The CX Files #1 - Isshu Rakusai

Isshu Rakusai is the Founder and CEO of Helpfeel, an AI self-service platform which boosts CX and service efficiency for over 400 businesses. Isshu is a serial entrepreneur and founded his first company in high school, where he wrote a note-taking application that sold over 1 million units. He created a new VoC data-driven methodology for self-service optimization.

Ben also works at Helpfeel and is producing The CX Files on behalf of Helpfeel.

Transcript

The CX Files #1 - Isshu Rakusai
===

[00:00:00] 

Ben Foden: Welcome to the CX files. Today's guest is Isshu Rakusai, the CEO and founder of Helpfeel, where I also work. He is a experienced founder and actually wrote a note taking application in high school. which has sold over a million copies, and he's gone on to grow this company, originally starting in Silicon Valley and now in Japan uh, with customers around the world, including, leading businesses across the range of industries.

Ben Foden: Now Isshu what's one customer experience you've had that was truly memorable and it could be as a customer or as a service provider.

Rakusai: Okay, so I think this is very difficult questions. I can choose one, and especially talking about the kind of online customer experience, I can remember the kind of only bad experience, so I cannot choose one very nice one.

Ben Foden: I see.[00:01:00] 

Rakusai: Yeah, usually will irritate me a lot and I'm asking to our customers to the same questions actually, but, they all say how do you say?

Rakusai: The experience online digital experiences of their services is really bad. But I can remember hearing nice customer support experiences of the kind of offline. So one of our customers are karaoke, uh, companies. They are one of the largest karaoke chain, in Japan. And so they have many stores in Japan. But, uh, kind of vision, I would say, of the customer support department is to, They answer all the customers questions, for example, the customer is now asking to the karaoke shop can I bring my own iPhone or smartphone and broadcast that to the displays or can I watch YouTube on karaoke store or even can I broadcast it. to the world [00:02:00] using the karaoke shop or something like that. A lot of other now guests will ask, or can I play my PlayStations? Or can I play online soccer game in karaoke? And can I host? the game events in that chop or something like that. some other will ask, can I bring my own wine and do parties? Or is there any, can I prepare for the special room to change my clothing, et cetera, et cetera. And all the answer is actually yes. So now they're building really nice cause, uh, experiences of the karaoke shop in there and now they're trying to build that kind of friendly hosting customer support experiences to online that can assist the customer experience customers questions and answering the same questions, uh, at their websites and using our services.

Rakusai: So that's something that I'm working [00:03:00] now yeah, for them.

Ben Foden: I see. And normally when we think of great customer experiences, most of them that I can remember have been physical in the real world.

Rakusai: Yeah,

Ben Foden: what do you see as the opportunity to create really memorable and really great digital customer experiences? Mm

Rakusai: OK, so compared to the kind of physical customer support experiences that online is really bad. So the people are not expecting them a really nice one. So they are abandoned already before before they ask.

Ben Foden: hmm.

Rakusai: Changing their experiences is most important. Challenging one first. we do is most people can not expecting that the digital can answer most of the questions.

Rakusai: So what we what our approach is [00:04:00] Word hints in their websites. So and showing what type of what categories of the question can be answered and also our approach is most of the people that cannot a nice questions. So we are now building our systems to generate some questions for their users when they just enter one word. And this type of approach will, create a better experience on the digital world.

Ben Foden: Yeah, definitely. I think, you know, putting, putting questions and And suggesting them for the customer, not waiting for the customer or, or making the customer, you know, go through all this work and do all this effort, make it, make it really easy for them. Right. And, and predict what they need. We often talk about the experience of an automatic door as a [00:05:00] metaphor for design.

Ben Foden: Right.

Ben Foden: And some of the best experiences are automatic, invisible, right? The door opens without you thinking about it. Now, with regards to the digital experience, one of the big things that, is necessary to enable a really frictionless and really smooth digital experience is having the right data and having the right content available.

Rakusai: mm

Ben Foden: and. A lot of organizations are in some stage of developing their knowledge management. Maybe they're aware of knowledge centered service practice. And especially in light of, you know, the AI tools that are out there today, of course, you know, content quality and content freshness is really critical.

Ben Foden: So how do you think about that process of getting content available, having the right data around it?

Rakusai: Okay, so I think the most important KPI of the digital online customer experience [00:06:00] is now self solving rate, which, means that how many numbers of how many questions I can be had been served, means that it's not just showing the answers, but they have to read it and understand and they change their actions. That's the goal of the digital customer assistance.

Ben Foden: Mm hmm.

