Unlocking UX Insights with Psycholinguistics & AI

Transcript

Userlytics x Receptiviti Spotlight Interview

Nate Brown (Userlytics):

Welcome everyone, Nate from Userlytics here. Have a really exciting show for you today. We have a one and only Jennifer Glista from the Receptiviti team here to talk a little bit about what they do over at Receptiviti. A really interesting mix of how they’re quantifying personality traits and allowing their clients to really get a better understanding of that as they build their products and solutions. without further ado, welcome Jennifer.

Jennifer Glista (Receptiviti):

Thanks so much, Nate. It’s great to be here.

Nate:

Awesome. I’m wondering if you could just give everyone a quick background of yourself, maybe, you know, roll on the team, any kind of background just so we have little understanding of who you are and where you came from.

Jennifer:

Yeah, for sure. I started with Receptiviti four years ago. When I joined the team, we were just launching a new API and I actually came with no API background, but was used to sort of building out teams in entrepreneurial way and building out sales teams with more of a consultative sales practice. My background is actually in finance, which being a human capital insights and analytics platform might not be obvious, but I spent many years on a trade floor, sort of working with all different types of people. So became very interested in who people are, how they express themselves and sort of what we can do and work with that information. And so when I had an opportunity to look at the Receptiviti platform, it was quite intriguing to me. And so I’ve been here since then and we’ve been working on lots of interesting projects.

Nate:

Very, very nice. So at Usualytics, we’re all about user insights and, you know, doing recordings and getting people’s thoughts and feelings. But the way that your team is approaching things is really why I was interested in speaking with you today. Again, being able to kind of quantify personality, all that. And I guess before we jump into too much nitty gritty, could you just give us a little bit of maybe a little origin story of Receptiviti? Where did things get started? What was the problem you guys were looking to solve?

Jennifer:

Yeah, absolutely. So Receptiviti was really formed to be the commercial arm to a science that our co-founder, Dr. Jamie Pennebaker, spent many years developing.

He was the chair of the Department of Psychology at the University of Texas at Austin. And Jamie developed this science called Linguistic and Korean Word Count, or LUC for short. And it really did a wonderful job at looking at people’s communications and really deriving different types of insights about people through their language. So Jamie put this science out to the academic community in its sort of initial form of technology. And that was great because the technology was leveraged by academics globally and sort of validated across all sorts of different use cases. So then, they’re started to become business interest in leveraging this technology to understand people that matter to companies. And so Receptiviti was formed to take it and really morph it into a forum that’s accessible more to the business community. So in its raw form, it was really leveraged by academics in psychology, sociology, computational linguistics, but not necessarily a type of technology that was accessible to the average person. And so that’s really what Receptiviti is here to do, to take this validated, wonderful science and map it to a way that the business community can make sense of and also really leverage in their own platforms and across different applications.

Nate:

Yeah, well, it’s a very, very exciting use case. I think something that I hadn’t previously seen on the market. So really, again, interesting use case. I’m wondering, have there been maybe certain industries that are more early adopters to this type of research or has it been a pretty wide breadth of types of companies that are interested in this?

Jennifer:

Yeah, absolutely. What’s interesting about the technology is that it’s widely sort of leveraged, but we do have key segments in our space, and those would be the HR domain, so human capital analytics, everything from assessment to helping with executive coaching. So that’s sort of one main segment is just understanding employees and just people at scale.

And I would say culture fits into that domain as well. The next segment would be mental health. So this is sort of a use case that’s very relevant to sort of the origins of where Jamie started in the development of the technology. And so mental health, it’s leveraged across all sorts of different types of applications. Everything from journaling all the way to sort of understanding different linguistic signals of people’s well-being. So it might be even in a coaching platform where we want to understand if various approaches are effective. Can we see that patients are improving over time? These sorts of applications. And then the one that’s probably most applicable to the Userlytics audience segment is generalized marketing. So this idea of brand and consumer understanding, even brand voice, looking at sort of brand as its own entity. So we have quite a bit of consumers leveraging the technology to work with different types of language data that they have accessible to them to better understand the people and audiences and customers that are important to their business. 

