If you want to improve customer service, increase sales, and build customer loyalty, then pay attention to this episode as you’re going to learn about how to leverage LiveChat, ChatGPT, and AI to improve not only customer support but also as a highly effective customer acquisition channel.

—

Welcome to the 2X eCommerce podcast. On this episode, I’m joined by Marius Laza who is the CCO or Chief Customer Officer at Tidio, a customer experience platform that serves over 300,000 eCommerce merchants globally, most of which are on the Shopify platform. In this episode, we discuss customer success and its importance in retention marketing and long-term eCommerce growth. You get to learn about how to leverage technology, particularly AI technology, in the right way and optimize for customer success.

Why should you read this episode? Firstly, you’re going to learn about the benefits of using AI-powered chatbots and LiveChat for customer support and lead generation, as well as how to use it as leverage to save time and improve communications. You’d also get to learn about the insights on making the most of your website traffic and turning more of your visitors into buyers by triggering the right chatbots at the right time.

Finally, you’d understand the importance of LiveChat as an underrated customer acquisition channel and how it can provide instant gratification to customers and lead to upselling and voucher opportunities. If you want to improve customer service, increase sales, and build customer loyalty, then pay attention. Without further ado, let’s get started.

—

Marius, welcome to the 2X eCommerce Podcast.

Thanks for having me.

I’ve been looking forward to this conversation because I feel like customer success is a missing piece in a lot of the conversations we have on the 2X eCommerce Podcast. Given the depth of your experience, you’re the man to guide us through, particularly in these trying times when every merchant and operator is looking to leverage retention marketing for long-term growth. Who is Marius Laza? I want you to take us back to your childhood and walk us through you being CCO at Tidio today.

There’s a lot to unravel both in experience and also in customer success so that’s a match there as well. I always worked in technology and I was always extremely passionate about technology even before I started working. I always had that bug of, “How can we use this?” Even when I got my first job in technology, which was SCL article writing, I was thinking that I could probably do this for free because I’m having so much fun. I was surprised that people were willing to pay me to do something on a computer and that was the beginning and that was years ago.

I had about five years of going through some SEO article writing, a bit of content writing, a bit of marketing, and community management. That’s when I started to get drawn toward customer success and customer experience. We can even talk about how the buzzword behind customer success and what that means in a bit. I then got the opportunity to work for a company called Instapage, which was all about landing pages and commercial rate optimization. I built a global customer experience team, about 70 or 80 people at one point, from all over the world.

I grew up the ranks to the VP of customer experience for them. I was always tied closely to customer success, conversion rate, optimization, and personalization. When I had the opportunity to join Tidio, of course, being a leader in customer experience, and working for a customer experience platform, it was a perfect match. I enjoyed the first year or so since I joined. I’m working now with a bunch of online stores but also other clients as well and leveraging customer success to increase sales.

That’s interesting. Prior to your SEO days, I also share SEO days with you and I know the impact of writing, what drew you to the net? What teenage years did you have? I want to go down to understand who you are. I feel you are techy by heart. Let’s go a bit further back. What attracted you to the internet and the web?

I don’t think this is going to be a surprise but it was gaming, of course. It must have been gaming, getting the first computer, and not being able to step away from it. I’m not sure how it happened, it must be typing speed game. I’m a fairly fast typer. For any of your friends out there that maybe are into that, my record is 136 words per minute or something, and the average is close up to 70. Fairly quickly, I started getting into typing.

I was gaming for a long time and then I started using the software. I had a nice stint of building maps for Counterstrike so I’m using the program that builds maps. I got into mapping design as a teenager. I then realized that maybe my typing speed is more suited to writing articles real fast. I was able to put out some volume and maybe rival ChatGPT with my speed of writing articles in my teenage years. it’s gaming to writing.

I also picked the bringing together personalization CRO to your role with Instapage. Do you want to speak to the importance of having that understanding so you serve customers as best you can from a CS standpoint, from a customer success standpoint?

I like the opposite perspective, the perspective of the visitor. I’m not even thinking about it from me as working at this company that does landing pages and conversion optimization or Tidio, which does customer experience, but me as the user of the internet. When you go online and you have an excellent experience, you don’t know the effort that went behind it.

I would venture to bet that whenever you had something unique, you went through something smoothly, you bought a product, you got something new, or even browsed the internet and it was nice, there’s some personalization behind it. There was a lot of SEO work and there was a lot of marketing behind it. You had a smooth and great experience because someone put a lot of effort into it.

