Revolutionize Customer Support with AI and Automation | #231 Ray Nolan

Please support the show if you liked today's episode:

• If you love the podcast, please get someone else to listen, too!
• If you enjoyed this episode of the Ecommerce Coffee Break podcast, please head over to Apple Podcasts, leave a rating, write a review, and subscribe.
• Share the podcast with your family, friends, and co-workers.
• Tag the podcast on Instagram @clauslauter and let me know what you like about it.
• If you like the content and would like to support the podcast, you can buy me a coffee here.
• Become a guest on the show or sponsor an episode.

Claus Lauter: Hello and welcome to another episode of the e commerce coffee break podcast. Today, we want to dive deep, deep, deep into customer support and why is it so important for e commerce sellers. With me on the show, I have Ray Nolan. 

Ray is founder and CEO of eDesk. com, an innovative customer support platform for e commerce sellers. Ray has founded, mentored, and invested in many digital commerce startups over a period of 20 years, including Hostelworld, He founded eDesk after seeing friends running a high margin e commerce business struggle with operational efficiencies and profitable scaling.

Today, EDAS powers thousands of merchants around the world using AI to help them provide faster and more personalized support to their customers. So let's welcome Ray to the show. Hi, Ray. How are you today?

Ray Nolan: Great to be here, Klaus. I'm a fan of your work, so it's great to finally put a face to the words. Yeah, great to 

Claus Lauter: have you. Ray, you said you saw friends running e commerce businesses struggling with their customer support.

Tell me a little bit what brought you into this world of helping, customers. 

Ray Nolan: Well, I suppose my journey has always been around dot com, so I've been in travel and I've been in various different dot com businesses. But a friend of mine had an e commerce business, and he was quite large, he was selling about a million bucks a month.

But he couldn't really lay his hands on all the data for profitability. He didn't match his shipping costs against individual orders. He had support issues. He wasn't able to manage the support cause he was selling on multiple platforms. it really was a bit of a mess. And also the sheer number of.

Software tools that he was hacking together to try and make his business work significant and were a challenge to him. So I just went in as a favor really. And this is something that's been in software all of my life. I just said, let me have a look around and I could see them that there were things that nobody was really doing in terms of helping with customer support in particular.

 But also in the general sort of management of this business. so yeah, It was opportunity for me and I like building software like building businesses. So that was the genesis of how we started probably eight or nine years ago. 

Claus Lauter: You mentioned a couple of things being on different platforms, getting customer support from different areas.

I think that's one of the biggest challenges that was out there. And I remember seven years ago when I started with my Shopify store. We had problems dealing with customer requests coming from Facebook, from Instagram, from Twitter, and not using the contact form. Tell me a little bit about the story, how you got all these different channels together in one platform.

Ray Nolan: You mentioned channels that you don't even sell on that can be a real source of pain because it's very easy to go. I sell on Shopify. I'll hit my contact form and all the stuff will come into my email box or my Shopify environment. And I can reply in there, but the real damage can be done by someone tweeting, you know, Hey, my stuff hasn't arrived.

Anything left. In the abyss for hours or days or weeks, because someone hasn't pointed that out or gone to that area. And so we set about trying to make sure that we were everywhere. You could sell everywhere. The support will be needed. We were connected to. So if you sell on. your own Shopify store and you sell on Amazon or eBay or Walmart or Superdrug or any of 200 and something connections that we have, consolidate all of the support tickets into one place.

And then what we do is uniquely terms of our product, we attach the order data beside the incoming query. And what that means is the support agent doesn't have to go looking. , I can tell you a story even from my own data, I'm a totally geek, so I would've bought a piece of technology and I'm a support person, so I wouldn't be hard on support.

 I buy this product and it, didn't work. I go online, see how if I can fix it myself, it says install a new firmware, rebooted it, all that kind of stuff that you do. And nothing's working. So I mail support and I go, hi, I'm Ray and I've got your product and it's really good.

