197. ConverseNow.AI Transcript

ConverseNow

Sat, Aug 5, 2023 . 9:59 AM

51:17

Owner: Jeremy Julian

SUMMARY KEYWORDS

ai, restaurant, order, talk, team member, guest, people, upsell, brand, ordering, drive, technology, drive thru, menu, work, experience, converse, pizza, call, store

SPEAKERS

Ben (60%), Jeremy (40%), Intro (1%) 

I

Intro

0:02

This is the restaurant technology guys podcast, helping you run your restaurant better

JJ

Jeremy Julian

0:14

Welcome back to the restaurant technology guys podcast. Thank you everyone out there in the the internet regions that gets to listen and watch now that we’ve got a YouTube and a live stream on LinkedIn. Today we are going to be talking about our pretty cool topic related to AI and voice and just kind of how AI has really taken over our world I was just talking with with our CTO the other day about just how much AI has become, I guess socialized in a way that I think there’s a lot of misnomers about what it is. And so we’re joined by a guest that’s been living in the space for a while they’re trying to solve some problems for restaurants. And so first, I’m gonna introduce you to Ben, Ben, why don’t you introduce yourself to our audience? Who is Ben Brown? What does Ben Brown get a chance to do? And then we can talk a little bit about your role at Congress now and how how you guys are, you know, taking on the world.

BB

Ben Brown

1:09

Great. Hey, Jeremy, thanks a lot for the intro. And for having me. It’s, you know, as a big fan of everything that you do, at the restaurant technology guys do. It is humbling to actually be on here speaking with you.

JJ

Jeremy Julian

1:22

I don’t want my kids know, because they like to make fun of me all the time. I like that, you know, they, they totally make fun of me about it. We were at a trade show recently, and somebody came up to me and they were like, Hey, I’ve heard your show. And they’re like, they’re like that this is ridiculous. So sorry, I’ll I appreciate you. But

BB

Ben Brown

1:39

you’re you’re a celeb in the space. So for those tuning in, Jeremy and I have gotten to meet several times on the conference circuit. You know, it’s a very, very dedicated group and restaurant technology. So yeah, I’m Ben Brown. And I serve as vice president of marketing and customer experience at Converse. Now, converse now is a voice AI company that automates drive thru and phone orders for some of the nation’s leading restaurant chains. We do that using voice AI powered virtual ordering assistants that take orders in the same way a physical team member would. So a guest speaks to the AI in the same way it would speak to a regular person, a natural, free flowing conversation where they can ask questions change their mind, you know, disregards the background noise and honking horns and barking dogs and all those things. And you know, provides a pretty good experience for the guests for the staff working at the store. And of course, for the restaurants bottom line. Well, we’ll get into detail there, right, I’m not going to go into it right now. But it’s it’s an amazing group to be a part of, we’re based in Austin, Texas. I’m personally based in San Diego and work remotely. We have a global team. It’s just an incredible group. And the only thing I’ll add to your note about trying to solve some problems, I am happy to say that we do have some members to back it up now we’ve been solving this problems pretty well. So yeah, it’s been good.

JJ

Jeremy Julian

3:08

Well, and I and as as you alluded to, in your intro, you and I have gotten a chance to talk on on kind of the conference circuit. And so I will be truthful. And this has been probably 18 months since I heard about it. Even 18 months ago, I was skeptical, I am certain a lot of our audience members are gonna be like, Yeah, you’re gonna have an AI, take my order, you’re gonna have an AI, you know, the amount of permeations that go on with my burger and go on with, with all of those things. So I’m certain we’re going to dig into that. So I’m really excited to educate our audience about, I guess, the I think everybody understands the why, but let’s dig in a little bit then on why do we think we need this kind of technology in today’s day and age? We’ve had drive throughs for 60 years, 70 years, 80 years, and they weren’t, they weren’t the speaker boxes work, the the ability to take that order and get it and ask the questions has worked. Why now? Why is technology? Why should technology be the way that people are thinking now versus you know what, we got a staff member, they’re gonna do this and take care of them?

