Latest / Honest Ecommerce / AI Customer Support That Drives Sales: Text's Aleksandra Kacperczyk
Transcript
- speaker-0: A lot of people that reach out through customer support, like the answer exists on the website and they could just search for it, but they're kinda just being lazy. And I know this first hand experience 'cause we run C S on a on an app that we have. and it's oftentimes just pointing them towards the documentation. speaker-1: Yeah, definitely. I I also know that firsthand because I'm the lazy customer sometimes. And yeah, y you can automate a lot of those situations where the customer is that either doesn't want to search for something or is lost. Because a page can be extensive, can have a lot of sub pages. So let's get those questions out of the way. and if that works properly, maybe the AI can help us sell. speaker-0: Honest ECommerce is a weekly podcast where we interview direct-to-consumer brand founders and leaders to find out what it takes to start, grow, and scale an online business today. Hey everybody, welcome back to another episode of Honest ECommerce. Today I'm welcoming to the show the product manager on one of the selling teams at Text. Alexandra, welcome to the show. speaker-1: Hello everyone. Thank you for having me. speaker-0: I'm excited to chat. so for those that don't know, could you quickly talk about the products? Obviously there's text, there's a few under your umbrella, but like what are all the the apps that you are kind of interfacing with and and and helping out with over there at your day to day job? speaker-1: Sex is the company behind LiveChat, which you might have heard of, some of you, because we've been on the customer service market for over two decades. And some other similar names that will be familiar are Gorgeous or Intercom in the same area. And some time ago, approximately two years, I guess. It's been two years, we decided to pivot and to change the way we do in customer service. and Hans Text, our newest application, which was built in with e commerce in mind. speaker-0: Absolutely. For our the sake of our audience is ask our guests, it's like why why should I listen to you? Like are you an expert? So take me back in time. Like what have you been doing kind of in your career to end up here to where we should definitely take what you have to say seriously? speaker-1: Don't take me too too seriously, but hopefully I will be able to share with you some valuable insights. So I represent text, which has been, as I already said, in the customer service industry for nearly 25 years, and we've seen a lot of customer interactions and conversations. And those are huge amount amounts of data we can learn from. And I would like to share with you our best learnings today. and as for me personally, I've been in the SAS for almost eight years, a little less in the e commerce. it's been two or three years. But I find this area industry super interesting. I'm very excited. It's so far it's been the most satisfying part of my career and yeah, that's why I'm so happy to be on the on the podcast podcast to also learn from you. If you're on a customer support team, you should really stick to the end of the video because there's gonna be an amazing offer at the end. speaker-0: Absolutely. W with your career, did you go to school for this? Was this kind of stuff that you started to pick up iteratively at each new job? Like how did the kind of evolution of your career work? speaker-1: Well, no, I didn't plan to to work in tech at all. it was something that was really unexpected because I I got my engineering's d degree in biotechnology, so it's not related at all. But I found the like basic programming courses we had quite interesting and I thought why not check out the the the the the tech industry. So I entered into not programming specifically, but into tech. And from there I started taking on more interesting projects, working with some technical documentation, some marketplaces, third party apps, and as time went by I got a chance to to switch to to the e-commerce area and that's what I did. speaker-0: Absolutely. Yeah. I always like to to show folks that there's no right way to kind of get into this fun little ecosystem that we exist in. speaker-1: Definitely not. I've had so heard so many stories about people, how they got into product, business. Like literally everyone has a different background. So and every background brings in something new, something fresh, so it all matters. speaker-0: Yeah, the experience is that it's so funny where I've talked to some people that have like, you know, they'll have mastermind groups or something similar where it's like everybody is a the founder of an e-commerce brand, right? And then I'll meet other people, they're like, no one can have the same business on purpose because the way you think about solving stuff when everyone's doing the same thing becomes a little bit hive mindy. Whereas if you have like innately different businesses and the way that you're solving things, a there could be a more creative solution that exists in a different ecosystem that hasn't been brought in to yours. So obviously there's pros and cons to each approach, but I think, you know, basically what I'm saying is go get in one of each. speaker-1: Exactly, exactly. I think that it's why you can't really copy somebody else's recipe for success because your path will have different challenges. Even though you can learn from those, some some aspects can be can be the same. Ultimately, the journey will be different. And if that was if it was so simple to just copy what everybody else is doing, well. speaker-0: Yeah. Yeah, exactly. We do a lot of competitor analysis, for clients and in within CRO and we have to really press preface it for folks to be like, We're not doing exactly what they did. We're just getting inspiration to give us hypotheses to go test and see if this is a good idea for your business. 