and then use cases adjust and account for those differences.
JP: The Heartbeat Ai platform uses text to analyze emotion, how accurate is text versus other forms of communication such as verbal or body language?
LN: Currently we are focusing on the language of text; it’s a way of communicating our thoughts and feelings such as poetry and literature. People may not express themselves through text but the same is true with body language, facial expressions, pulse (biometrics), etc. If I ask a question in a safe environment, you are more likely to open up and express what you are truly feeling. The setting and how you ask the question is important. People are more likely to open up if the tool they are using is designed to help them. If they respond with a short three word answer expressing anger, you can ask follow-up questions asking why they feel angry. In an open survey environment (not controlled by invitation), 40% of text is emotional that we can use to analyze. We also looked at bank app reviews on Google Play Store such as TD, 20% of the text was emotional.
HG: It has been very difficult for companies to sift through all the reviews to get a better understanding of how people feel about them. I wouldn’t express my anger in a review if I didn’t care. What is more dangerous for a company is apathy and void, the explicit lack of emotion. I will leave and go to a competitor and not say anything. An expression of anger in a review is a final cry for attention and responsiveness; if they don’t respond it will go into that void, they don’t care about me and I don’t care about them.
Humans are looking for a human relationship with organizations that they haven’t been able to get. Some organizations have done this really well and that is where they have gotten brand loyalty.
Incredibly powerful in terms of the bottom line for a business, I am more likely to refer my friends, take risks on new products, share around the world how great you are without getting anything in return, because I care about you and I am emotionally attached to you as an organization.
JP: Is glassdoor.com a reliable way for an organization to measure employee satisfaction?
HG: By that point it’s too late. The majority of negative reviews are from those that already left the organization; while you are there it’s too risky. Organizations should read reviews to determine what kind of anger and what kind of joy, but you need to ask for feedback when an employee first joins and then measure over time. If a company is consistently asking, particularly in times of change, whether in someones role or change management initiative, you can track and respond before someone gets pushed to a point where they write a negative review on Glassdoor. People write a negative review because they want management to finally listen to them because they didn’t listen while they were there.
JP: Could AI driven emotion analytics have prevented the New Coke fiasco?
LN: Yes, it could have prevented the New Coke fiasco, Uber fiasco, etc.
This happens when organizations, leaders or people are blind to the emotions of others. The others being employees, drivers, shoppers or customers.
We recently worked with a large telecommunication company in Australia. They asked their employees two opened ended question, how likely would you recommend us to your friends and how do you feel about this company. The results were categorized into positive, negative and neutral sentiment, and it showed that most people expressed anger towards the company. The data was used to set KPI’s based on employee emotions to help decrease anger, increase joy and build trust. They are including this data in a predictive modal and a year from now they will be able to predict employee turnover before it happens and avoid other risks.
JP: What financial return range should an organization expect if they adapt AI emotion analytics?
HG: If a company is open to being a human-centric organization, emotion analytics can help prevent releasing bad products and tone death advertisements like the Pepsi ad with Kendall Jenner. It’s a risk management tool to test how your audience feels before releasing at a larger scale. By understanding your customers, employees, etc. you can build authentic connections that are far more powerful than any typical transactional relationship. For example, I have an iPhone and I love this device. By adding an emotional dimension to their brand, Apple managed to cultivate loyal advocates and legions of lifelong paying customers. However, if Apple were to change the device I love thereby taking away that emotional connection, I would be quick to switch to another device.
LN: Actionable advice for a brand manager about to release a commercial, test with your target audience and the general population. If your non core audience reacts differently it could become systemic and negatively impact your brand. Make an informed and empathic decision.
JP: Is there any direct correlation between emotion and sales?
LN: I’ve been doing this for many years, I was at Nestle for 3 years doing marketing mixed modelling trying to correlate sales with advertising. It is such a complex thing to do, I’ve never seen a model that works all the time with high accuracy. As a customer, you are affected by billboards, TV ads, competition, etc. When you walk into a store, you may be in a different mood, in a rush to find what you are looking for or browsing around and picking up various items. Marketing mix modelling is very limited, they take what data is available, which is roughly 20% of the data needed to make a proper model and then look at past behaviour to try and predict future behaviour. But if something happens in the present, like the competition puts a product on sale, the marketing mix model is at a disadvantage because it doesn’t know about this situation. While at Nestle we invested a few million dollars in marketing mix modelling but we eventually dropped it because it didn’t pay out; it was roughly 50/50 accuracy and we might as well flip a coin. The best approach is to continually test and build your gut feeling. Just like a good tennis player, you know where the ball is going. You no longer need a model, it becomes more of a skill.
JP: If sales are high but the emotion expressed by customers is not very strong, should you be concerned?
LN: Heartbeat is a tool to measure explicit emotions, the ones you express. Emotions are much deeper, there are also implicit emotions that come from the subconscious. I studied a lot of neural science about how your subconscious influences your shopping behaviour. For example, Zara has an interesting way of influencing shoppers to come back to their store when they release a new clothing line. You may not be passionate about Zara but something happens in your brain that acts as a trigger. When Zara releases a new clothing line, you instinctively have a need to go to their store. They train people at a subconscious level to be brand loyal. Another example is ice cream, when you are sad you may crave it and when you eat it, it satisfies your emotions and you become happy. After eating ice cream it becomes the hero, this is why people overeat, take drugs, etc.
To learn more about Heartbeat Ai or to get in touch with Lana and Hollie, please visit http://heartbeatai.com. Also be sure to check out the Heartbeat Ai blog athttp://heartbeatai.com/blog
— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
The Influential Series is an independent publication where I explore the minds of thought leaders in the marketing and technology space.
Other interviews in this series:
My Interview with Hilton Barbour — The Influential Series Issue #1
“Meaning is the new strategy” Martina Olbertova — The Influential Series Issue #2
Originally published at https://www.linkedin.com on October 31, 2017.
Originally published at www.linkedin.com on October 31, 2017.
Are you a social media manager? We are sponsored by Falcon.io.
Falcon is the platform for every kind of social media marketer.
Falcon helps you manage all social channels and teams on one collaborative platform, create quality content with an easy-to-use editorial content calendar and social media calendar, and so much more.
Join Our Writing Contest: What is Social Media Marketing?
Write about social media marketing, earn points, and you could win an Apple Watch, Samsung Smart TV, and so much more.
Need help with your marketing plan? We have experts for that. Let’s Talk.