Rakusai: yeah, itself solving it. answering the questions. Now, quality of the search, it is getting really nice recently. So using the, an AI search, definitely, the answer, the hit rate gets very high now. As it talk about, it's getting more important how prepare the contents can yeah, cover the. The range of the questions of the users, and there's so many approaches, but the most basic approach is analyzing the queries of the [00:07:00] users only for this single issue is really hard right now. So when you prepare really nice kind of search engines, now that people get started using that more frequently, which means that we can, obtain more kind of real role users. Question data more and more.

Ben Foden: Mm hmm.

Rakusai: yeah, by analyzing it, you can now detect what type of contents is needed and which lacks now in your customer management systems. and having that type of PDCA is kind of starting an entry point so you can prepare for the better customer knowledge.

Ben Foden: Yeah, that makes sense. I think, you know, PDCA, right? Plan, do, check, act. I think a lot of teams have a sense of this. But maybe they don't have the right tools in place, or maybe those tools are not well connected or integrated. When [00:08:00] you talk to customers, and I know you've talked to a lot of customers in the space and help them with this problem.

Ben Foden: What are some of the common challenges that you see and what are some of the opportunities around this?

Rakusai: the current common challenges, most of the company do not have very nice search kind of customer engines.

Ben Foden: Mm

Rakusai: the problem. In this case, they still using category type FAQ systems or how to scenario based chat bot.

Ben Foden: hmm. Mm hmm. Mm hmm. Mm hmm.

Rakusai: the. that is prepared already by the company. And most of the cases, they cannot choose from that. They have, they have more wide, wide range of the questions. 

Rakusai: So when you see the analyzed datas, you can get the voice of the customers when you use this type of the systems, it's getting really hard for the entry point of that of the PDCA, actually, and second one is when you use [00:09:00] search. But in this case, when the quality of the search engines is very bad, then, you cannot actually improve the search engine's hit rate. So you can get what users want by analyzing the queries, but you cannot fine tune the algorithms of the search engines. 

Ben Foden: Right. I think and that fine tuning, that adjustment based on data that's coming in those search queries that people are typing. That's a capability. That's a skill. And do you find that customers need to develop that skill. Do you find that that's something that they want to hire an expert to do?

Ben Foden: How do you see the landscape around this skill of making a search finely tuned?

Rakusai: We have actually more than 300 customers or enterprise customers of our engines, but most of them do not hire special person of the tuning. Of the search. [00:10:00] So they did it manually right now. The problem is that actually even when we hire a kind of special talent, there is a limitation of the tuning by the systems. For example, the kind of legacy full text search or even the kind of modern AI search. What you can do is just. How do you say? Adding tags or, registering the kind of similar words, or, changing the title of the page a little bit or something like that. So it's very limited. For example, the bad thing of the tagging system is when you, add, just simply more and more tags, the results get worse. Well, from the some lines.

Ben Foden: Right.

Rakusai: So,

Ben Foden: Too many tags is not helpful.

Rakusai: yeah, yeah, yeah, yeah, yeah. Means that they have more bad search results.

Ben Foden: Right. 

Rakusai: The problem is there's no regular kind of methodologies of tuning the [00:11:00] search.

Ben Foden: I see.

Rakusai: Thats 

Rakusai: the 

Rakusai: problem of 

Rakusai: the 

Rakusai: current technology. 

Ben Foden: I think, everybody who's watching this video has used a search engine and they're familiar with Google and they're familiar with these major search tools. You know, some of the leading names in customer service, they've used these different search tools. And I think a lot of people watching this also have had experience trying to add the right tags, adding too many tags, not having enough, not getting the results that they want, not having visibility into that.

Ben Foden: And this is where I'd like to ask you, what is the opportunity here and what is a better tool for this? I'm asking about Helpfeel, but how is this something that's different from the other tools that are out there?

Rakusai: Okay, so helpfeel is we have the kind of technologies for answering that questions. So we have developed predictive input search systems. And in this case, our system will build kind of intentions of the users. From one content and [00:12:00] we generate, an average 50. Intention questions from one content. So that we enable the customers to tune that, hit rate, too. we're gonna also we also are using AI to generate this type of questions. And. Usually when you imagine, okay, generate the 50 questions from the content that then the management will get really harder, but we developed a kind of special script. of a language models can expand the questions from the one line. So it gets really easier to maintain the questions. Also, it's, it's compared to the tag system. Using a tag is usually hidden. So the end users will not understand or why the answer has shown to

Ben Foden: Right.

Rakusai: when I type these questions, but using our systems questions, [00:13:00] results always contain the same words that users has typed.