I like it. One thing doesn't work and I've done all the things. I've done the install. I've done the this, this, this. Can you help me? First question, Matt, I can't find your order. where did you order I ordered it on Amazon. Then I get, I can't find your order. And then. I get a few more.

And so I think there were five different communications before the person found my order. And then what did they say? Have you tried installing the firmware? And so they turned somebody from a recently happy customer, definitely on the side of the support agent into kind of raging bull. I'm going, gosh, lads I gave you this information, but it's so far down the thread.

obviously, I was very nice to the person and We solved the problem eventually, but that's a typical of the situation. So attaching the order is super important for us. So that means request comes in, query comes in, the order is beside them, the address of the customer is beside them, the tracking code is beside them.

You can actually press the button and see the person's house on a map. So in case they live, for example, apartment block, you can go, Oh, I see you're in an apartment block. Maybe you should check downstairs. It was definitely delivered. We know it was delivered, et cetera, et cetera. The first essence of what we did was A, consolidating all the tickets from everywhere and then B, collecting, connecting the order data.

And that just means when you think about consolidation problem or solution in our case. That means you don't have to have someone who's the Twitter expert, or the Shopify expert, or the Amazon expert, or the eBay expert. All the stuff comes into one place. You've got the same team of people solving all the problems.

And if it's an individual solving their own problems, then at least they're all in one place. They don't have to keep checking to other places. Adding the social network side of things was really important for us. So we added Facebook, Twitter, Insta tick tock. I'm sure missing one. And just to make sure that those tickets also came into one place.

And that meant you weren't worried that there could be somewhere that someone's asking the question that you're not able to answer. So consolidation was hugely important. And it's something that we still do. We still write. Integrations with new platforms with new places you can sell and the complexity the e commerce world means that if you get into distribution on platforms, maybe outside your own web store, you're kind of all in, you probably should try to be on all the platforms.

 If you're on one platform maybe look at another one and maybe be able to Amazon stores versus one and maybe be on Walmart as well, or depending where your geography is. So consolidate was the big thing. I think we support about 10 million. say we support the tickets, but 10 million tickets come through our system every month from various customers of ours.

So it's a pretty big business. And we also add. Sla's to the tickets. So, for example you might have an internal SLA just for yourself. So I want to get back to everybody within 24 hours of 48 hours. But, the platforms will also have an SLA requirement. So they'll go Amazon will have a 48 hour SLA.

If you don't reply. Your ratings are going down. Your internal Amazon ratings are going down. Not your star ratings, but the ones that they rate you on, as well as your star rating, obviously, your customer rating. But that might mean you don't feature in the buy boxes often, and it might mean that your rating in general doesn't, across Google or whatever, doesn't work, depending on where you've sold.

So hugely important. So we have different SLAs for different platforms. So you can say from my site. I want to be a six hour support, person but on Amazon, I'll meet the requirement or I'll try to beat it and make it day SLA 

Claus Lauter: I didn't know that, that Amazon has SLAs for customer support that's something very 

Ray Nolan: important to know. It's as simple as. If you close your business at 530 on a Friday and I reply to a support ticket saying, thank you so much class that was really helpful at 535 and you don't reply to me till Monday, your ratings going down.

, what we do is we have autoresponders that make sure that customer gets, you're more than welcome, you know, so it's those kinds of things that matter. It's all about nuances in customer support. 

Claus Lauter: Let's dive a little bit deeper into responses. I think most customer support requests are pretty much the same questions.

It's like, where's my order and something like that. obviously there's a lot of things you can automate hopefully and help your customer agent with their life, with their work. How do you do it from your side? What kind of tools do you offer? Yeah. 

Ray Nolan: , so we were always at the forefront in AI. And what we do is we categorize the ticket as it comes in.

Okay. Into 30 categories, everything from WISMO, where is my order, want to say thank you, I want to change it, I want a bigger one, smaller one, bluer one, et cetera, et cetera. So we categorize the tickets and then we don't let AI do the response, because AI is prone to a few issues, as you may have heard.