'Cause there's so much different and I've belabored that point. It's you know, things as simple as like how they're structuring sales. You like you don't know if they're losing money on that sale. You don't have the insights into how much they're paying to acquire a customer. It's like You know, just do what do what's best for your business. speaker-1: Yeah, exactly. Mm-hmm. Definitely. speaker-0: You mentioned earlier that you have twenty five years of data to go through, and that you have gone through and you've kind of got some insights you've learned from it. Where do we start? What what's kinda what's the first big unlock or a big interesting learn? speaker-1: Well, I would like to start with I don't know if it's going to be controversial, but it's my take on on customer support these days, is that a lot of conversations that seemingly look like just support conversations are in fact buying decisions in progress. And your inbox, wherever you're handling your day to day interactions with customers, is where you can actually sell more. And that's the starting point, I think, for the conversation. speaker-0: Absolutely. So you're saying that the people that care enough to stop what they're doing and actually open you know, a chat widget or whatnot and a conversation with you on your online store, these are high intent buyers. speaker-1: In some cases, yes. And it can be noticed in various ways. I believe that whenever you're chatting with a customer or even w if you're not chatting yet, even if they're just browsing your store and you're seeing them browsing, you can look for those high intent signals. Think for of a customer who has repeated repeatedly returned to a specific product page or has been going through one product page to another and you're noticing These are similar products. You can guess based on that that they're probably comparing these two products or that they're wondering about something if they haven't yet converted. You can step in, take action, ask them some if they need help, hopefully resolve their solve their doubt resolve their doubts and help them make a great purchase. another example would be returns or refunds questions, which look like simple, basic policy questions, but when people ask those, they often try to assess how much risk a potential transaction will involve. So it's your chance to establish trust. Not only solve like answer and solve the the ticket, but actually build trust with that potential customer. And if you establish that foundation Making approaches will be much easier, even if it it doesn't happen on the same day. speaker-0: Absolutely. I think with the first part of that with w if you see these these high intense signals and you reach out to them, I'm assuming is it like a widget on the store and it's just blop, I'm a real person. Can I help you compare product A and product B? Is it as simple as that? speaker-1: Yeah, it can be like that. It can be a real person, it can be an I an AI that sees those signals and is instructed to react on them. It's it depends on what setup you would like to have. speaker-0: Hey everybody, just a quick reminder. Please like this video and subscribe if you haven't. We're releasing interviews like this every week. So don't miss out. Now back to the interview. Absolutely. Yeah. And then going to the next part of that, you know, a lot of conversion rate optimization is just people do not buy because of fear, uncertainty, and doubt. And you you had mentioned that then they're trying to de risk the purchase by asking questions about these policies to see how that lines up against their, you know, their risk acceptance. So what you're saying, you know, is like, all right, you can answer their question about your return policy, link them to that policy, but also give them more information about why you're trustworthy. You know, we've got 10,000 five star reviews. We've sold enough of this product to go to the moon and back. We've got, you know, this whatever these cool like, you we are featured in Forbes. We were featured in GQ. Like what, you know, whatever these social proof elements are, you should also be sharing that. Because that original signal that they kind of flagged was like, I don't trust you enough to think I'm gonna keep this. How do I return it if that is the reality? So then you gotta go, well, here's the reason that people don't return it. speaker-1: Yeah. and even answering such such a question timely, not making the customer wait for an answer for a day or two, is a signal that they you're signaling that they can trust you. You will answer them timely and