BB

Ben Brown

4:16

Ya know, it’s such a good question, Jeremy. And, you know, I think the same question was probably posed to Henry Ford, when horses were working just fine. As a primary mode of transport, the same question, you know, was leveled to Steve Jobs. When, you know, computers were, you know, for a very niche group of people that wanted to build them themselves and not really marketed for a mass audience. Drive Thru has been ripe for innovation for literally decades. And you know, what’s working is working, right. But the question is, is there a way to make it better? And the answer is absolutely. So, if I was, if I was to pose this question to you as a consumer, right as a pair are some who would go through the drive thru the order food? You know, maybe when your kids are screaming in the car and they just want food now? What if we could make drive thru lines go faster? What if we could make it easier to hear that person on the other line? What if you felt you know, more cared for and more value that because you weren’t feeling rushed to get through an order? You know, what if your orders were more accurate, right? I mean, the list which are all

JJ

Jeremy Julian

5:28

problems that every drive thru deals with order accuracy, getting the right getting understanding what the guest was looking for pushing them through the line, you’ve got that family, I, you know, many of our longtime listeners know that I’ve got a family of six, four kids. And so driving through the drive thru is a lot of work, you know, driving through the drive thru and trying to figure out who who wants what and, and such. And I know that you know, and you guys have been around for a while. So it was even before the labor crisis hit, but some. And so, back to the skeptic that I talked about just a few minutes ago, I was a skeptic, because I think a lot of people, when they think of voice AI, they think of my cable company, or my cell phone company, or my credit card company, where you just scream into the phone representative until you can actually talk to someone, you don’t engage in it. Because so often, they don’t get you what you’re looking for. We all know what the experience could or should be, when we go to a drive thru because we’ve been doing it for 60 years. Or the mode to your point, ripe for disruption has not changed in six years, you go so Squawk Box, you may or may not be able to hear the person, they may or may not be able to hear you. But you’re gonna shout into that box what you want. And hopefully the bag that comes out is going to have the food that you ordered, or the desserts that you ordered or the drinks that you ordered when you get up to the window. Talk to me a little bit about the innovation and why why the technology is so much different than what a lot of skeptics think of when they think of calling American Express. Or they think of calling you know, at&t, and they’re pissed off and they want to get their bill figured out or they want to get you know, that’s what a lot of people think of when they hear voice AI and go no, no, that’ll never work for me.

BB

Ben Brown

7:12

There’s no question about it, Jeremy, you are hitting it right on the head that as it currently stands, the perception of voice AI are those archaic systems that we are all unfortunately used to that tear your hair out feeling when you have to press a bunch of buttons. And you you just cannot get a question answered a one size fits all methodology that doesn’t work for hardly anyone. So happy to say that voice has come a very, very long way. You’ve noticed that a couple of times now Converse now has been around since 2018, quote unquote, before AI was cool. You know, far before the the open AI chat GPT trends took off. You know, we we’ve built our AI in house. And we have really seen it from its inception to where it is right now. And you know, the the technology there, it’s natural, it’s free flowing? And at this point, yes, you’re you know, that you’re speaking with a virtual assistant, oftentimes, you know, the AI will introduce itself as a virtual assistant. So you know, you’re not getting fourth. But from there, you just talked to it, like you talk to a person, right? You just say yeah, I’d like to number one, you know, catch up extra pickles, and, you know, actually make it the number two. And, you know, I’ll take a medium Diet Coke. Oh, and like, can I, you know, can I use this coupon that I got and the I’ll just, you know, process all of it, and repeat the order back to you so that you know you’re getting, you know, the order that you’re asking for. And boom, you just drive up the staff take it in, you know, without an issue. It’s as if a number one of their colleagues took it. There’s no real training needed on their end. It’s just really seamless at the back of the house. But yeah, going back to the guest experience. That is that is the number one most important thing that we focused on, from our inception in 2018 was what is going to create the experience that guests wants to have, you know, this isn’t about people tolerating technology, it’s about people embracing it, right? Why then people flock to the newest MacBook or the newest iPad or the newest iPhone, right? Because it’s something that’s cool that they love to engage in. And it really blends, you know, the idea of, you know, art and technology, and that’s exactly what we’re doing. Right? Not necessarily for consumer hardware, but for the ordering experience that people are looking for. Well, and