 My favorite one is autoparts. com or carparts. com. Both the same company listed US company who do many, many millions in sales. And they were trialing the AI. We were letting it generate the AI, the response. So a customer comes in and says, I can't get this tow bar to fix my car. And AI being AI and trying to be super helpful says, no problem, sir.

We'll send you a new car. So we quickly realized that while that was in early testing, we were not going to let AI generate the responses. So what we do is we check, we categorize the tickets and then we use template replies. That the customer can write and customize. So the standard answer to your WISMO, to your, where's my order one, maybe, Hi Ray, I write to you and I go, when can you tell me when my thing will arrive?

 Your template reply will go, Hi Ray, your order that was placed on Tuesday, 25th, was shipped on Wednesday, 26th, will arrive with you on Friday. And here's the tracking code. Thank you very much for your considerations that we want to put in there. That's all automated. No user agent, no support agent had to touch that ticket at all, and it generated and included the tracking code, importantly, and then it could be, I bought a lot of stationary for my office and I need to claim the expenses.

Can you send me an invoice that I can submit? And our system will interpret that. Create the PDF invoice, attach it to the reply, and send it back to the customer. All again, without the input from the support agent. What we're really trying to do is take the drudgery out of support, and make it so that what I call domain specific support is in the realm of the seller.

So domain specific is this shirt going to run in the wash? Why does it run in the wash? Not, where is it? Or, can I change it? Or, that's logistics. Frontline at first level support, if you like. I can see a world where in time, That's all the sellers will do is domain specific and they will, AI will do most of the rest of it

and it just means that, the agents tend to be happier because it's less tedious on them to support and because they've got all the tickets from everywhere in one place. they stay longer. This is also a HR issue. People in support don't tend to stay as long. If you can make tools that make their life easier, they go, well, this is a better place to be.

If they stay longer, their onboarding costs aren't as high, and so on. 

Ray Nolan: I 

Claus Lauter: think it's a very important point that you made there because it's a people business and customer support agents can be quite boring if you're doing the repetitive tasks. Being a custom support agent can be a profit part of your business.

So if you drive into the part of, live chat. Obviously you can generate potentially more profit for your business. Do you support this? What do you do on that 

Ray Nolan: front? We have live chat and we have a bot that does chat too. interesting. You talk about making it a profit center because Early on in our journey, we started to categorize things as presale queries.

Presale queries are really important and obviously on many platforms you can't chat to the seller. You can chat on, Shopify, and we have the tools to do that too. But let's say on Amazon, I see the shirts and I go, is it available in green? It's not listed in green, but I wanna talk to the seller.

Realistically, pre sale query, you probably have an hour to get an answer back. Otherwise I'm going to go to somebody else and ask. So we categorize and we move pre sale queries to a different section so they could possibly go to someone who they're like higher priority maybe, or they're going to Mary or John or Ray, who's a better seller, who's going to engage with the customer better and say, yes, I have it and here's a link to buy it and all this kind of stuff.

So pre sale query is a huge part ' of the way we handle support. And I think it's something that's, missed in certain areas and really lots of the software that's out there, both it be within the platform, be a platform like Amazon, et cetera, or even within your email box. It's very hard to go, which is the important stuff, which is the opportunity.

 People with general support queries probably okay to wait an hour or two, but people with pre sale, they're not waiting. They're just going to go somewhere else. 

Claus Lauter: I 100 percent agree. You mentioned before that you're dealing with a huge volume of customer support requests every month.

Obviously big sellers, they might have tens of thousands, hundreds of thousands of report requests coming through or of support requests coming through. How do you deal with. This volume, when it comes to reporting, monitoring all the kind of statistical side of it, 

Ray Nolan: we've got a huge level of reporting within the business, within the product, which will tell you who the agents are that are doing the best, who's got the best turnaround, who's more likely to maybe issue a refund on somebody else, which products.