And so this is where there is potentially quite an interesting overlap between what Userlytics and Receptiviti is up to. Because your platform obviously is helping to extract a lot of that language from people. And so that is sort of raw data. That’s something that could be absolutely further analyzed to understand more about the people who are actually leveraging the platform.

Nate:

Yeah, I mean, I think you put it really well. And I guess clarifying for maybe someone who, let’s say, isn’t the Userlytics audience who is interested in kind of the service that you guys offer. Could you help us understand when you say raw data, right? So what type of information from someone getting or using a tool like Userlytics, what kind of data or raw data would they be feeding into your kind of algorithm or service?

Jennifer:

Yeah, that’s a very important point. So thank you for asking that question.

What we work with is really any type of language data, so either spoken or written communication, but we want it to be free flowing because we want to really understand the person behind the language. We’re very interested in not just what someone is talking about, so the content, but we’re also interested in the functional language. We get really excited by functional language, so prepositions, pronouns, all those little linking words are actually about 50 % of what we communicate. And those give us really interesting clues about who someone is and how they’re orientating themselves with the world. So although we might be speaking about the same subject matter, the way that we’re creating our sentences helps us understand the orientation. So is someone talking about a platform in a very self-focused way or a way that is very much mindful of the family that they’ll be working with, et cetera. And so language data is really important to us. That can be often spoken and transcribed language data. So if you had, for example, a focus group where you had a transcript of the focus group, that would be quite applicable. Even just a live interview where you have someone sort of walking through a platform to give a summary of what’s easy and what’s difficult. That is really valuable language data that can really help us understand much more about that person, their thinking style, how mentally burdened they were through that journey, and much more.

Nate:

Yeah, and again, I think so valuable, so key, because those can be difficult things to really identify if you’re a researcher, let’s say watching an interview back or kind of going through a recording that you had done. It can be difficult to kind of identify those things. So to be able to have a service or a tool that can kind of automatically do that for you is a really, again, interesting use case. I want to come back quickly to, again, the Receptiviti as a company, as a team. I know that you’ve been on the team for a couple of years. Can you kind of maybe give us an idea of how your team has changed, how the tool has changed. Maybe you guys were initially going towards one product or service and said, hey, I think this is a better fit. Yeah. Interesting to learn how you guys have adapted.

Jennifer:

Yeah, it’s a really interesting question. And you’re absolutely right. The initial early days of Receptiviti was very much focused around understanding people at scale and actually culture within organizations. So, you know, lot of the early focus was going in to help teams be able to forecast, different culture implications. You know, they might be going through a merger and acquisition and wanting to understand how that is impacting various levels of management within the organization, or if they can understand if there’s key kind of cultural centers that have big impact within an organization. And the technology is wonderfully applicable to that. But what we found is that it does so much more in so many different domains. we actually adjusted our main deployment to create an API.

And so that was a really interesting pivot for the company because it meant that it was sort of the first foray into being product agnostic and sort of industry agnostic. So, different types of companies or use cases could now start to leverage the technology for their applications. And so, through the API, companies showed up to actually build applications on top of the core science that we offer. But we also then continue to evolve to say, API meets a certain type of customer where they’re at, but we need to continue to evolve so that we can service more consultative type companies or people that maybe don’t have access to the same technology skill set to leverage the API. So we went down the path of building user interface. So that’s something that we’ve been building for a couple years now and is still in development. And also continuing to evolve our API offering. And this is sort of a newer project that we’re working on where we’re starting to think through the deployment of API in augmentation with large language models and some of the newer technology that’s now available.

Nate:

Okay. Awesome. And I guess maybe for someone in the audience that is listening and you they’re hearing you speak about API, for you, for your company as a service is API required to use your type of service, or is this something where a customer can come to you and say, Hey, I have a set of data. Can you guys just crunch the numbers and give me some kind of insight from that.