even if they want to make that return, that will also be handled with priority. So you're actually proving to them that you take their questions and requests seriously. speaker-0: Yeah, absolutely. You know, you take their potential business seriously and so that you you're gonna get back to them quickly. I guess that kinda leans into my next question here. And you alluded to it a little bit earlier, which is, you know, getting back to people quickly. Some businesses can't afford to staff with live chat twenty four seven with a human agent. and obviously there's been crazy advances in AI this year, it's everywhere, everyone was talking about so so I guess you're in this this ecosystem day to day, like how are you seeing AI and humans working together solving these problems for customers and like the customer support area? speaker-1: Yeah. I can share some more insights about how we restructured our own support team and how how human agents collaborate with AI to actually serve our customers. So that's also interesting, but I want to start with a more general answer to your question. I think that we're past that moment where AI only automates the repetitive stuff and it's actually skilled enough. to handle more independent, more risky transactions potentially, or at least tasks, right? Activities. And one thing is that it can spot those high intent signals, as I've already mentioned, it can react on them, it can proactively approach customers. So and if you let AI do that, you're putting a lot of trust on it because Like letting someone proactively approach your customers is a is a sign of trust, right? So but you can make that choice. You can you can test AI in a way that gives you confidence to to let it handle those interactions with your customer. And it can turn out to be beneficial because it can not only handle high traffic for you, it can answer twenty four seven day or night on weekends. And in multiple languages, which is also a huge benefit. and I remember one chatting with one customer of ours who selling who sells mattresses. And he said that the problem his business had was that customers would reach out to him late at night. and he was surprised by those emails he was getting or chats. He couldn't answer them because it was in the middle of the night. And he realized that people were frustrated with their mattresses, and that's why they started going on the internet, browsing mattresses, looking for alternatives, ended up on his store and started asking questions because picking the right mattress is not so easy. he couldn't handle those those questions on time, but imagine that you have an agent who doesn't sleep. You you can let it handle those chats, those conversations. So even such a unexpec une expected need can be addressed with AI. speaker-0: Yeah, I think that it is just wild how much we can get off of our plate these days. have you seen any kind of big claims about Lyft from AI? speaker-1: left in terms of revenue. speaker-0: Revenue f sales attributed to kinda just letting it, you know, all right, we're gonna let this thing kinda do what it can. Yeah, yeah. speaker-1: Yeah, I've seen some pretty impressive results. for example there's this one a customer of ours that's a pretty recent successful success story. It's a Finnish brand, nouveau, and they sell business grade refurbished computers. So a pretty complex business with a lot of technical conversations I I can imagine. I don't know ma much about hardware, but I imagine it's it's complicated. And they had this problem with a lot of traffic coming in on weekends and in the evenings. And every morning, every Monday, their team would start the day with picking up the conversations from the backlog, having to deal with with a lot of with a lot of conversations and unanswered messages. And they managed to train AI in a way that gives them a very spectacular resolution 98%. And once they saw that they can trust they can trust AI with their questions, they allowed it to recommend products, to be more proactive, to actually approach customers who were on the store just browsing. And with that they have increased sales by over four hundred percent in a matter of a quarter. So that's pretty impressive. We're very happy for them. They're probably happier. Great story, all in all. speaker-0: absolutely. So they historically it was all human agents and these things would pile up on the weekends and then they'd have to then it then Monday sounds terrible if that was your job, just getting through this backlog and getting caught up and refamiliarized. Yeah. It got to a point where they they were just kind of letting the AI run 247, try to handle all the conversations that Cody said 98% were just it could just handle them and then the other two percent obviously would probably get kicked up to a human agent or something like that. speaker-1: Well, actually in terms of traffic, I think their AI handles seventy percent of traffic. Gotcha. So yeah, so as far as I know, their support human support team who are trained computer specialists, they still handle some chats, emails and phone calls. They still have work, they still have they just have less on their place. They can focus on more complex cases. Yeah. instead of running from chasing each