JJ

Jeremy Julian

9:52

we talk a lot about it on the show. Almost every single episode I talk about the guests experience and delivering technology for technology’s sake is worthless. I’ve said that 7000 times because technology that doesn’t try try to solve or ultimately solve business challenges for consumers, or for the business doesn’t get adopted, it has to have to solve a problem. And so my guest experience needs to be as good or better through this AI than it is with a 16 year old that started, you know, three hours ago and has very little training on the menu and has very little training of who I am and what I want. So talk to me about the power of AI and how and why AI has the capability to replicate that experience, but make it better every single time because that’s the thing that, that a lot of people that aren’t in the technology space don’t comprehend is that the AI is self learning, I was explaining to somebody the other day I happen to drive a Tesla, I’m fortunate enough to to to have a Tesla. And that thing is constantly learning. And so every day that I’m going out, if there’s different road conditions, it’s changing the GPS, and it’s changing all that’s going on learning from the other cars that are around and voice AI and AI in general is constantly doing that. So talk to me talk me through a little bit of that, and educate our listeners, because most people do, they just think it’s this box. And it just goes in and it’s pulling from a Dewey Decimal System. I’m pulling these books, but it’s not that way. Like I know that that sounds really archaic. And at the same time. i It’s hard to comprehend how much computing power and how many times that can evaluate the words and change what it’s doing based on what’s happening. While it’s self learning.

BB

Ben Brown

11:41

Of course, happy to walk through it first. humblebrag for for a Tesla owner, that’s great.

JJ

Jeremy Julian

11:47

I feel bad thing. Like my 15 year olds like Dad, can you drive me to church camp in the in the Tesla. I’m like, No, I’m driving you in the minivan, dude, like you need to be cool when you’re going to church camp. So.

BB

Ben Brown

11:59

But yeah, so so the way simply put, think of AI, like a regular team member. But instead of that team member learning as it goes and taking basic training, it is pre trained, we program the AI for each restaurants menu, and integrates directly into their POS system, into their drive through hardware, the base stations, etc. So that it’s starting with all of the training that you would otherwise give a regular team member, that’s, that’s ground ground zero. And that takes place before we go live in the first store with the brands. Now, take your hardest working team member, you know, those golden team members that everyone, you know, everyone wants, where they’re genuinely, you know, fun loving, they’re endearing, they’re very hardworking, and they just want to create a great guest experience. Now managers, you know, sees opportunities to improve and educate those team members with teachable moments all the time. So imagine every manager in every store that you operate across the country, talking to that team member at once, and that team member actually being able to listen and absorb that information, and never being able to forget a thing. That is what AI is like, Yep, it’s a self learning model, where the more conversations that it takes in, the better it gets just like your Tesla. And each guest gets a better experience moving forward. So, you know, we’re live in nearly 2000 stores across the US, we’re operational in 46 states, and we process you know, more than a million conversations every single month. And all of those conversations feed into the AI. So that it’s not just learning from, you know, your stories, it’s actually learning from other stories as well. We’re keeping all you know, proprietary data, or data proprietary, but there’s 101 ways to you know, say the word pizza, there’s 101 ways to say the word taco. So, you know, there’s a lot of cross pollination that takes place for everyone to win. And, you know, just imagine like a powerhouse team member that has experience working in virtually every other restaurant out there. And they’re bringing all of those key learnings into your brand. So, you know, it’s, it’s a pretty Win win scenario.

JJ

Jeremy Julian

14:31

Well, let me think, again, I appreciate you educating pupils, because because again, I hear these, you know, super binary because they just haven’t spent enough time digging into how the AI continues to learn again, I spent some time talking in this was even before Chatri PT talking to our CTO about I’m like, Why? Why are we able to do this and with the cost of compute and the cost of storage continuing to go down. It allows you to build these models. Talk to me about the model building because with a lot of AI It needs to build a model. So let’s take that burger chain that’s got six different, you know, combos, and they’ve got them individually. And, you know, any kind of guests self ordering technology, whether that’s voice, whether that’s web, whether that’s kiosks, you have to figure out those permeations of how how people are going to order them, and build them into the system where you’re going to hit roadblocks and have bad guests experience. How, how does the modeling work? Do you have to have that best team member sit there and tell it how, how Suzy orders it and how the you know, the lady that’s you know, the grandmother that’s been ordering is, is it self learning, talk to me a little bit about how it learns what is going on within the menu, because without it, you’re going to deliver a worst guest experience?