Jennifer:

Yeah, very good question. So typically companies will show up looking to leverage the API or our user interface. So if you show up with data, you can upload the data to the user interface. It will do a lot of the data preparation for you and help you visualize it. The team from time to time will engage in consulting arrangements. If there’s a big project or if a company actually is looking to put together a platform that has a repeatable use case, leveraging our sort of team’s expertise can be quite valuable. And maybe I will elaborate a little bit on the team’s expertise, if that’s okay, because it’s a really unique group of people. And that was one of the things when I joined Receptiviti, I was just taken aback by just the depth of experience and understanding.

So our team is predominantly comprised of computational linguists and psychologists. We obviously have a technology team and then more of a client-facing side of the organization. But these skill set that is readily available within Receptiviti is one that is just so unique. A lot of these people actually have worked with Jamie in academia over many years, sort of in development and sort of growth of his initial offering to the academic space and have continued to evolve to be really fundamental and key thinkers in certain domains. So one of the areas that, you know, our technology can be deployed in is to understand the linguistic similarity, for example, of how two people are communicating. So does it mean that they’re very much sort of in tune and committed to the conversation that they’re having? So we have a member of our team, for example, who really has a deep domain expertise in this unique space. And so I think it’s just wonderful when we’re able to offer that to our customers, make sure that they’re really able to leverage the extent of sort of the know-how and industry expertise that we bring to the table.

Nate:

Yeah, I love it. The team is key and I’m glad to hear that you have some solid teammates with you. One thing and kind of an interesting segue here, but wanted to talk about one of the things that’s really taken over all industries and is really a hot topic and that’s AI. Obviously AI is, some are scared coming for everyone’s jobs, things like that. But wondering how your team is leveraging AI if you are, what’s kind of your view on how AI may be changing the market that your team is in?

Jennifer:

Yeah, great questions. Firstly, in terms of how AI is changing the market, it’s been pretty positive for our team. And I’ll explain why. When I started here four years ago, the sort of consumer acceptance of this type of technology was much lower. And so with the emergence of a lot of the new technology and also more of a retail deployment so that the average consumer is getting exposure and starting to understand just how powerful and capable a lot of this new technology is. It’s really changed the sentiment towards sort of anything to do with language, understanding, and insight. So that’s been really interesting to observe in real time, to be honest. So the enthusiasm, I would say, for our type of technology has really grown.

And I’d grow on that to say that there’s now just much more of a sense of urgency within companies to take stock of the data that they do have available and try to understand how they can leverage that. So, you know, lots of companies are showing up and saying, you know, we’ve got these wonderful data assets. We want to understand sort of the people because it’s, you know, communications. Can you help us really categorize that? And quantify what’s going on in our data. And we’re absolutely able to help with that. So it’s very exciting times. In terms of how we’re starting to deploy AI. I sort of touched on it a little bit earlier, but what we do really well is measure people. So we can understand at this intersection of language and psychology who people are as expressed through their language. So this concept of projected personality, drivers of behavior, social dynamics, we can understand people’s thinking styles, obviously expressed emotion and all of the things that go along with that. But what we do well is provide the measurement to that. And with measurement comes repeatability. And that’s where some of the newest technology falls a little bit short. So in large language models, very good at sort of like this network mapping to create a really solid summarization and understanding of what you’re asking it to do, but the repeatability aspect is not quite there. So, LLMs, for example, are not good at quantifying information and then leveraging numbers to create repeatability. So that’s where it’s really a nice augmentation, where we’ve found quite a bit of value in first working with Receptiviti’s API to score language data. So for example, I might have a focus group, I might take that transcription of the interview for every individual, score it through the API to understand values. Values across all sorts of different frameworks to understand personality and drivers of behavior, for example. And then I might take that and sort of change the scoring mechanism into a way that the large language model can understand so that the large language model can help me interpret the meaning of those scores and also provide things like communication recommendations or some sort of value add insight about how I can best action those scores. So that’s where it’s been a really nice project and continuing evolution of sort of the combined forces of these two types of technology where we’re really bringing the measurement and repeatability and sort of like science and accuracy. And then we’re really able to leverage what LLMs do well in terms of like, you know, summarizing large bodies of knowledge to be very focused on sort of the insights that you’ve been able to produce to the LLM to help you accomplish what it is, whatever it is that your business is setting out to accomplish.