To me, the shortest road to offering that experience is personalization so you have to know who your visitors are so you can give them what they were looking for. That’s the thing that I’m trying to replicate and help our clients replicate for their visitors. People have a better experience when they’re browsing the web. If you want that for yourself, you should want that for your visitors.

To help people, you need to know them from that fundamental. Let’s talk about Tidio. How many merchants? You’ve amassed lots of reviews on the Shopify app store. You seem to be predominantly eCommerce-focused, particularly on the Shopify platform, the biggest eCommerce platform on the planet right now. What is your size? How many merchants do you serve? Let’s jump into that.

We have a lot of strong metrics. We have the highest-rated app. If you search for Tidio anywhere, you’re going to see that we are always at the top of the ranks. We also have the volume behind it and I like that. We have about 300,000 active installs right now, globally, mostly online stores but also other websites as well. We are also getting about half a billion views of our widget every month. Massive volume and still an incredibly high satisfaction rate across any network should tell the story of what we’re trying to do. That’s where we are at in terms of those numbers.

We have a team that’s mostly based in Europe, about 200 people. We also have a few people in Latin America and someone in Canada as well. We are growing the team globally because we are working with clients from all over the world. We don’t have that focus maybe if you will. We wanted to see if we can expand this to everyone.

We have the free plan on one end, we have Tidio Plus on the other end, which starts at around $300 a month. We have this volume because we allow everyone. If you’re a small business, start slow. If you’re a larger business and you need our help setting it up, we can do that for you so you get a dedicated customer success manager, and we work with you on that. We have this range there and we are trying to work with them.

That’s a lot of merchants you’re serving and it warrants the size of the company. I’d like to ask, is the majority of your team, customer success in that self, trying to help this large volume, 200,000-plus, of merchants who use or who generates half a billion page views through the Tidio app?

We have to give credit to the product, a lot of it is self-serve, and anyone can sign up. We have this open feature policy so there’s not a lot behind any type of paywall. You can sign up and use it and you can use it yourself. We leverage our own tool to automate some of this. We have our own chatbot on our website within our application.

With the high volumes of new signups that we get every day, they get in the product and, luckily, the product is set up in a way where you can get up and running. Our average up-and-running time is eight minutes so you can be live in eight minutes. You can use our own chatbots and if you have any questions, you can reach out to our support team, of course, 24/5, which can help you. That’s the most we can do with the high volume of users that we are getting.

That’s when we get requests from some of our larger signups, larger clients, where they do want that one-to-one relationship, and we do have a plan for that. That’s where we have the human intervention but we were able to automate products and chatbots to help the huge volume of merchants help themselves and build what they need to support their visitors.

You’re product-centric and it reflects in the manpower behind Tidio. The elephant in the room, we’ve been going back and forth on nuances prior to this interview, which is important. I want people to understand that. Could you explain the difference between customer support, CX, which is customer experience, and then customer success? They’re all three different concepts, right?

Yes. I’m excited to talk about how people still see customer success as a buzzword and they also saw AI as a buzzword. ChatGPT then finally showed some power behind it, turning the buzzword into something sustainable. Let’s take it from the top. Customer experience is the full experience that your client or your visitors have with your website. That goes beyond a department of customer experience or customer support team. That’s marketing, that’s the product, and that’s the true experience of someone going through your website, through your service. To me, that encompasses everything.

You have customer support, which is a bit more traditional, and a bit more reactive in my view. I also think it’s a solved problem. I don’t want to undermine how difficult it is to offer excellent customer support. I managed customer support teams over the past ten years. I was a customer support agent for a few years before that so I know how hard it’s to set it up and I know how hard it’s to do it but it is solved.

Get multi-channels, take all of your channels in one place, set up a strong support team, have good internal procedures, set up your FAQs, set up your help center, and maybe create some other responders in there, maybe a chatbot. You’re then trying to find this match between the first response time, you can be fast, and then quality, customer satisfaction rating. If you can be fast, have high quality, and then if you have everything in place, you can get there. It might take a while but you can get there and solve customer support. Anything that’s reactive can be fixed. Any problems or questions, you have it done.

That’s where customer success comes in. Is it just an interaction that happens in a bubble? You get a request and your support team fix it, is that enough? Customer success takes that interaction and wants to get to the root of it. Who is this person? What problem do they have? How did they get on our website? What can we do to get them to come again and build a relationship with them? To me, customer success is identifying the true needs of your customers and then expanding that relationship by understanding them.