BB

Ben Brown

15:45

Yeah, absolutely. And what we do is we’ve created a number of in house platforms that are dedicated solely to training or AI. So we’ll like we have a tool that allows the AI to get trained on your menu and every possible permutation and combination within that menu so that it’s deeply familiar with the brand. That actually sits on top of a very strong foundation of general conversational AI, and restaurant specific AI language that’s used in the utterances that are exchanged. So going back to that note on, there’s 101 ways to order a pizza. So for onboarding a pizza brand, right? There’s that base layer, that just talks about the way that people converse, in general. And then there’s the restaurant in general layer that talks about basic restaurant speak. And then there’s the cuisine specific layer, which is, okay, we have data about pizza, we have data about burgers, let’s apply it to this concept that has the same, you know, type of menu, and then we have that that layer on top, that is very menu specific. So that’s really tailored to the brand. So if you think about it this way, the models that we build, there, it’s meant so that about 80% of the work is done even before the brand signs up, right, because the AI is ready, it knows how people order food, it knows how people order that specific type of cuisine. And really, it just needs to be trained on your exact menu. And that includes your hidden menu items, your specials, right? Things that are unique to the stores, you know, specific promos going on, like fundraisers, et cetera, you know, if if a store, you know, has different operating hours, or if there is, you know, I don’t know, a blizzard, that sweeping through in the store is closed, you know, our AI senses that it will shut down when it doesn’t have to go. And then likewise, you know, it operates when it needs to. So we’ve just really created, we’ve created this, this virtual team member that can just, it just needs some topping off in order to work with a brand new brand. So that the training process, you know, takes very little time.

JJ

Jeremy Julian

18:11

I love it. And I think to myself, I mean, your San Diego guy. So you know, I was just talking to somebody who’s who I was with him last week, and he just moved to Orange County, and he’s like, Man, I can’t get myself enough in and out because he was from the middle. And so but he’s doing this, you know, protein only or, or carnivore diet or something. And I was like, you know, and so he’s telling me about the lettuce wrap. And they told me about, you know, the protein silent, he’s telling me about the Flying Dutchman. But at in and out there’s three different ways to order that protein style. There’s the letter, you can call it the little trap, you can call it protein. So is that all built in? Because I know when I program pls because that’s my day job. It would just be protein style. So when Ben walks up and says, Hey, can I get a lettuce wrap burger? I’m going to touch the button that says protein style. How does the how does the system learn that protein style means the same thing as lettuce wrap and build that model? So that it knows letter Trump says is there something some web behind that you guys have that says this is a look like like my Google Photos? Or my you know, this looks like this kid or this looks like that kid on my Google Photos? Is that kind of how it works?

BB

Ben Brown

19:17

In a sense, yes. What we do is we you know, we meet with the brands to talk about all of their menu items. And if there’s a new menu item that our AI isn’t already familiar with, then yes, we will do that look like comparison. And so, for example, we’ll take pizza, you know, you can say pizza, you can say pie and they the AI will understand that the word pie will mean pizza. Whereas, you know, with your lettuce wrap example yes, it’s a very easy discussion, where we will you know, get the menu items from from the brand, what are the regular menu or secret menu or you know, special limited time items, and we do that look like analysis to find any and every possible way that someone could reference that item. So that it does, you know, funnel into the the actual POS selection, which the AI, you know, submits on the restaurants behalf

