Personality assessments are common, but Crystal CEO Drew D’Agostino has pioneered a new approach – Artificial Intelligence – as outlined his new book “Predicting Personality.”

I’ve always been fascinated by how personality impacts business relationships, even encouraging my employees to take personality assessments such as StrengthsFinder (now called CliftonStrengths). I think a better understanding of yourself and others leads to better communication and talent development. So I was excited to speak with Drew D’Agostino, the CEO of Nashville-based technology startup Crystal, about his “Personality AI” app and new book: Predicting Personality: Using AI to Understand People and Win More Business.

Crystal uses artificial intelligence to predict anyone’s personality based on their answers to a questionnaire or existing information they created, such as biographies, resumes or social media profiles. By analyzing millions of data points within this information, Crystal is able to provide personalized, situation-specific advice. Thousands of professionals globally use Crystal to communicate more effectively, write more persuasively, and build trust faster with new people.

Crystal’s revenue grew by 100 percent in 2019, doubling the size of the company. They have raised $7 million in funding from Salesforce, Hubspot and more, and have 3,000 corporate customers. Drew and his co-founder Greg Skloot were recently named to Forbes’ 2020 list of the 30 Under 30 in enterprise technology.

How did you get into this whole world of personalities? That’s not a subject most people major in at college.

Drew: “I got into this world from my first company. My co-founder and I were 23-year-olds running our first venture-backed software company. We didn’t really know what we were doing. So our executive coaches taught us about DiSC^®, which is a common personality framework. At first I was pretty skeptical. I saw personality types as kind of horoscopes, and I didn’t really put too much stock in them.

But then I saw them play out in the workplace, helping us communicate with investors, customers, and our team. I saw how it helped me understand myself better. This was further backed up with more understanding about the scientific research behind personality models. So I became kind of addicted to that and could not unsee them in every conversation or every interaction. It just is fascinating to me how these individual tendencies could impact communication and relationships so much. It also unlocked a lot of really key things to understand about my own personal relationships and to help them improve, like with my family.

So I dove into them while we were working on that company, kind of as a user of personality models. And then when we were fired from that company, as I outline in the book, we decided to try to build a product that could do that automatically. It could tell us what other people’s personalities were like before we met them and open up that data for anybody. So that’s how Crystal emerged.”

There’s been work in this area for many years. Why do you think now’s the time for more widespread adoption of personality assessments?

Drew: “Yeah, there has been lots of work in the area of personality assessment and training and coaching. There are companies that have been around for decades doing that. The reason why now is really interesting, because there’s never been enough data to get personality types or personality profiles without assessments. So it’s always been restricted by who you could get to participate in the big assessments.

But now that everybody has a lot of data online and available, with different communication and social media channels, it opens up an entirely new area for AI to predict personality. So that’s what we built Crystal on and what we wrote the book about. It’s all about how you can now analyze data instead of just relying on assessments, and you can understand people on a whole different scale. Within the world of personality assessments, that’s a complete paradigm shift. That’s what Crystal is leading forward.”

Tell me more about the “Predicting Personality” book.

Drew: “We got the opportunity through Wiley to write our first book last year. We put together an overview of the best personality models for business and how they work. It includes a lot of practical insights about using personality data in sales, marketing, recruiting, and leadership, and also a compact overview of technology and where the technology is going. So for instance, how is AI actually helping people understand each other and win more business? This is the title of the book. We’re seeing it as kind of like the core foundation of all of our thinking. And we hope that it’s a really good guide for a lot of people into the world of personality data and personality insights for the first time.”

So you are looking at personalities of employees in your company, and also potential customers and partners, vendors, the whole gamut. What information do you need to do that?

Drew: “We just need a text sample to get the minimum of around 200 words. And that text sample can come from bios or LinkedIn profiles and resumes. The more structured, the better. And those have varying degrees of accuracy, but we just need a solid text sample.”

I’ve preached for years that while most CEOs spend 70-80% percent of their expenses on people, they don’t focus enough on people management. Years ago when equipment or something was the biggest expense, organizations would have a whole, dedicated crew focus just on its maintenance, building up tons of expertise. To me, the first step is giving people a language to talk about people. Is that what a Crystal helps you do?