CX is the entire experience, support is reactive, fixing issues. When you know things hit a fan, it needs to be fixed, putting out fires. Success, from what I’m hearing you say, is understanding the customer and their needs and saying, “How can this be improved? How can they come back again? What happened?” Is that right?

It’s perfect.

With success, what key metrics do you look at? From the explanation, it sounds qualitative. I’m sure it’s quantitative because you need to apply it at scale. How should merchants approach customer success?

There is a tiny bit of overlap. Even in my own job with my customer support and customer success teams, I do run a lot of meetings together. Customer success managers will support agents at the same time. To me, that overlap might be even CSAT and an NPSs survey. You might want to learn about the quality of your interaction with your support team but also, how does that person feel about the product or the service that you’re selling? Those survey metrics and those NPSs scores start becoming the first customer success metrics that you could be looking at.

It then heavily goes outside of customer support with even sales. You can increase sales with customer success. That’s where you are identifying opportunities within a support interaction and upselling your visitor or cross-selling or doing any type of lead generation activity. Now you have a plain support interaction that turns into something that generates revenue. You can have revenue metrics for customer success. Traditionally, I come from a Software as a Service background so retention and churn are huge. You can apply some of the learnings from customer success in SaaS and customer success in eCommerce.

That makes sense. With customer success, it’s also looking at the unknown shopper. They might have questions, for instance. I tend to ask questions sometimes on eCommerce experiences that I have because I might have shipping concerns or I might have product quality concerns or usage concerns. I’m speaking to someone, live chatting with them, and interacting with them. You would say that that’s customer success activity as compared to customer support. Customer support is supporting an existing customer, right?

That’s right. Again, a bit of overlap. For example, there are tools like ours that leverage both. Now you have a client about to ask about the shipping issue and you can wait for the question and respond to it, which is more traditional reactive customer support or the system can detect that they have an open order with a shipping delay.

It can even automatically respond and say, “You’re probably asking about this order and it is delayed. Do you need more help?” That does blend the customer support and customer success aspect behind it. The point is, what about visitor interaction? How do they feel? Did they feel the difference of not having to chase down an answer and understanding, “This company knows me.” They know what I need and they’re here for me.”

With regards to resolution solutions as well as pre-sale inquiries, what is the data suggesting in terms of communication channels in eCommerce? Are customers or shoppers gravitating more towards instant chats for instant feedback so they can make decisions or get information? Are they willing the drag with email?

I purchased a flask and it had issues with leakage and the email is slow. It might be the issues with customer support in terms of their resolution times. What does the data show in terms of convenience standpoint? Where are customers interacting with brands? Where do they feel most convenient? There’s supposed to be a choice. What channels do you see a lot of customers use?

We talk about this a lot with our clients and the short answer is people want to talk instantly. I am being fully transparent here that they want to talk to a real person so that’s still a big thing in the space. We have powerful chatbots. Now, with ChatGPT out, there’s solid technology that can automate this but there’s still a stigma against it.

I would say that the channel is LiveChat, talking to a real person, and getting an answer quickly. It’s mimicking real-life experience. If you go to a store, you expect to find a person. We hear about this constantly. We want to have a real-life in-store experience online and we’re afraid that the chatbot can’t do it. The reason why is that we’ve all seen a bunch of underwhelming chatbots going online and we know how wrong it can go but it can also go right if you do it correctly.

What about availability? Not every team has the ability to be available for 24 hours. With chats, there’s an availability challenge or requirement for it to work well.

I need you on my sales team because that’s what we do when we pitch it because that’s the inevitable conversation that happens. People want that real-life experience, a personalized experience. They have agents and they can do it. We ask them, “What about on the weekend?” More often than not, they don’t have people on the weekend that are online. What about during the night? That’s the same issue. What about repetitive questions? You get the same question every time.

It’s truly wasteful for everyone if you go in with a generic answer. All of these cases go to the chatbot. If even if you have 24/7 availability, at the very least, the chatbot can route questions to the right people, and they can answer the most basic questions. Everyone agrees that you want to get an instant answer and if a chatbot can deliver it, that’s still good, you don’t have to talk to a real person.

From an agent perspective, do they log into a Tidio app? Let’s say I want to give our team, a team of three, CX or CS support agents, Tidio access. On a weekend, something pops up. How do they relate with customers if they’re not necessarily at their desks?

