Customer Experience Superheroes - Series 10 Episode 2 - NLP connecting us to what matters

The Customer Experience Superheroes is now in its 10th series. We've introduced you to CX professionals and practitioners from across the globe. Seeking out new thinkers and rule breakers who dare to experiment to achieve better outcomes, when others are content to rely on convention. 

In this episode we join in conversation between Eli Lepkifker, co-founder and CTO of Tagado, and Christopher Brooks, global CX specialist and MD Lexden CX Consulting.

Eli shares his story, from the origins of research at Citibank to a CEO leading a growing customer feedback company, shaped for the future, not constrained by the past.

Having seen the shortfalls and failing of standard CX measurement approaches, Eli set out to find a stronger outcome. One which focused on what was said and not the score it came with. As a researcher he wanted to be purely inspired by customers, not scores. 

Listen to the podcast to hear how Eli and Tagado have landed in the world of CX, and where the future will be taking them. 

Transcript

SPEAKER_01 0:10

Hello and welcome to another episode of Customer Experience Superheroes. My name is Christopher Brooks and I'll be your host throughout this series where we will bring you some of the excellent talent you can find in the world of customer experience, making progress and changing the way in which we both think about customer experience and the solutions available to us. And it's in the solutions space that we come to today for this episode. We're going to meet Ellie Lepkifka. Ellie is the co-founder and CTO of Taggado. Now, Taggedot is like no other company. Having had a conversation with Ellie, I've really understood the value of his uh former role working within an insight department. And that's led him to understand just what we as customer experience professionals are looking for from insight. And no surprises, it's not KPIs. It's quality insight. But that's enough from me. Let's go and join the conversation I had with Ellie and find out just what makes Tagado and the way he thinks for me the future of Voice of the Customer. We're really delighted to be catching up with Ellie here from Tagado. So Ellie, thanks for joining us on the CX Superheroes podcast and welcome.

SPEAKER_00 1:38

Thank you. We're very excited to be here and talking to you.

SPEAKER_01 1:40

As we uh we I said in the prequel, it's a really interesting topic that we're going to be talking about today. We're going to be talking about NLP, which is one of the underplayed opportunities in customer experience as far as I'm concerned. And we've taken a while to get to where we got to today, but what you guys are doing is incredible. So I'm very much looking forward to you sharing the story of how you are going to change the world of customer experience, which I'm sure you're going to. But before we get there, obviously you've got a past. Some people don't know what your past is. Could you just give us a bit of an introduction to your journey to where you are now and what you've learned along the way?

SPEAKER_00 2:17

Sure, no problem. So I've been working in the industry for about 15 years. I was working as a software developer, team lead, data science, team lead of data science. Last position I was a vice president in the NLP department in Citibank. And throughout all my career, the major thing is that I noticed that there is a distance between the RD and the actual customers. A lot of times, like even in a large corporation, we are when you are the product manager, you're speaking to some kind of stakeholder, but at the end, you're not talking to the end users. It was very career, something that was very missing for me. I want to understand how the end users are feeling about the product. And that's why I started to think about Tagadoo. That's why we raised, we founded Tagadoo about two years ago. And for me, I wanted to solve the use case of someone working in the company, a product manager, an RD department, a QA, to have a close touch into the actual customers and to understand how he is what he's doing actually affects the customer. Because sometimes this, let's say, uh distance is very hard for someone doing a job walking and doesn't understand what the actual result of his work. So that's why, for me at least, that's why the journey brought me to Fantagadu.

SPEAKER_01 3:30

Excellent. And and you know, I I did some research on the three reasons that change doesn't happen. And one of them is exactly what you say there is that I don't see how I contribute to the change we're trying to achieve. And uh and it is um so look, what we're talking, we're talking to with you, Ellie, not as someone who has an idea, but as someone who's been there, worked within an organization, seen the value, but not been able to realize the potential. And what you've decided to do is to step outside and create something which will actually fill the gap. So, first of all, how how's that been as an experience? Going from the comfort and safety of a uh an analyst desk to actually setting your own business up. Has that been an in during lockdown as well? Let's not forget. How's that experience been for you? Enjoyable?