Are generating most support requests. Another interesting one is a shoe seller in Manchester, actually in England, who would order shoes and they're custom made. But it turns out that all the size sevens were actually on a size six template, but he wouldn't know because he's going this is just part of the general run of the mail, lots and lots of support tickets.

Then we highlight, hold on this product in this size. Is causing 20

 percent of the issue. That's a real saver. And we are very much into like, helping you schedule people seeing when your peak times for support are. Cause it's really a lot of people tend to say, well, I need people from nine to five, or I won't shift. And then you go, well, actually you got almost no queries except lunchtime in the UK and 5 PM or whatever, usually different you can see when your busy times are, you can see your busy products. You can see which way you want to cut it. We can give you the reports on it. One 

Claus Lauter: thing I'm interested in with my store, I had an older audience and they'd like to pick up the phone and call in which was a bit of an issue because obviously then tracking whatever was happened was a bit difficult for us back in the time.

How do you deal with this? We

Ray Nolan: Have our own call handling, voice call, or we integrated one or two other. So if you've got an existing telephone system and it can integrate with us, we carry that too. And so when the phone rings and it's got a number. You can assign the customer to that number.

So when that customer calls back, you can say, Hey, Klaus, how are you? that's again, great for a presale environment so that you can recognize the customer. the world is moving away from voice. Certainly at the younger age, I actually don't want to talk to you.

They just want to know that you're on it. So it's important that we acknowledge the support ticket and we say we're dealing with it. , not in a sort of robotic way, but, we can give them assurances that we're on this and we expect to be back to with an answer within whatever time I think customer support's interesting.

Most people are happy even if you can't support it immediately, that you can say, I'm on it and I'm going to have an answer for you within four hours or by tomorrow. Once you set that expectation, people are happy. So when you leave them in the abyss and they're not sure whether you opened it or read it or whatever, that's a huge thing.

So yeah, it's very important to get that to people. And again, be it on Twitter, text, SMS, or, WhatsApp. We do WhatsApp support. Increasingly, WhatsApp, watch this space. If you haven't used WhatsApp for support, it's a stunning environment. a, it's asynchronous, but it's very, very personal. You're on my phone, I've allowed you to talk to me on WhatsApp, can kill the line of communication any time I want, but it's great because I get to have a little chat with you without having to, and I'm on the move, I'm on the bus, so I want to see where my thing is, so WhatsApp going to Really race ahead , in terms of support. So we want to look out for 100 percent agree. 

Claus Lauter: I think what's up is the one marketing channel that we'll see massive growth coming up. There's specific markets like the middle east, for instance, where what's up is really big in e commerce and other countries is a little bit less.

Now talking about the technical implementation, how does that work to get all these platforms together? How do I get started with edesk. com? 

Ray Nolan: We have a free trial. People tend to sign up for it with Shopify. We've got a special, environment that is for Shopify entrepreneurs, which is sort of a cheaper product.

It's a full system, but you can start really small. And then if you do integrate with another channel, you can add them in, but basically you just hit a connect you enter your Shopify password and connect the two bits together. We don't keep your password and like that, but you allow us access to your order and your support data, and then it's all there in one place.

And then you can add new channels as you see fit. As you expand or as you want to move to include all of your support environment. 

Claus Lauter: Okay. Very straightforward. What kind of steady onboarding time, how long does it take to get started? 

Ray Nolan: I think if it was Shopify only, I think you could say five minutes or less.

 It really is a case of, you know, your password to thing. you into Shopify on the right page. onboard there. It's 2 or 3 clicks and you're good. And then you'll see within our the future tickets will appear in our environment. By the way, they also appear in your old environment.

So the fear. Shouldn't be there. So you can say, well, I want to see how this looks, but does that mean I have to cut off my existing way I support? No, they're just replicated into our environment. And then when you're ready to actually use ours to send, you stop sending on this and you start sending with this.

Okay. 