We have the full spectrum of it. You go into the dashboard where we have the connection so you have the LiveChat. We have ticketing and we have multi-channels so we connect to your Instagram, Facebook, email, and, of course, your website as well. You know for sure that all of the conversations are landing in the dashboard so you can see them there. You also know for sure that they are tied together. If someone reacts on Instagram, you can see the full conversation and not just a standalone question but the history of the conversation with that person.

You can answer directly and you can do it from your mobile phone. We have our Tidio mobile app as well. If you’re getting started and you want to be there and people get excited with their first few clients, you might want to answer from your phone wherever you might be, and then you start working with the chatbot, “These are my working hours. If I’m offline, then the chatbot comes in. If I’m online, I can answer myself.” That’s how that dashboard experience works.

I do know there was a mobile app but it’s convenient to have a mobile app. I remember when I was excited about Shopify, I had my first store, I had the app, and it would give me notifications. I love those notifications. Knowing that someone on the other side needs your help and it pops up on your phone. You do the same if you’re like WhatsApp. If your friend or a client reached out to you on WhatsApp, you will respond. It‘s amazing that there’s a Tidio app right there.

Here’s a big thing as well, that’s exactly what happens but the thing that I see the most and pains me the most when I talk to our clients is when they get their first visitors, it’s going to take a while until they get their first interaction. You should be excited when you get your first visitor on your new online store when you get your 10th or 100th visitor but it’s going to take a bit until you get your first order and the first interaction.

With our platform, you can see your active visitors on the site so you can snoop around and say, “This is the first visitor. Maybe they don’t have a question but maybe I can reach out to them. Maybe they spent more than five minutes on this product page. I’m going to start the chat myself and type something in and ask them if they need any help.” People are spending a lot of effort into bringing that traffic and, of course, a small percentage of it has the intent to talk to the support team. Most of them are browsing and won’t reach out to you so that’s your time to maybe reach out to them and interact.

It goes even way beyond that because you don’t even have to be interface and pop out and say something. You can tag them maybe for the next time, “This visitor came in so next time when they come again, I’m going to know that it’s their second visit.” A lot of the chatbot automation work happens on the backend without the client seeing anything or knowing and it feeds your customer profile so you understand your user base a bit better, “This customer spent a lot of time on this page. This customer came in for the fourth time. This customer is coming from this location.” A lot of it is building that customer visitor profile so you know who’s on your website and learn from them.

That’s interesting. What about the other piece like ticketing? There’s an issue, I reach out to you, my order is meant to come in today, and it’s not here. We’ve tracked it and it seems to be stuck. That’s a ticket. What happens?

Within the same dashboard, in our case, we convert LiveChat into tickets. In many ways, that’s the escalation method. Despite our best efforts, not everything can be handled on the spot. When you want to have the ability to go back to a ticket, you turn the chat into a ticket and you can even do the opposite and turn the ticket conversation into a chat. Everything still stays in the same panel so you still are getting this profile of customer interactions that has a nice timeline that you can go back to.

What I’m getting now is that Tidio is an alternative to typical customer support platforms with the ability to interact one-to-one through its chatting system.

We cover everything.

I have another question about customer success. There are different types of merchants reading this. There are merchants who are starting. They’re processing about 1,000 orders a month. I’ll say they’re in the $1 million or so revenue mark. You have some processing 5,000 orders a month and I’ll put them in the $5 million bracket. You have others processing 10,000-plus orders a month and I’ll put them in the $10 million plus.

In these brackets, you guys have customers in each of these brackets, what are the best practices for customer success for these calibers of teams, 1,000, 5,000, and 10,000 orders? How many interactions should they expect? How should they structure their teams and structure their processes in order to maximize their MPS scores and be the best they can be from a CS standpoint?

You’d be surprised to see that we get people that, before Tidio, only had the Contact Us button. You got the contact us, you click on an email address, and it opens your Outlook or something. They then turn on Tidio and you get massive volume almost instantly. You don’t even know how to stop it. In the beginning, there’s always an operation person. It might be you, the only person, dealing with everything but that means that you’re now the head of the customer experience, you’re the operations person, and you have to do it.