SPEAKER_00 4:18

Amazing. It was amazing for me being only in the RD department, working on feature all the day, planning features, creating projects, getting, let's say, getting project done. I had a success with getting a project done in the Citibank, and uh to be to do everything. I'm doing everything. I'm doing sometimes the sales, I'm doing pitches, I'm uh I'm helping with all kinds of things with company with finance. So I'm doing all around with the RD, of course, because I'm the CTO. But for me, it's amazing to do all those things together.

SPEAKER_01 4:50

So you're having to learn how to run a business as well as to take your something you're very passionate about and take it out to people.

SPEAKER_00 4:57

Yeah, it's amazing because I can talk. Sorry, if I I can talk to the actual customer, get I can do the sales speech, I understand what the customer is missing, then I can go to the RD, plan the feature, and help to make sure that it's implemented. So I have all the paths, and that's what I was looking for before. I wanted to be in touch with the customers. So right now I'm in touch with the customers because I founded the company. So for me, it's amazing.

SPEAKER_01 5:20

Yeah, you read my mind. That's exactly what I was gonna ask you actually. It was in your RD function, were you able to connect with customers as you're doing now? And you've just answered it. So you've got the opportunity, and and are you finding by talking to customers that's influencing the shape of what you're developing?

SPEAKER_00 5:35

Of course, but a lot. Just for an example, let's say that we have just the we closed the Monday account like a week or two weeks ago, and even during the POC that we had with them, they had all kinds of requests and features and things that were missing. And then I went to the RD and we did it for them, and it helped us to close the deal. So for me, it's like a raising that we can go so quickly about it. You can talk to a customer, you can see what they need, you can implement it very quickly, and you can see the foot of what you're doing. So, this is something that was very missing for me, at least in large corporations such as let's say Citibank, and you're very far from the end customer.

SPEAKER_01 6:10

First of all, congratulations, that's a blue chip organization there. So exciting for you, exciting times. But what we've got now is we've got our listeners just salivating at the mouth, wanting to know what it is that you're you know you're doing. So let's give me kind of the elevator pitch in terms of what is it that you're doing and how is that adding to the world of customer experience management.

SPEAKER_00 6:33

Okay, great. So I think that in a very short line, what we're doing is giving actionable insights on a lot of the customer feedback on a large amount of customer feedback, textual data. Some companies have tens of thousands of tickets, of support tickets, sales conversation, they have a lot of textual data, and then they're trying to find what people are saying, what people complain about, what people are asking, what are the they even sometimes say customers have suggestions of what to improve in the product. And I think what we do very well is we know we use NLP technology to give a very specific insights on for each business. So we know how to adjust the technology to give actionable insights for specific business and to go over all the customer feedback automatically and give you a long list of actionable insights. I think this is the thing that we are doing the best. And on top of that, we also build a BI tool to help you navigate into the data yourself and find the insights that are more interesting to you. So if you work, let's say, in a specific department and you don't care about the insight of all the customer feedback and all the data, you want to find the insights that are more interesting to you. So we help you very quickly to navigate in the specific things that are more interesting to you and find those insights. And when I say insights, I'm talking about with several customers that are talking about the same complaint, several customers that are asking questions about the same thing, or they have a suggestion about how to improve the product. So to find those micro trends is very difficult, and to be able to give a specific team in the company the specific insights that are more interesting to them, this is something that we are, I think we're doing very well.

SPEAKER_01 8:09

Brilliant. And you can tell you're a former user in terms of landing on those two things, the ability to get at the data. The second one is and not everyone's data, just my I just want to look at my stuff. So actually, that that's a you you know, a user would only be the one who could think like that. Now, when we talk about NLP, you know, natural natural language processing, this has been around for a while, and you know, help us out here because sometimes there is a confusion between perhaps the volume of quantitative data you get back in a survey versus the contextual verbatim you're talking about. That you're having emails and complaint records and things that customers provide, of course, as well as the verbatim boxes in surveys as well. So the insight that you're bringing together is from many different sources within an organization, would that be right?