JJ

Jeremy Julian

20:19

Well, and again, I it’s gonna get better and I just you know back to the learnings, it’s like the build your own, you know, I mean, we’ve all been to a mod or you know Blaze type pizza and you’ve got the meat lovers and you’ve got the, you know, all of their names that they have for them. And I can say meat lovers with extra pepperoni and no cheese, whatever, you know that that is relatively binary. But then you go to the Build Your Own the build your own burger, the build your own pizza, the build your own salad concepts, they turn into a much more complex, complex set of rules that you’ve got to build, whether you’re building it in the POS, or you’re building it into an AI solution, like you guys have, which I think it’s just it’s, it’s amazing. Talk to me a little bit about accents. Because, you know, you talked about this 100 store chain that might might have some people in the south and people in northeast, I was at a recent customer engagement with somebody from New England. And we were in Texas, and we were in South Texas, and you know, the guy Oh, where are you from? And he said, Rhode Island, but the way he said it, the guy’s like, where’s Where are you from? And like, didn’t understand what he was saying? Because he just jet legitimately didn’t understand what the guys, the guys accent? How do you guys normalize for things like that? Because you know, it? It’s hard. I mean, it’s sometimes even as, as a human being understanding somebody that might have a thick accent, whether it’s foreign, or it’s even domestic. That’s, that’s tough from time to time. Talk to me a little bit about that.

BB

Ben Brown

21:46

No, yeah, no, it’s it’s a very top of mind thing that, you know, we’re happy to have accounted for where we digest so much information, that we have the ability for the AI to learn from it, right. Because we’re operational in 46. States, we’ve heard just about every dialects that you could have in the US. So, you know, when you’re taking in, what, more than a million conversations a month, and they’re geographically geographically distributed in the way that they are. We have swaths of data that we’re already pulling from, so that the AI can correct itself from, you know, errors that it may have made due to miscommunication from an accent. So we really have that that data set, or that that data Mote on our site? Where, because we’ve been there, done that. It’s something that, you know, we’ve we’ve sort of taken care of, and it’s a really, really hard hurdle to get past.

JJ

Jeremy Julian

22:45

Yeah, well, especially because every one of those permeations of the southern accent from Alabama, with the pie, and the way he says, pepperoni might be different than the guy in New York City, and how the guy from Long Island is gonna say that, and so yeah, really, really critical to doing that. Yeah. And

BB

Ben Brown

23:01

I want to, I don’t mean to skip ahead, if you had this question, you know, lined up for later, but a lot of people ask themselves, how come I can’t order through Alexa? Or through Siri, or what have you? And you know, how come the big tech companies like Apple and Amazon and Google, like, how come they didn’t just seize this instantly? And, you know, the services like Alexa and Siri, they’re great. I got to meet the Siri co founders, actually, before they were, you know, the people who found that before, you know, Steve Jobs approach them. It was a it was a great conversation that they led. But it’s, it’s excellent for general discussion. The reason that companies like Converse now are thriving, is that we’re working in a very niche space, that will be very difficult for the big tech companies to focus on given their you know, just their sheer size, where they would have to develop entire teams, or even a spin off brand to deal specifically with restaurants. The AI from a Siri or an Alexa gives you that base layer, right of like basic conversational utterances. But there’s not a lot of restaurant specific language. And given how every restaurant is different, and every store could potentially have different hours, different promos, etc. You know that that is not something that the large tech companies are necessarily eager to cater to on such an individual basis. It’s much more of a one size fits all.

JJ

Jeremy Julian

24:40

And I think that back to guest experience if it’s not going to be done well, we’ve all had the experience with Alexa. We’ve all had the experience with Siri where it’s like I don’t understand what you’re saying. And you can’t do that. When you’re a consumer facing brand that is selling something to somebody. Siri is not likely Apple’s not selling me something likely when I’m looking something up on Siri, I’m looking for directions, I’m looking to text my wife, I’m looking to figure out what the temperature is going to be. I’m not necessarily looking to go buy a MacBook, I promise you if they made it to a point where I had to buy a MacBook, and you know, I was spending two grand on there, they’d figure it out. But they’ve chosen not to. And they oftentimes will say, whether it’s Siri or it’s Google Voice, or its Alexa, they all say go to the web, go open your browser, because you need to continue this year, which gives you the ability to have a multi threaded conversation. Talk, let’s talk about opt out. So I’m now in the drive thru. I’m, you know, I’m not having a great experience, I’m not able to get what I want. You know, my wife, I, again, talk about having four kids, there’s a nationally recognized chicken brand that is closed on Sundays, that has a meal that one of my children will not eat, unless they talk to a person to order that meal. So she can’t use the app, they have a fantastic app. She’s good with driving through drive thru. But she can’t do it on the app, because it doesn’t allow her to order the food the way that my son likes to have it. And again, I’m not picking on Chick fil A, because he’s probably in the point 0000 1% of people that want what he wants, but she can’t go to the app, if my son is eating with us. When we go to Chick fil A I’m not picking on Chick fil A, but that happens. And so in your world talk to me a little bit about if I’m in that drive thru, and we we noticed that it’s going poorly, the I noticed is it’s going poorly the guests notice is it’s going poorly. How does that interaction change back to a live human being talk to me a little bit about that? I know the answer to this, Ben, but I want our listeners to understand that it’s not 100% of the time and you have to do it, no different than McDonald’s learn really early on, when you only allow people to order through the kiosks, they get really pissed off when you force you know, especially Americans to do it your way. They don’t like it. And so, talk to me a little bit about about how you guys get around that opt out, I don’t want to use the AI to deal with my ordering today.