Drew: “That’s a really good way of describing it. Yeah. It’s a language to describe behavior, communication styles, motivations, and tendencies. So these are things that bubble up every day anyway, either beneath the surface or explicitly. But it’s very hard to talk about it when you don’t have a standard, because everything seems very subjective. Crystal puts that standard together and then makes it available for everybody on a team, so they can all use the same language.”

Do you have a good example maybe in your own business career of where your understanding of personality helped you with a problem?

Drew D’Agostino, CEO, Crystal

Drew: “Yeah, I would say Crystal has taught me a lot about interacting with people. If we’re talking in DISC terms, I’m a very high I (Inclusive, Creative, Outgoing, Talkative). Crystal has helped me really form solid relationships with people who are on the opposite side of DISC: C (Accurate, Analytical, Structured, Purposeful). Previously, I would send emails with a lot of expression, really short and very casual. I tend to work in a very creative kind of chaotic way. That would be very frustrating for my C colleagues.

The best example is with my business partner, COO Greg Skloot, who is a very strong C. Understanding each other better frees both of us up to dig more into our strengths without expecting the same of the other. So Greg knows that I’m not going to provide everything in a very structured, organized update every week. It’s kind of an ongoing joke. We have these weekly updates. He makes them very methodically and very thoroughly, and I forget to submit mine half the time.

But then on the other hand, he knows that a lot of times my more chaotic, spontaneous style leads to product innovations that he could not stumble upon in a more structured, predictable way. Because we’re not in a fully mature market, we have to be comfortable making decisions with limited data. That’s where I excel. Because we know these things about each other’s personalities, we can basically understand the downsides and blind spots that each of us has. We do the best we can to cover up for them and let the other lean into his strengths and take advantage of that. That’s the best and most frequent example I have.”

Have you used it from a hiring perspective of “Hey we need more of this type on the team or we’re looking for a particular personality to fill a particular role?”

Drew: “Yes, but it’s with a caveat. With most roles there’s not just one personality type who can excel in the role. It’s more about understanding the expectations of how we’re going to want that person to behave. So before every hire we do a full role assessment to understand exactly what we’re looking for. Let’s say we’re hiring a new customer support representative. We try to understand our perspective on it, the people who are going to directly work with them, and anybody who’s been in that role before. Then we get a full DISC profile of the ideal person. That could be influenced by a lot of different things, not necessarily intuitively either. And we also might not agree. So we actually created a tool that helps us do that to narrow down the profile of the person for that role. And then we can look at other candidates’ personality profiles and see how stressed out or how energized they might be by getting the role.”

Now, it has to be a little interesting going to work. Do any of the potential hires worry about going to work for a company that specializes so much in personality?

Drew: “I think we attract people because of it. I mean people are endlessly fascinated by personality. We tend to attract people who are just very curious, very interested, and very relational too. So overall it’s been a very net positive for recruiting.”

How do you see the industry developing? What would you expect to see five or 10 years down the road in terms of CEOs using personality assessments?

Drew: “I think there’s an overall awareness of personality models that’s picking up. DISC has been kind of a constant for a long time, but you’re starting to see the proliferation of tools such as the Enneagram and a few other of these newer models. It’s not a new thing, but just the distribution of it is becoming more widespread. People are starting to lean on these things more and having more open conversations about personality types and understanding why people behave like they do. So I’m not sure I can really predict exactly how it’s going to shake out. But I do know that more and more people are becoming aware of these real psychological differences, and then trying to use those in their management, hiring, recruiting, and sales.

Our goal is to become the leader in the whole personality space. When you look up personality information online, we should be a number one source for that stuff. When you’re looking up information on a specific person, Crystal should be the best way to find that. People are using this for all kinds of things right now. We’re trying to create that generalized source to find anybody’s personality, and to support them with that data.”

How would you suggest a CEO get started in this whole world, who maybe hasn’t done much with personality assessments before but wants to better understand his or her team?