I’m also curious to know your thoughts because I know that this is your background as well but I feel like, in the beginning, you have to have someone to set it up and it might be you and that’s okay. The volume fluctuates, of course. For most people, it is low at the beginning so that gives you the luxury of time. Maybe you can set up some things, set up some automation, and think about the structure of your support system. Of course, whenever an interaction pops up, that’s an opportunity for a sale so you can throw in some customer success as well.

Small businesses getting started have a bit of the luxury of time. They do need to have some type of operational work behind it to set everything up and then they wait. Of course, it’s important when you hit that notification because every sale counts. Now is not the time to miss a sale and that’s why we even have this notify operator system to notify the team when someone important is on your website. You can set it up in many different places to make the most of the small traffic that you have. That covers this small segment.

You then go into the middle. I’m not sure what you want to call it, maybe the scale part of the business. That’s when you will get the first team members, operators, and people that can do this for you. You might have your first operators. We have had a pretty story that’s not officially out yet of how we did it as we were scaling and how we got up to ten support agents. The volume kept growing and it doubled but we didn’t have to hire twenty support agents.

For a while, we stopped at around ten despite the volume doubling because we managed to create chatbots that were assistance to each support person. For every supporter person, the chatbot took 50% of that volume. It took us about three months to set it up. I don’t want to recognize that it’s considerable time. Still, the need for an operations person, maybe a dedicated operations person at this point, is crucial. For us, it doubled the headcount. the chatbot itself does as much work as ten support agents. That’s what we had at the time and now we have a bit more and the volume has increased even larger.

When you mentioned an assistant through the chatbot, I posted this to social media where I said, “This ChatGPT tool should be an assistant in your organization. Whether it’s for copywriting, for CX, or writing called emails, every employee should have ChatGPT, a use case for ChatGPT, to improve and increase their productivity.

You mentioned that it took you two months to build it up for these ten employees but their productivity has doubled because they don’t have that availability. There’s some AI bot there helping. It’s profound in that sense, what you said. What about the scale? You are now ultra-scaling, doing 50,000 orders a month. What should your CS look like and availability? I know this is nuanced. A merchant selling furniture would have a complete use case for merchant selling protein shakes. I understand that. How does customer success change at that skill in eCommerce?

Also, this is where it starts to matter. A 1% increase in your conversion rate, that scale can have an impact on your ROI. I like this part of any business as they get there and when they are there. We know that we can go in and make a small change and make a massive ROI impact and get them to sell a lot more.

It’s a lot of strategies. This is where you have a lot of strategies. This is where you have a lot of trial and error. You can think about all the days of A/B testing and pure conversion optimization. At the end of the day, that’s what you’re doing, even if you want to stay in the chatbot space and even use the chatbots and automation, to your advantage.

I want to give some even specific examples. Now that you have a large book of business, you have a large set of customers and former visitors on your website. When they jump in and they have an interaction, there should be some type of internal notification for your support agent or your live support agent saying that this client has purchased this product before or this client is returning for the third time.

This is all of that hidden automation that people don’t talk about. If I start talking about chatbots, everyone believes that it’s some type of popup or something where you type in, “Please take me to a real person,” or, “I want to talk to a live agent.” That’s not it. I don’t have the full data but I would say that at least half of our automation is hidden and, at scale, it makes that impact. Still, you have to have your support agents, you have to have the operations person, and you might get into strategy quite a bit so you can properly set up these chatbots. On our plan, Tidio Plus, we do it for our clients so they can skip the learning curve and they can get to the results.

What I’m hearing from what you said is that this is context, it’s about getting deeper into who we are speaking with. A lady walks into a yoga outfit store and she fits the profile of someone who’s purchased in the past. That familiarity from your support person or your sales rep has to be there in a store environment. You do what she expects you to do whether it’s to leave her alone until she asks you a question or you reach out to her and say, “Hello. What are you looking for?” There’s that context. You need to implement when you are scaling and that should help them faster so you could get into the next one and increase your efficiency.

It’s things that you don’t even think about like, are they visiting from mobile or from their desktop? Maybe she’s on our way to yoga. Maybe give her the order number or whatever she’s looking for. If she’s in the car, on a bus, or something, she’s probably trying to make this interaction faster. If you have that information, if you identify the cohort of users that are always on the run, maybe you want to slightly adjust the customer experience for them.

If they’re on a browser, on their desktop, or on their computer at home, take your time and do a deeper interaction. It’s small things like this but you start noticing behavior patterns and you start noticing segments when you have the traffic and when you have the volume. At scale, everything becomes a bit more powerful, the impact is a bit bigger, and the ROI can be massive. That’s why the full-scale segment of our users or our clients gets a lot of value. I enjoy working with them on these projects.