SPEAKER_00 9:04

Yeah, so we collect data from internal sources, like support assistants such as Zendesk, Intercom, and such, from social media such as Facebook, Twitter, Instagram, Reddit, and from public review websites such as the app Google Play. So we take all this data, we combine everything together, and the insights that we produce are across all the sources. So you get an insight. Let's say people are suggesting to add PayPal as an integration to the application. Let's say the specific insight. You can see the trend, and like 30 people arrested for it in the last month, and you can see that it divided between all the channels. People are talking about it in the customer support, they're talking about it in the public air review websites. And in this way, we can create an insight with a lot of people together, a lot of customers together, and see a large trend by looking at all the sources.

SPEAKER_01 9:48

It's the utopia, isn't it? If a voice is the customer bring bringing it all together, you get the ability, and obviously your integrations, but let's assume they're seamless. You're integrating within organizations, pulling all the data together, aggregating it, using NLP to find the trends and the patterns that exist in there. And then if I'm head of product innovation, or I'm head of a service, or I'm head of logistics, I can now dive into the slice that's relevant to me, I guess. That's the ability to do that. And so I've done this before, but I've done this using paper and post-its and rooms and putting everything up because I had this situation many years ago working with a gaming company, and they'd found that it was difficult to really understand using the conventional voice of customer platforms, what the text was telling them. They could get a score, but they could get nothing from the sentiment. So they could get the score, but nothing from the sentiment. And we said the only way to do it is to throw it up in a room, put it all on the room, and let's eyeball this stuff or walk around and pull it out. It was great, but I remember leaving the room and them saying, We can't do this every time there's a challenge. We just haven't got the physical space, we haven't got the resource to do this. What you're telling us is that you can now do this, you can have the quality of spending time bathing in the data, but in a very efficient way. Is that was that always the dream, or what you wanted to get to in terms of Tagado as a solution?

SPEAKER_00 11:22

Yeah, I think it was my dream. It was I think a lot before Tagado. I wanted to, but but now we are making the dream happen. And I think the major thing that makes it work right now is the development in NLP in the last few years. When you have large models that you can understand context and the understanding the actual conversation and not just count the keywords in the text, you can actually understand the conversation, and also the thing that our experience is very vast on the bringing a data-driven project to life. So we know how to do it well and we concentrate, we make that we understand very well that you have to adjust the technology specifically to a customer. It's very difficult in this area to build something that out of the box you have the same topics for all the customers. Let's say that you want to say that you create topics such as asking for advice and questions and bugs, very general topics. But in the end, the customer is looking for something very specific to his use case. For example, a company that does shipping, they want to see what are the issues with the shipping driving or the is the item is broken, very specific topics. It's not enough to do something very general. So we are we build the way the machine learning operation in a way that we can detect very specific topics for customers, which is very, very important. And also we have now the tools because of the advance in the technology in the last few years.

SPEAKER_01 12:44

Excellent. And and it'd be good to get your opinion on this, and I'd be good to understand how you display information. But for me, we've had a longstanding obsession with numbers way before customer experience. We've always been obsessed with numbers, and but with customer experience, we've inherited that, and we look at an eight or a seven and we say make that more, and you just kind of think, I can't do that, I don't know how to do that. But if I look at the information that comes with it, the text and the verbatim and the feedback from the customer, I can understand why think that we're getting a seven or eight, and I can now do something about it. So, how do you display information back to your clients? Are you using scores or are you being quite true to yourself and just how are you doing it?

SPEAKER_00 13:26

Yeah, so it'd be great to show you the demo, but I can explain it and explain it in words. So we give you out of the box a very long list, prioritized of actionable insight. And every insight is it as a title. You understand very quickly what is it talking about? For example, the first insight, user are asking to integrate PayPal. You can see the trend, you can see how many people are asking for it, and you can see what type of accounts, let's say that more enterprise accounts, are asking for this feature and they are missing it. And then the next insight, people are complaining that when they press on the button, it doesn't do exactly what it's supposed to do. So that's a bug. The 20 people are complaining about it. So you go over the insight and you understand very quickly what are the questions people are asking and what are the complaints. And if you are more interested in a specific area of the product, let's say that you work more on the mobile application, you can filter all the insights by mobile application, and then you get all the complaints, all the questions, and all the features that user are asking for, specifically for mobile applications. So you understand very quickly what is the status of the customer feedback, what are people asking for? After a few seconds of a few, let's say after a few minutes of going over the insight, that's it. You don't have to go over manual feedback yourself, you don't have to read. Obviously, we back the insight with the actual feedback. So if you want to investigate further, you can go inside the insight, you can look at the trend, you can look over on the summary, they also give you a summary of the feedbacks, and then you can also go to read the feedbacks yourself. You don't if you're not sure that you got it accurately or you want to verify it, you can go and look for read all the feedbacks that are related to the specific insight. I think that this view really gives you a very quick understanding of what the customer is saying.