Drew: “Well I’m biased towards Crystal, and we’ve created this little path, so the way to do it would be first to sign up for a free account and get the team to sign up for a free account. And then you can create a group report. Before you make any decisions, you should really understand the lay of the land personality wise. You should understand where you are strong and weak, and where you are balanced and imbalanced. You can see if, for example,  you’ve been hiring people who are just like you. If so, there are going to be weaknesses, because you’re going to have a company that kind of mirrors your own strengths and flaws. It’s something you probably know at a gut level but have never really looked at quantitatively.

So that’s the first step. Look at the personality map, and see where everybody falls. The next step would be to start using personality data in communication. So prepping for every really important interaction. Just taking a second, checking what their personality profile is and saying, okay really, how should I be approaching this? Should I be approach it in a way that I prefer or is this person very different from me in how he or she thinks about something? That could be the difference between having a conflict or a misunderstanding vs. having a breakthrough or really clicking in a conversation.”

There are lots of consultants out there. Should a CEO engage one, or can the typical CEO manage this largely alone?

Drew: “Well the industry so far has been very dependent on consultants, and they’re still really an important part. But I think one of the beauties of Crystal is that it’s so simple. Personality models are not that hard to understand, at least the popular ones. You can get it from just a little bit of reading, and tools like Crystal help use that in practice in a way that has never really been possible before. You’ve needed consultants to interpret any of the personality assessments and get everyone talking about it in a certain way. So I believe consultants and coaches will continue to be the backbone of the personality assessment industry, but Crystal is kind of like a front door to it that was never there before.”

With so many personality models, why did you head down the DISC path? Why do you think it’s the best or maybe the easiest, or do you think it’s the best and easiest?

Drew: “Well we started with DISC for a few main reasons. One is DISC is the simplest of the models, with four main types. Then you can break it down into 16. But really its simplicity makes it very friendly to machine learning. You can get very accurate responses most of the time. If you’re talking about something like trying to predict accurately the 16 types every time, the numbers kind of explode and you have a lot more inaccuracy. So DISC is very helpful just in the structure of it for making predictions. It’s also open source. So you don’t need any licensing or anything to use it. It’s very old. A lot of people know about it. It’s been used and tested in companies for decades.

And then also DISC is the only other model that’s validated to the Big Five pretty closely, not perfectly. But it has a lot of backing in the Big Five personality traits. I’d call the Big Five personality traits the most scientifically valid personality framework. It’s just that Big Five traits are not useful for business directly. So yeah, those are the main reasons. But now we’ve expanded the product to have more personality assessments in it. So DISC is where we started, but we’ve gotten into Enneagram and Myers-Briggs, and some other things.”

What’s the biggest pushback you hear from people about personality assessments? I remember when we did some things at one of my companies, some people refused to take them. They didn’t want to be put in a box with their comments. Do you see any of that? Maybe that’s just a certain personality.

“There are definitely people. I mean we get thousands of assessments each day, and hundreds of accuracy feedback points each day. Very few people who are just resistant to personality assessments as a whole are using Crystal, because I think those people self-select out in the beginning. They’re probably not even signing up. But once they’re in, there are definitely people who send inaccuracies in the algorithm. We are 85 percent accurate when we provide a prediction, which means that about 15 percent of the profiles will not be accurate. And then we try to get the feedback in there so we can understand where it’s inaccurate, where it’s accurate, and then feed that back into the algorithm to make it more accurate in the future. We get lots of feedback like that, and it’s built into the product because it’s really important to improve it.

And then there are also people who have broader concerns about the nature of communicating based on personality. They wonder if it is really good to be changing the way you interact based on someone’s personality. That’s more of like a deep philosophical question, which as evidenced by us building this product, we believe the answer is yes. Because you have much better interactions when you take the time to be empathetic and understand the other person. But there are others who think that it’s more manipulative or adding some kind of artificial layer between people, which I think is a really good conversation to have. But ultimately it’s a good conversation, because thinking about what others think is actually a really good and productive habit in business.”