Thank you for that, Marius. Let’s speak to ChatGPT, AI-powered tech. In my predictions episode, one of the first episodes this 2023, I was like, “There is a need for store owners right now in the Shopify ecosystem, and with the power of the APIs to have AI connect to first-party data within Shopify.” This AI bot is aware of all of the orders, their statuses, and all of your customer information. At the snap of a finger, you can either query it and say, “Who’s ordered over $200 today?” It makes you agile. It’s almost like a query engine on the one hand.

Also, from a chatbot or a customer success perspective, it makes your chatbots cleverer. Your chatbot is also able to know, “This person has purchased from us seven times in the last twelve months. They are a VIP customer.” There’s a better turnkey solution. Especially if there’s no agent or when they’re about to hand it over to an agent, they give that agent that context. The experience for that person is different from someone who’s exploring, researching, and trying to figure out what to do. Do you agree or have I left certain bits out of that use case?

There are even more use cases of course, but you covered it quite well. It seems like you agree, it’s finally happening, and it’s not a buzzword anymore. It’s hard to deny at this point. What we’re going to see moving forward is going to solidify the power of AI correctly. We’re doing a bit of that. We have five AI features already. We started a bit earlier with the first AI features. There are some that are almost perfect for what you would need.

Think about support teams and regular questions, where do you go as a support agent? You go to your internal knowledge base or, in a lot of cases, you might even have the article in the public help center. It’s probably 1 of the 2. We probably see over 90% of questions are questions that exist somewhere in either the external or internal knowledge base.

With standard NLP model AI, you can have the chatbot learn your internal knowledge base, learn your external help center, and pick up on the answers, and it’s perfect. In many interactions, it’s perfect. It gives the perfect answer quickly. It doesn’t send your client on a wild goose chase with a link, it doesn’t have to do that. You’ve seen ChatGPT and how precise it can be with the answer if it has the knowledge.

We have this thing called Visitor Says so the client doesn’t have to take the responsibility of asking the perfect question. We identify questions regardless of how you ask them. You can have spelling errors, you can ask it in ten different ways. The chatbot would recognize what you’re talking about and give you the proper answer at the right time. This is our core AI feature that we had for quite some time and we are working on some advanced stuff already. I can even share some with you.

That’s incredible because that product expertise and product awareness, especially for a merchant selling somewhat complex or hard-to-use products, products that need instructions. If the AI is able to cut to the chase, understand, and go through the manuals, it’ll be helpful. What do you think about the future of customer success and customer support teams? Are they going to shrink off the back of AI? With ChatGPT, the ability for it to be articulate based on the data set it’s referencing is incredible. How do you think it’s going to impact particularly in 2023 where margins are being hit by inflation? What are your thoughts there?

I like scaling without having to scale the team. You’re scaling your revenue and you’re getting more orders but you don’t have to grow the team. This is exactly what we did with my example, with our team getting support members and then the company doubling and tripling in size without having to hire initial support.

For that, this is maybe the more conservative approach but it’s valid that people are right to do so. We have a feature that we call the Magic Wand. Instead of just using the chatbot to give the answer and have a full AI interaction, we have the support agents type in an answer and then use the Magic Wand to get the suggestion from the AI or rephrase the question even or the answer. We believe that that’s a nice baby step toward the solution.

We were playing around with writing a quick answer, “No, we don’t have the product.” You click on the Magic Wand and it says, “I’m sorry for the inconvenience, it looks like we don’t have the product in stock right now. We are going to get it on this date.” You turn a one-minute interaction into a five-second interaction. You click send and your support team is now supercharged and can handle a lot more requests. You have the confidence that the answers are still perfect. Of course, they can edit the answer before they send it. It’s going to increase the scale of what the support person can do.

It takes me to two questions and one is the tone of voice. If you are buying from Kylie Cosmetics, it’s a totally different experience than if you’re buying from Tesla, two extremes. The AI, does it adapt to the tone of voice?

We do use ChatGPT for this and we’ve been playing around with it quite a bit. We have the full connection. We have our own AI models as well but we also leverage their platform. We have the full power of what you see in all of those TikTok videos online. What we’re playing around with was responding RAP forms. We did have a bunch of internal RAP responses. Of course, you can give it the tone.