Nate:

Yeah, I think a really cool dichotomy of kind of AI and kind of what your team is doing. One of the things that I’ve seen both in just personally using AI, LLMs, like ChatGPT, and also how Userlytics is using AI and our analysis of sessions and things like that is it does a really good job of like summaries and kind of gives you a lot of averages, right? So to review all the sessions that here’s kind of on average what we’re seeing in the session.

And so it sounds like, correct me if I’m wrong here, how I’m kind of interpreting what you’re saying is that your team is kind of giving more of a personalized view on the different participants in a session and kind of giving you an outline on them and their personality versus just like a generalized summary of everyone together. Is that right?

Jennifer:

Yeah, absolutely. So like, maybe it’s helpful to jump into an example. You know, I might be a customer who is leveraging your platform to interview different consumers, and I might be interested in two different types of sneakers. So let’s say we’re interviewing them about Nike and Adidas. So we would probably go through and ask a series of questions about the different shoes and sort of what they like and dislike and, you know, different types of, you know, unique characteristics and I don’t know, the strings, all the different consumer product attributes. And then after that, we might say, okay, that was very interesting. We have a lot of like tactical data that we can leverage from that focus group. But now let’s try to understand what was different about segments of customers.

So at a high level, what is a difference between a Nike consumer and Adidas consumer? And so we might be able to learn that in our hypothetical example, Nike consumers were much more driven by achievement. Maybe their thinking style was a little bit different in that they had less analytical responses. Whereas maybe Adidas consumers were much more driven by affiliation or this focus on sort of the team and the broader group that they were thinking about when they were using their sneakers. And so that we would be able to produce in the form of a measurement. And then basically from that measurement, can help companies understand how to better communicate to those audiences, how to potentially segment audiences going forward so that they have like attributes so that they can sort of categorize them. Maybe they’re better able to deploy marketing spend because they’re more effectively targeting different types of audience groups. It’s all these things that a really solid understanding of the individuals and therefore like consumer groups is able to produce. 

Nate:

Yeah, and I guess maybe a follow up question to that. Customers or clients of yours that come to you, let’s say for the first time. Do they generally have like a really good understanding of who their audience is, know, their consumer market, or is that something that you help them find, or is that something that they ideally kind of have a good understanding of before coming to your site?

Jennifer:

Yeah, we have customers that show up with language data and don’t really know what’s possible or what to do with that, but know that we can be helpful. So it’s everything from that to pretty focused and specific questions.Typically speaking, when customers show up, it’s because they have a good sense that they’re missing a piece of information that’s quite important for their use case.

So, trying to understand and be able to ascertain measurement so that a lot of these decisions are not based on a hypothesis or gut instinct is really important to a lot of our customers. So they show up and, you know, want to really be able to accomplish either, audience segmentation or understanding or even, you know, better understanding their own brand voice and how the brand voice is coming across and maybe even consistency of brand voice across different channels and how they’re communicating. And that’s all stuff that we can absolutely help with and do very.

Nate:

I love to hear it. One of the reasons that I wanted to have you and your team on today was because I see a lot of potential synergies. I know a lot of the teams that we work with at Userlytics are really interested in understanding that information. Maybe don’t know how to get to it or how to get to the insights.

So let’s say for someone that is using something like Userlytics, right, and they’re kind of doing all their research and interviews, I feel like you’ve touched on it a little bit, but what are some of the ways that a tool like ours could be used in tandem with something like Receptiviti?