Rakusai: So, it's easier to find the right questions from the list.

Ben Foden: Sometimes people don't know what the right question is to ask.

Rakusai: yeah, yeah, yeah, yeah. And especially in the customer experience, that domain is limited. For example, when you are karaoke shop, people will ask only what karaoke shop can do. So it's not, it's totally different from the, like Google or ChatGPT or Perplexity. So that variety of the question is large, but the number of the answers are limited and also the knowledge base is very poor. So we have to solve this problems. So we have to help the company to more, uh, friendly and enough knowledge base and providing a better search experiences to the end users.

Ben Foden: Right, right. I'd like to talk a little [00:14:00] bit about the problem. What is the problem that this solves for businesses specifically? We're not talking about the end user here. But, you know, let's say you're you're working in a company.

Ben Foden: You know, you're, Interested in customer experience. You're trying to develop your skills. You're trying to move your team forward. You're facing various challenges, you know, what, what is the challenge that customers bring to you that, that this really solves?

Rakusai: Okay, I see. I see many customer management team in the company is focusing more on kind of physical team management more and more so they do not telling to the management of the company that they should invest more in the online digital services. That means that it's hard to show the ROI. That's the kind of hardest challenges that the kind of CX team is now experiences in the company. I think,

Ben Foden: Got it. So looking at the landscape today.

Rakusai: [00:15:00] Mm, mm,

Ben Foden: of tools out there. You know, let's say that there's a team. They're evaluating their options. There's a lot of different tools for AI. You talk about AI chat bots, of course.

Rakusai: mm,

Ben Foden: is a really big one. There's also AI search and all these different options available. And I think a lot of people are kind of concerned about the risks. They're excited about the opportunity, but there's some kind of gray area or

Rakusai: mm,

Ben Foden: And what are the things that you're seeing that they really care about and that really make a difference for them?

Rakusai: mm, mm. so as you know, kind of chat GPT or some now AI model itself. It's something like, it's really big innovation but you can't use it as it is. So you have to wrap it carefully when you want to adapt to your services. [00:16:00] When you use as a consumer, when you use chat GPT as a consumer, it's really fun tool or very kind of innovative and convenience tools, but you can't use it directly to represent your company customer support to end users. So we take that approach using the chat GPT or these kind of AI models most conservatively. And that means reducing the risk of failure of the AI, as small as possible. 

Ben Foden: Right. I think the, you know, we talk about hallucination risk, right? Of AI, you know, kind of imagining an answer that doesn't exist.

Rakusai: Mm-Hmm.

Ben Foden: we've talked already today about, you know, making the right data and content available as an input for AI to generate quality answers. Right.

Rakusai: Mm-Hmm.

Ben Foden: One of the big things that people think about with AI that I've been hearing a lot, I think it's really common is the idea of using AI to get efficiency, right?

Ben Foden: To save time to save money. And if you're an organization that's really prioritizing the customer [00:17:00] experience, efficiency is important, of course, but it maybe is not everything. There's more to the customer experience than just making it fast. What are some opportunities that you see for using AI to create a better customer experience, a more satisfying experience?

Rakusai: Mm-Hmm. . So current customer experiences of the digital world is really bad as we agreed already. So our approach is now u using the. For of course, the first create to build the new answers and registering it to the knowledge base. First. Also, we are now trying to use that AI's answers directly to end users to but our approach is now trying to prepare answers

Rakusai: before users has asked or we have list of the users questions already in our database, so we can prepare all kind of answers, not [00:18:00] in real time, but before they ask. And check all the answers before showing to the end users and this will reduce the hallucination and keeping the efficiency of the AIs.

Ben Foden: I see.

Ben Foden: Definitely. I think if you're expecting bad customer service, you're expecting a bad experience. That's the default online. Most help centers, even modern chatbots, the normal feeling is a negative expectation. You think it's going to be bad, right?

Ben Foden: And then if you have a tool that predicts what's Your problem suggests a question for you. That question is thoughtfully created. It's, it's matching to your intent then that creates a good experience, right? Suddenly, you know, what you expected to be negative is now positive. I think this is an opportunity, whatever tool that you're using, if you're predicting intent and you're matching what people need preemptively or before they can put it [00:19:00] into words themselves that's going to create a better digital experience. Um, when you look at Helpfeel itself, when you look internally now, of course we have, for those of you listening in, we have a customer success team. We have various elements of our service department. Of course, we have a sales team. And when you look at how Helpfeel delivers a customer experience, what are some of the things that you think go across departments? So, I think normally in the traditional model of the business, you have the support department and the sales department. You have marketing, and they're very separated. And they have their own metrics, and they have their own priorities.