BB

Ben Brown

27:03

Of course, with guest experience being the most important thing for us, we always want to make sure that guests have the opportunity to speak with a person if they want to. Just as you very well know that there are times when someone just prefers to speak to a person, you know, whether it’s age or geography that influences that or whether, you know, they’ve had a bad experience before, where they just saw the movie AI robots, you know, it, there are a number of reasons. And, you know, it’s no hard, no hard feelings on us, we don’t take it personally, we know that, that there’s a subset of people that want that human touch, and we always give them the opportunity. So you know, there, if, if the guest wants to speak to a person, you know, they can say that we’re the agent or something like that, and that’ll direct them to a human being that they can talk to that will serve them as needed. So that your your son could have his custom order at Chick fil A, just the way he wants it, you know, I’ve been there too, right? And then for, for, for just ensuring that smooth experience, right? It’s, we recognize that AI, while it’s exploded in the media waves, there’s still some time before it really becomes that mainstream default way of, you know, taking orders at restaurants. I mean, it’s eminence, but there’s a process that that needs to be followed. And right now, you know, we’re looking at those early adopters and those innovators that are embracing AI as an ordering method, and we just let the results speak for themselves. What we’ve noticed is that maybe someone might request a person the first time they talk to AI, but then they go back in order again, and they’re like, oh, like, I’ll try it this time. Oh, you know, that wasn’t so bad. And then the next time, like, you know what, that was actually better than when I ordered with a person, it was quicker, my food came out faster. My order was very accurate, you know, nothing was wrong. And okay, and then it keeps going until all of a sudden, that person really prefers the AI over a person because those results are speaking for themselves. So we just let the, you know, we let our products deliver the results and that great experience people are looking for and you know, the the adoption curve follows.

JJ

Jeremy Julian

29:38

Well, I think Restaurant Brands and you know, branded consumer brands are going to have to decide how do they want to let that happen. I know years ago when when Starbucks first release their app, you know, almost everybody’s still stood in line and wanted to talk to their normal barista. Now, you know, nine times out of 10 when I walk into a Starbucks, nobody’s sitting in line. Everybody’s standing there waiting for the barista to make the make the drinks that they ordered already ordered on their app, they don’t actually want to talk to you human being they voted on their app, and now it’s gone. Because they to your point, it’s been better, faster, smarter. What does it what does the AI do with conflicts? Something goes out of stock? Something, you know, is is on the menu, but now they don’t have it? How does it deal with that, and I just don’t know, this is a little bit of my curiosity. I ordered this, you know, this, this pizza, and I asked for olives, and I just ran into olives on the pizza just before that, and the AI is in the middle of taking that order. And it goes to submit it to the POS and, and recognizes that we no longer have office or, you know, whatever it whatever that might be.

BB

Ben Brown

30:40

I mean, in the sense of AI being just like a team member, it reacts to that situation in the same way a team member would in that it receives that signal saying this item is no longer in stock, and it will kindly you know, share with the guest, hey, look, we don’t have that item in stock, can I interest you in something else, and it might suggest a similar menu item. It’s the same as when a guest will you know, just try to mess up the AI. Let’s say that you’re at, you know, a Mexican concept and someone tries to order sushi, which clearly, you know, the restaurant would not sell, you know, it would say sorry, I don’t carry that item, or that item is not available, can I interest you in something else? So I’d say relatively simple just trying to emulate, you know, what it really is like to have a physical team member to the greatest extent.