Our larger clients do have their own brand voice, they do have that. We are working on training the model to recognize it. It’s using your own internal answers and internal knowledge base so that you can find your voice. Right now, it can do different types of voices. Of course, it can make the message friendlier or it can make the message more exciting or shorter so that’s straightforward. As you go into a specific brand tone, we’re getting there.

I run a woman’s brand. One of my experiments back in the day was a woman’s fashion brand. We were processing a hundred orders a day with that brand. We were targeting 18 to 24-year-olds and we first realized that 18 to 24-year-olds are broke, they don’t have much money. They want to buy $20 dresses. The interesting thing about that demographic is there’s a certain way we had to talk to get them. The brand was even called Miss Slay, slay, and all that stuff.

We then moved onto the 30-plus demographic and started to sell bandage dresses, dresses that you’d use for dates, and whether you’re looking for maybe your partner for life, those kinds of dresses. You have an event like a gala or whatever. The AOV grew and we had to also change our tone of voice. That would be interesting. I’ll give you an example again. I need to go back to that story. We had this amazing customer support rep and her name was Mariana and she was in the Philippines. She got the culture. She was on social media. She knew all the celebrities and all the female influencers. She got that tone of voice.

When she left us, it left a gap in communication because that was the tone of voice we wanted to connect with. If you then have something like the Magic Wand feature in Tidio, you are able to normalize anybody’s tone of voice whether you have offshore resources or even inshore that they don’t get the tone of voice per se and normalize it. What do you think about that use case?

It’s all about the volume again. You have enough volume and you have enough conversations and then you can trade the model. Still, it’s the two-stage approach. The first stage is freeing up some time for Mariana to work in her zone of genius. She can work with clients and talk to clients. When she has to answer a boring support question, she can use the Magic Wand and send the answer and that’s fine. The Magic Wand might even work in her own voice anyway so you all know the difference. She can spend more of her time doing what she likes and talking to clients and understanding them.

We also noticed that we’re filling up sometimes from our support team and, suddenly, we have 1 or 2 support people that are masters at lead generation. They notice this client, “They have a question and I’ll fix the problem.” They could do much more on Tidio Plus, “Let me send them to the sales team.” Before you know it, we’re getting leads for the sales team from support interactions.

Before, talking to 2 or 3 people at the same time, you don’t have time to think about, “We fix the problem but they need an implementation specialist here so we can build a stronger chatbot for them. Let me send them to Tidio Plus.” A lot of these things didn’t happen before because we didn’t have the time like how support agents are when they have to multitask, it’s difficult. We fill up the time and they can be available. Of course, stage two is having the AI do some of that work as well.

I’ll add to say that LiveChat is the most underrated customer acquisition channel. Brands are not leveraging it enough I have to say. A lot of people have questions. It’s important not to delay that instant gratification people have because people are ready with their credit cards from my experience. Especially in fashion, they’re like, “Would this dress fit?” They would’ve read the description but they don’t want to go back and forth. They need that answer and t then you could upsell or you could give them a voucher. There’s so much stuff you could do off the back of LiveChat because you have that opportunity to speak to them one on one.

The final question is, what are our suggestions or what’s your advice to eCommerce brands in 2023? In most sectors, there are many headwinds Customer demand is suppressed a bit but not necessarily in every sector. How should operators navigate growth through customer success in 2023?

It goes back to what we talked about at the beginning as well because it is about making the most of your traffic, you’re getting that traffic. Usually, it’s high-intent traffic that you’re getting. Maybe marketing teams are doing their best to do that. In my experience with clients, you would be surprised to see the relationship between traffic and sales.

We have a bunch of clients that have mid-level traffic but incredible sales. You would be shocked or surprised to see how much they are able to sell by getting high-intent traffic and then turning more of their visitors into buyers by triggering the right chatbot at the right time if you want to put it that way. Building that nice flow of understanding, let’s see what we are able to bring on the website and then what they’re looking for. If we can give them that, that’s critical.

For that, you don’t need the volume, you can survive the headwind. There is a wave now with ChatGPT and the advancements in technology. You do have to make the most of what you have. It’s not about bringing even more people to your website. It’s bringing the right people and then leveraging technology in the right way to get the job done.

Incredible stuff. Marius, we could go on and on but it’s been an incredible conversation I’ve had with you. Marius Laza, from Tidio, thank you for coming on the 2X eCommerce Podcast.

Thanks for having me. It was great.