Ben Foden: But when we look at CX today, it really crosses over the whole customer journey. Right. There's pre sales. There's the body of the experience. And then there's there's post sales. There's support, you know, sports normally right at the end of that. how do you look at building a comprehensive customer experience and kind of building that across departments?

Rakusai: Yeah, that's very [00:20:00] interesting questions. think the, how to say, how to kind of, Uh, hundred. How to positioning the customer support department is now changing. I think I can remind of the CEO is now changing a little bit. A little bit. that's that's the kind of trend I'm seeing now. Traditionally, as you know, the customer support team has position as a cost center in profit and loss PL inside the company. But, they can start using like the tools of the Helpfeel or kind of modern customer support tools now they can get the voice of the customers. and when they build their kind of better experiences of the website, then, of course, the conversion of the websites gets better, or now start hearing the voice of the silent customers or the cool customers, before the transactions or before buying [00:21:00] the, the items or products of the company and the customer support centers can now show the report of analyzing the VOCs. So it, it's the customer support center is let's integrated more and more getting close to the marketing team of the company.

Ben Foden: Definitely. And I think there's a natural opportunity there where, you know, I think traditionally that the different departments have used different terminology or different words, but it's really been a lot of the same themes, right? You might have, the sales team, obviously listening to customers, talking to them every day.

Ben Foden: Understanding what they want and what they need, right? You have the marketing department who's often is looking for testimonials or looking for quotes from customers, to understand where marketing should be focused. And of course in support department, you have VOC, right? And really all of these are examples of VOC.

Ben Foden: They were all in their own different [00:22:00] silos. And now I think VOC is becoming this more central thing and support is really an opportunity because they really have that. That is their, their domain. That's their area of expertise.

Rakusai: Mm, Mm, Yeah, correct. Yeah. Now, now, using using that kind of database like CDP customer data platform, you can integrate the whole kind of datas of all the customers. In the customer journey, right? yeah, the first step of the customer journey is, of course, as you know, uh, you because you're a marketer, you know, you understand that they know the company, they know the brands, and they become a kind of estimated customers, buying customers, existing customers.

Rakusai: After that, customer support center is usually kind of position to the last piece of the puzzle. of the, uh, customer journey. So almost the end of the customer journey.

Ben Foden: Right.

Rakusai: But that now, [00:23:00] when you start using the tools like Helpfeel, then many kind of, not, not kind of buying customers, they is going to start using the digital support on their website. So now it's positioned to the marketing. And so, I mean, I want to say that kind of digital customer assistance experiences now. Um, covering a wide range of the customer journeys. It can benefit the first part of the customer journey plus the end part.

Ben Foden: Absolutely. Yeah. I, and I think, if you're doing research about a product, or a service that you're thinking about buying, you know, as a consumer or as a business, um,

Rakusai: Uh,

Ben Foden: you go to a company's website, you read about the products and. I think a lot of people today would like to read even from the knowledge base, but maybe they don't have an opportunity or the business [00:24:00] is not making that, you know, easy to find.

Ben Foden: But I definitely see the overlap there that you're talking about. Um, now I want to talk a little bit about the future. And, and you can talk a little bit about the roadmap at Helpfeel and some of the things that you're excited about. Um, but I'd like to kind of dream big a little bit and, and kind of have, have an open imagination about, you know, where do you see things going?

Ben Foden: What do you think is going to be really important in the future? And what are, what are some opportunities, you know, you could talk about AI or whatever technology.

Rakusai: Okay. So, of course the goal of the customer experience is solving their gap, reducing the gap, but the knowledge between the customers and companies, that can, of course, building more happy customers, plus happy companies. That's a goal and more and more company are now investing to the kind of physical customer support. [00:25:00] maybe some, some of the company investing 10 times more money to the physical customer support compared to the digital customer support. But now, now, now these kind of evolving the AI solutions is now showing it's become more possible to have really nice, customer experience online world. And we are working hard to realize that visions now. really excited to the futures now.

Ben Foden: Got it. And are, are there some new features or some new products or, even if we're not talking about concrete plans, but things that you would like to see in the future from Helpfeel.

Rakusai: Okay, so the first first call is, of course, we have we prepare for the answer answering the every questions of the users. But after after it's done, it's finished. Then we definitely want to work hard more and more to change the customer support department to the profit [00:26:00] centers. So adding more features that can, for example, for example, optimizing more than landing page that by solving the questions of the customers or detecting the customers more and more and more clearly and precisely by the natural word, we can definitely find more problems or that we can improve of the landing page or the whole customer experiences more and more and we want to create a methodologies to put this type of data into the whole customer experiences. of the customer journey.