Claus Lauter: Tell me a little bit about the pricing How does that work? 

Ray Nolan: We price on tickets, or users, mostly people go for tickets now 'cause it scales with them. If you go to ess.com/shopify, you'll see the Shopify, program and

it has an entry point of 35 euros a month. For 300 tickets, and then it goes up to 141 euros a month for 2000 tickets. So the ratio I would think of, depends on the industry, depends where you're selling. Obviously, if you're selling memory cards for cameras, the support is probably zilch because there's not much can go wrong.

It's clothing anymore, but typically for every 10 orders in most businesses, you're going to get a support ticket. So if you buy the 300 ticket program for 35 euros, it typically suits someone who's doing 3, 000 orders a month. And so you can go up to that. But the great thing is it scales, so grows with you.

 If you're looking at support tools, don't think about it as necessarily. They're all in the same place, but you've got to think of how efficient is this one versus the other.

This is not about somebody saying mine is 35 and his is That's a whole payroll to have that you didn't have because you had a bad product. So we're very much about making sure that you can get the max efficiency by using the AI, for example, to just make sure a lot of the humdrum stuff gets done And the beauty of it, support that really matters is that domain specific support. The other stuff, people get the answers and they get what they want. But the domain specific one wins you customers. That's the one that's where I get the personal response and I said we also have it in green if you want to buy that one, here's the link to buy it in green it's, getting the best out of your support team if it's one individual then so be it.

That's great too. 

Claus Lauter: A hundred percent agree on that one. The customer support is so important. I used to work in hospitality business and that's basically customer support, pure, a hundred percent, nothing else. And if you make a customer happy, they will come back. And that obviously increase your customer lifetime value.

Before we come to the end of the coffee break today, is there anything that you want to share with the listeners that we haven't covered yet? 

Ray Nolan: one last thing is translation and internationalization is a key topic. A lot of sellers are reluctant to sell outside their core market.

So I speak German really well, but I want to sell in France or do I want to sell in Spain or the UK? And I can get by and so on, but really am I going to have to hire? Hebrew speaking support agent we have agents now, we have companies now who are supporting in China. In English with not English speaking agents. So the ticket comes in English and it says, Hey, where's my stuff? It gets translated to Chinese. By our software, the Chinese agent responds to it in Chinese, in kanji, different characters, et and when customer receives it back, it's in English again.

And so you now don't have all of the pain of dipping your toe in the water in a different marketplace, or even supporting Spanish. If you're in the U S I can take your Spanish career. I don't have to have a necessarily a Spanish speaker beside me to do the support, my English speaking support agents can do that.

And that's a huge saving. You'll have all seen Google Translate and it was okay. This is perfect. There's no sense of, oh, this is robotic when it comes back on the first time. You should try it. Everybody should have a go, go into chat GPT or wherever and say put a typical response to a support ticket in and say, translate this into German and you'll see perfect German back.

Claus Lauter: you should make the best out of it. And it helps specifically in being more efficient. In your case, giving a better customer support. And that's why I like AI so much. It's really the features make such a difference there, right?

Where can people find out more about you guys? 

Ray Nolan: It's all on edesk. com. And if you go to edesk. com you get directed to the pure Shopify environment and you'll see. The Shopify entrepreneur pricing, which is different because dedicated to Shopify. If you want to go broader and go to everything it's the same price per ticket.

It's just that you probably buy more tickets. It's the entry point for our Shopify program is so low at 35 euros a month. It's kind of a place that you should try you don't have to sign up for a year you can sign up for a month play with it.

You've got a free trial. It has dot com. And then slash slash Shopify, if you want just to Shopify environment. 

Claus Lauter: Excellent. I will put the links in the show notes. Then you're just one click away. Great. Thanks so much for giving us a really good insight in how you can optimize your customer support.

And I think every listener should reach out to you guys and see what kind of difference this can make. Thanks so much for your time today. 

Ray Nolan: Great. Thanks for taking the time to chat to me.