Jennifer:

Yeah, the Userlytics tool is great at sort of sourcing a lot of the information. Facilitating really extensive product journey investigations, know, focus groups, really trying to understand much more specifics about a product journey or a user. So in doing that, you’re collecting a lot of communications from those customers. And so you can leverage those communications to just get a different level of understanding about who they are, what they’re driven by, what their needs and values and motivations are, and even things as simple as how that actual segment went. So for example, if a customer is providing a user journey of a website, for example, they might be speaking through what they like or don’t like about that website. And by recording and transcribing their communication and running it through Receptiviti’s API, you would get a much more granular understanding of things like what we call cognitive load. So how burdened were they by that journey that they just went through? Was the website really intuitive or not?

And so, we can get a much more granular understanding about sort of their thinking style, but also, you know, in understanding sort of some of the expressed communication, things like frustration or more topical things that were driving that individual so that maybe the platform would be able to adjust its communication to better meet that person.

Nate:

I think you hit it right on the head. I think a lot of times researchers, they want to be able to facilitate the research a little bit easier, right? So for someone who are already using Receptiviti or is like, hey I’d love to use a tool like this or service like this, but you know, where am I gonna get all this language data from. That’s something where Userlytics can help source the people run the research itself pull that information. But I think again, really important is that Receptiviti kind of gives you that other, I would say next step understanding of who those participants are, right? Cause you can only do so much as you’re watching the video yourself, maybe running the transcripts through AI or things like that. But I would say your tool does a really good job of giving some really insightful things on the personality traits. So one again, I just, really appreciate you coming on, giving us a lot of good information. I think there’s probably a lot of researchers I’ve worked with and work with now who maybe weren’t even really thinking about this. So we’d love to kind of give the exposure for it.

Where can the audience and we’ll, you know, we’ll probably put up some of your contact information on the playback, but where, where can your audience connect with Receptiviti if they’re interested in using a service like yours.

Jennifer:

Yeah, for sure. We just have a general mailbox at sales@receptiviti.com. So it’s spelled R-E-C-E-P-T-I-V-I-T-I. And so if someone were to reach out there, that’s always monitored and we would certainly get back to them really quickly.

Nate:

Awesome. And I know we were chatting before this, you mentioned that your team has some kind of trial offer to maybe give people some exposure to the insights that they could get through a service like yours.

Jennifer:

Yeah, absolutely. So oftentimes when companies call us up, they won’t necessarily know exactly what they’re looking for. So we’re happy to help guide them in terms of what success looks like for them and how they can best leverage the platform and how to best leverage the platform in terms of the technology. More of a consultative sale is fine, but we also do have a free trial of the API that people are more than welcome to try out and if Userlytics customers want to come forward, we’ll certainly extend that to be a more of a broad free trial.

Nate:

Lovely, well, highly recommended from my end as well. I will say that a trial offer goes both ways. Maybe someone’s using your service already and they’re like, we’d love to get more data or maybe target a different audience that we can’t reach right now. That’s something that Userlytics can most likely help you with. And we’re also offering a free trial just to kind of get some expertise on the tool, see if it’s something that would fit well. But with that, Jennifer, again, thank you so much for coming on today.

Any kind of final thoughts or any parting wisdom to those who are maybe interested, haven’t used something like your service before, any kind of final thoughts for them?

Jennifer:

I mean, lots of thoughts. I think, you know, just reach out. We’re happy to help. I think a lot of people really underappreciate the value of their language data set. So if you have sort of the language data set that’s produced by something like Userlytics, then there’s a next level understanding that you can certainly derive from that. And we’re happy to help you navigate sort of how easy it is to get actually some of those things.

Nate:

Well, I couldn’t have said it better myself. Jennifer, thank you for coming on today.

Hope you have a great rest of your week. For those watching, keep an eye out for more of these spotlight interviews as we talk to just industry leaders and try to gain some more understanding with user experience. So Jennifer, thank you and I’m sure we’ll be chatting soon.

Jennifer: Okay, thanks a lot, Nate.