Ben Foden: Definitely. And, before you and I, we've talked a little bit about how to get started. For example, if you're, if you're talking about the subject of marketing, you can go and get a degree in marketing at college, but there's no CX degree. There's no support [00:27:00] degree. And there's kind of a lack of concrete knowledge or concrete skills, um, compared to say marketing as one discipline or one, uh, one department of a business. And, you know, talking about methodology is one thing, but also just, understanding and the skill level, you know, the foundation, um, what do you think about this, this problem or this difference between CX and other focus areas.

Rakusai: yeah, the problem is that there was kind of no methodologies how to kind of Analyze the user's voice because it was simply the problem. It's very simple. It was simply natural language it was not analyzed by the numbers. But thinking about the marketing, they use the they are analyzing every day, but they see all the numbers. For [00:28:00] example, the number of the lead customers or number, number of the sessions or number of the users who have added the items to the shop cart

Ben Foden: Mhm.

Rakusai: of the conversion, et cetera, et cetera. But they are not looked at the, the real user's desires directly because it was simply not possible.

Ben Foden: Right.

Rakusai: the, only the way that they can do is just use an interview or something like that. But the number of the user interview you can do is very limited, and there's a risk that you hear only the small part of the large user base, right?

Ben Foden: Mhm. Mhm. Absolutely.

Rakusai: Well, there's now, but we have more chances that we can hear the user's voices equally and efficiently and more kind of rich user's voices and to have the better kind of service.

Ben Foden: Definitely, I think owning VOC, owning the [00:29:00] user voice, being the, the kind of amplifier of the customer, is really the opportunity and, and there's really, there's all these different things around VOC. You have that content. You have that natural language. And then how can you use it?

Ben Foden: How can you analyze it? What conversations can you have inside of the business, right? And you could talk to the marketing department and say, Hey, we've learned this. You could talk to sales, say, Hey, we've learned this, talk to leadership. Right. And every meeting you're bringing in that VOC data.

Ben Foden: Right. Um,

Rakusai: So additionally, that role was limited only to sales department or the marketing. And customer support is, they're not joining that conversation usually in the company.

Ben Foden: right. So as, as a CEO, I mean, when you think about this situation, you understand the landscape and you understand the importance of VOC. How do you think about this internally and, and, You know, would you be inviting this, the customer success team into a sales meeting, for example, or [00:30:00] a planning meeting for marketing?

Rakusai: Yeah. We, we are kind of B2B company, so we do not have customer support role in our company. So we have only a customer success now. Yes. Mm-Hmm?

Ben Foden: Um, right. But if you have that customer success person, like, and they are talking to the customer, they're getting that VOC data. They have the firsthand experience

Rakusai: Actually,

Ben Foden: and then.

Rakusai: yeah, we use Helpfeel by ourselves, yeah. We provide all kind of knowledges or kind how to use the Helpfeel or the, any kind of the. problems, questions, they're putting that everything into the Helpfeel. And so, yeah, this is helping us to getting the users voices.

Ben Foden: Definitely. All right. So I think we've talked quite a bit today. Um, I thank you for answering some of my questions that might have been more difficult. I'm trying to kind of push you a little bit and, and challenge a little bit here to. To dig in a little bit more deeply. [00:31:00] Um, now I, before we finish, I'd like to just close with a question for those people out there who are trying to increase their skills.

Ben Foden: They're trying to level up their team. CX leaders, CXOs, people out there who really want to get better at what they're doing and get better results. What is an opportunity or what are some things that you'd like to share with those people,

Ben Foden: something that they can do to improve?

Rakusai: Okay. I think learning more about kind of digital marketing methodologies. It's a customer support experience. It's getting more and more important. So these, these kind of two categories of the department's getting closer, as I said, mentioned already today. So trying to use more on the kind of digital analyzing a marketing tool is definitely improved.

Rakusai: Their [00:32:00] to, to, of course, the skills of the department and plus. How, how you should choose the tools. It's, it's, yeah. Yeah. We, we improve definitely.

Ben Foden: Definitely. Yeah, I think understanding the right metrics to look at and maybe looking outside of the normal or traditional metrics for support.

Rakusai: Yeah.

Ben Foden: Definitely sounds good. 

Ben Foden: All right. Um, well, I guess that's all for today. Um, thank you very much. Um, this has been the CX Files podcast with Isshu Rakusai, CEO and founder of Helpfeel, and we'll see you next time. Thank you.

Rakusai: Thank you.