case. but the AI in it in the cases it handles, in the chats it handles, it's correct. It or or it returns a meaningful answer ninety eight percent of times. So yeah, we could say that it's yeah, it ha just has meaningful conversations with customers almost always. speaker-0: Yeah. but there's also something to say a lot of the a lot of people that reach out through customer support, they're just not like the answer exists on the website and they could just search for it, but they're kind of just being lazy. And I know this first hand experience because we run CS on a on an app that we have and it's oftentimes just pointing them towards the documentation or pointing them towards something that already exists. So like that's such an easy handoff that, you know, it doesn't need to even cross somebody's desk or you know or dashboard or or whatnot, you know, d any sort of distraction like that stuff's so easy to do. It's the it's the it's the actual complicated stuff. True bugs, true edge cases, you know, larger accounts that are looking for more hand holding, things of that nature. speaker-1: Yeah, definitely. I I also know that firsthand because I'm the lazy customer sometimes. And yeah, y you you can automate a lot of those situations where the customer is that either doesn't want to search for something or is lost. Because a page can be extensive, can have a lot of sub pages. So let's get those questions out of the way. and if that works properly, maybe the AI can help us sell. Yeah. speaker-0: But I think it goes back to the beginning of the conversation that we had where it's like if somebody is taking the time to reach out, they are more than interested in making a purchase. So it's like, you know, you should definitely be happy about, you know, and about the opportunity to potentially solve their problem for them with your product. speaker-1: Yeah, if if they're reaching out, that's a strong signal. you can also reach out to them, which is which is actually what got this this customer such a high result in sales because the AI agent started being more proactive. So it started sending chat invitations like, Hey, I'm I'm like I'm seeing you're looking at this page, can I help you with it? or some more contextual questions and people would would interact with those. and that ended up in those higher sales numbers. But yeah, customers who reach out to you are already down in the purchase funnel, basically in the selling journey. speaker-0: Absolutely. They're just looking for a couple of answers. Now, Alexandro, is there anything that I didn't ask you about that you think would resonate with our listeners today? speaker-1: I would like to encourage merchants to experiment with this mixed model of having human support, human customer service reps, and AI, and not being afraid of combining both and giving the AI like authority to talk with customers because there are ways to make the AI agent sound like your brands, like your customer service, human customer service representatives. And I know that some business owners are afraid that DAI will be pushy to salesy. And I completely understand those reservations because it's up to you, to your it's up to to you as a founder to define how you want your business to be run and how you want your sales assistance and support reps to talk with customers, but you can totally adjust that experience and make the eye more friendly or more professional, more chatty or more concise, speak in your voice. Basically what I'm trying to say is that you're in charge, you have the steering wheel, you can define the how it communicates with the customers and the what, what it does in how it reacts in various situations. you're the owner and you're the captain of that ship. speaker-0: Absolutely. That all makes so much sense. Now, if I'm listening to this episode and I'm curious to learn more about you and about text, where should I go? What should I do? speaker-1: You should go text.com and I would encourage you to create a free account to to test it out and to connect your store. We have an integration that allows you to do that in I don't know, in a minute, even or less. So it's not it's not a lot of work and just play with the application and see what it can do for your store, for your business. See the all the functionalities, the capabilities in action. See how the A agent recommends products, how it connects chats with orders that happened in relation to that particular chat. And hopefully you will find find value in the in the information we're giving you and in the capabilities we're we're offering. If you're on a customer support team, you can go to text.com, start a chat with our AI agent or support, whoever is responding to chats, and you can mention that you've listened to the podcast and we are happy to share with you our newest merge, which includes T shirts and sweatshirts. So just mention you've listened to the episode and we will tell you how to get it. No strings attached. speaker-0: Absolutely. And those are comped, you're gonna give away discount codes for those, right? speaker-1: Yes, yes, that's how you get it. speaker-0: Mm-hmm. Awesome. Alexandra, I can't thank you enough for coming on the show today and sharing all those amazing insights. speaker-1: Thank you.