SPEAKER_01 15:06

Brilliant. So, again, your positioning is very much as a solution that can inform future changes and outcomes that can be fixed. It's not a it's not a reporting capability that says, aren't we doing well or we're not doing well? You're you're allowing people to have actionable insight that they can take away and start to affect change with, which at the end of the day is there's no point unless we have that, is there, really? You want to democratize it, help others understand these are the issues, and this is how you can take them forward. Have you got any examples where you'd like to really make a difference within organizations?

SPEAKER_00 15:42

Yeah, of course. So we have one of our uh one of our one of our customers, so they use the platform for a lot of time, about two years now, and they made actual changes. They fine-tuned the roadmap on actionable side that they found in our system. They I can go into more details, but they found something about the IOX, and then they make a change in their plan, the specific roadmap, and they change the plan according to the insight that they found on our system. And I can give you another example, for example, in a recent example of a product manager that he knows exactly what are the insights that are more interesting to him because he worked only on the software department, on the API. So he looked at some of our insights, and then they made a change in the product, and they want to see if the trend of the complaints have gone down. So we have a feature in our platform where you can put an event with a date, then you can see that the specific trend has gone down and there are less complaints about it. Right.

SPEAKER_01 16:33

So it's very useful for product manager to so you can actually make a change and then track not the score, but the fact that the noise regarding the negativity has dropped away. Yeah, brilliant. So you can celebrate the success that you've actually achieved. How does that make you feel in terms of a few years ago your solution didn't exist, and now you know you're giving organizations, you're empowering them to make changes that are making a difference to their customers' lives, to their employees' lives, and ultimately to the business. So, how does that make you feel?

SPEAKER_00 17:07

I can feel amazing. I can give you another example. So, for example, also from the recent customers from a large company that you and our serving, the product manager, they look at the summary of the insight that we give them and they copy it to their presentation to prove to the management why they are developing something, why they are why they are adding improvement to the feature, why they're adding more to the feature that they are working on. So they use our system to prove to their management why they need to make changes in the product. And this is something that we're really excited about because it because they actually use it to prove the leadership why they need to make changes to the product. So this is something that was happened recently, and we're very excited about.

SPEAKER_01 17:45

Yeah, great. And that evidence is a lot of the time senior leaders they need evidence they can trust to help inform their decisions. It's not that they want to make the decisions themselves, they just need to trust the information that's coming to them. Yeah, it's really important, isn't it?

SPEAKER_00 18:03

Yeah, a lot of companies today they have competitors. Most of the companies today they don't work in a road that they don't have any competitors, they have a lot of competitors or several competitors, and then they are in a race to add more features to get to the solution. That's the race because they they need to add more customers, and a customer goes into the website, you look at the list of the features, you compare it to the competitor. So they are more and more features, but then in the end, when someone uses one of the features, they find out it doesn't work. There's a lot of issues with it. And then large companies they get into a state that they want, wait, let's stop. And now we want to improve the features that we already have, because some of them don't even work. It creates big bad reputation, and it can create bad reputation and can uh it can uh affect the customer attention. So I think a lot of companies, at least large companies, they go into a phase when they want to improve the product instead of adding more and more features, and I think at that position, our solution is a great solution because we really we can very quickly make them understand where the major issues with the product.

SPEAKER_01 19:08

Brilliant, brilliant. Now, this is a curiosity from my perspective, but I was looking earlier today at one of the big gaming companies, and I I stumbled upon a feedback survey of theirs, and then I went to a forum where the image was from, and I saw in that forum a whole lot of conversations around, oh, I wish I get this survey because I could tell them about this product feature that needs improving. I'm a good customer of theirs. Why don't I get the survey? Surely you should get rewarded. And I thought to myself, this is publicly available, really rich information telling you lots of things that could be better or customers are looking for. Can you use? And this is where my curiosity comes. Can you use your solution as kind of a competitive analytics tool where you can go and look at publicly available information and then present back to an organization, say these are the things that are bothering our competitors and they're the same as ours or different? Have you used it like that before? Could you use it like that?

SPEAKER_00 20:09

Yeah, so we had customers that used it. So analysis that we are doing today, you can just you can come to NESCAS, let's do the public analysis on some other company, only or also obviously only on public data. You can just plug into public data of some other company and do the same analysis. You can see what people are asking for, what are the feature requests in the specific topics, what are the bugs. So you can very easily show analyze the competitors and the market, and something that we did we are doing, yeah.

SPEAKER_01 20:36

Yeah, brilliant. And and how you obviously the challenge you have running your organization is you've got to think about yesterday, today, and tomorrow. But let's just hear about tomorrow. What's next then is as you get over that hump, which is the one that's always been the issue, or two issues. One was about the quality of the NLP, but also about the efficiency of working with such contextual data, you've cracked those two, tick-tick. What's next?

SPEAKER_00 21:02

I think there are several ways to go forward. I think one of them, let's say that I some someone, one of them that I think is very important, is to close the loop. So if you have you get the actionable insights, and then someone uh improves something in the product, and then you want to notify the users that it's fixed. So the actual customers feel that they are very that someone is very personally personal, they have a very personal uh relationship with the actual company. And I think this is something very important. So if you had a bug or an issue, someone made a change in the product and it fixed the issue, and then you want the customer success team to go back to those customers and tell them, look, the issue is fixed. We fixed them, we are notifying you right away. So I think that's something that we are very keen to step forward to. To have even automatically, let's say you are the people that are complaining about it. Have a button that says, Okay, I want to inform all those people that the feature was fixed, the issue was fixed, and then you have and then you can go and press a button and they will get a message the issue was. Fixed, we have been concerned about this issue, we understand your pain, and that's it. We fixed the issues. I think this is something that we are really keen to proceed with.

SPEAKER_01 22:08

And yeah, I have a massive smile on my face at the moment because we're working with a client globally, and it has been a focus for every time we talk to a customer. Let's go back and let them know about the progress that we're making on the improvement that they've requested. And okay, it may not always be the specific improvement, but it's demonstrating we're listening to you and going back. I love the idea of a tagged box that kind of when we fixed it, we can then pull out and say, You said this is what we did. I saw a very interesting piece from Gartner that showed that 95% of organizations collect the insight, but only 5% go back and tell customers what they did with it. So I think you're way ahead there of others, Ellie, which is great. The world you live in must allow you to really think about what comes next. But the thing that you bring to the table, which I think many um don't, is you've sat there as the user, as you say, in the RD department with a list of what if. So that that that's brilliant. I think if people want to get hold of you, what's the best way to speak to you, Ellie?

SPEAKER_00 23:12

Yeah, so there are several ways. First of all, they can go into a website, they can sign up for a meeting, and it's what it might be me, it might be the CEO, one of us will take the meeting. Another way is just sending me an email. I will answer every email that I get, and I will be very responsive. And also on the LinkedIn, they can just uh add me in the LinkedIn, connect to me, and I will talk to them very gladly.

SPEAKER_01 23:30

Right. And then I mentioned a couple of times about internationally how do you find it works from a different language perspective? We are you still progressing with new languages, or is this something that you've managed to take care of as well?

SPEAKER_00 23:45

So it's a very good question. Right now we are more focused on English, but if a customer has the other languages, we we can easily translate them and then give them the insight combined in English. If they have some feedback in English and some in other languages, we can very easily combine them together, translate them and then give them the insight with the other language. If they are specifically 100% a different language, it's something that we are working on, but it's not there yet there.

unknown 24:09

Yeah.

SPEAKER_01 24:09

Okay.

SPEAKER_00 24:09

So we have to adjust the NLP to a specific language.

SPEAKER_01 24:12

Sure, but that's something for tomorrow. But uh that's great, that's great. Wonderful. Well, look, thank you so much for giving up some of your time. Really excited to hear. I think we'll probably have another conversation, Ellie. So thank you so much, and I wish you and the organization all the best for the future.

SPEAKER_00 24:29

Thank you very much. It was very exciting to be here.