JJ

Jeremy Julian

31:31

Last question before I want to start talking about the statistics and what you guys have seen as far as adoption curves. have you guys gotten to a place where you guys are, you know, drive thru is is obviously huge. But I know, I know from my own personal experience, the takeout ordering process, especially at anything other than a fast food restaurant, and quite frankly, all takeout ordering experiences over the phone, I am still one of those people that from time to time, I don’t have seven, I’m gonna have 700 apps on my phone. But I only have 700 apps on my phone because I’m, I’m in the industry, I don’t use them. Oftentimes daughter because I’m driving home from baseball practice, I want to order food from my family on my way home. And it’s awful. It’s really bad. It’s you know, in a lot of brands, they haven’t resourced that appropriately. And so you’ve got it goes to the front desk, and then the front desk team member transfers into somebody you’re sitting on hold, by the time I get the order in, I might as well just drive it to the store and order the food, oftentimes. And that’s been my guest experience. And so it prevents me from doing that in the brands that I have that experience. And so, to me, a phone call in order is not a whole lot different than a drive thru order. Timing is obviously different, because I’m not two or three cars back but getting an order into the system through AI and, and whatnot. Have you guys started to tackle that yet, then

BB

Ben Brown

32:53

I would actually say that that’s the first thing that we tackled before drive thru. So we’re an omni channel platform, we serve phone and drive thru, in addition to, you know, really any other voice ordering channel that you can think of. So phone is, you know, we have some very prominent film customers, those that I can publicly talk about, one is Domino’s Pizza. And you know, we’re all across the country with them. And the same thing with Wingstop. So both those brands phone is a very, you know, strong source of revenue for them and a very helpful one. Yeah, and, yeah, phone isn’t necessarily growing at the same rate is, you know, at inap ordering, but it’s, it’s a really prominent channel. And a lot of people just like you and me really want that experience of calling up a restaurant and placing your order, and customizing it in a way that maybe you can’t find online. So what we do is, you know, we configure our AI to, you know, really solve that pain point for the guest. And it really goes that step further. Where, look, you’re calling to place your order on a Friday night. And that in your call is not getting picked up, you’re hearing busy signals you’re trying to call back 234 times. And beyond the busy signal. You know, maybe people just aren’t picking up, they’ve got too much going on, or they just don’t feel like it. Or maybe you’re trying to call in 10 minutes before closing and you know, a tired team member is not going to want to take that order because they just want to go home. You know, we automate that entire process. The AI gives you all the time in the world so you don’t feel like you’re talking to someone who’s super rushed on the other end where, you know, they have to get back to you know, making burgers or pizzas or what have you. The AI will, you know, take your, you know, fundraising code or your coupon code or whatever it might be. And it just really, you know, gives you that experience that you’re looking for. Picks up after the first ring And right around the same time you, you get to speak with a person if you want to as well. And we do that for, you know, QSR concepts as well as, you know, full service concepts where, you know, maybe it’s, it’s a higher end group that you want to order from. And yeah, we take their phone orders to,

JJ

Jeremy Julian

35:23

yeah, and I was just on the phone with somebody else’s trying to solve a different problem earlier today. And he and I were talking about somewhere close to, to, you know, some very large percentage, more than 50% of orders that come into restaurants don’t get answered in an appropriate amount of time. The statistics that he gave me was pretty significant. And if I’m calling for a revenue opportunity, you got to figure out how to solve that problem. And so I guess where you’re able, Ben, talk to me about the success that you guys have seen, you know, statistics, you know, throw throw some, you know, brag on what you guys have done for a few of the brands, I know, you alluded to Domino’s, and, and Wingstop, just a few minutes ago, but I know that you guys are focused on solving this problem. You guys continue to to build momentum and solve challenging issues. Staffing is hard drive thru, you know, and quite frankly, people don’t want to sit there and deal with the guests over and over again, they’re fine to make the food and put it in the bag. But like that same, that same repetitive number one, bringing out the point of sale, automating that, and getting them out of the processes is something that can add value, even to the staff members, and allow you to have them do things that are guest facing that aren’t, you know, this bad, bad experience. So where are you where you’re able to talk to me a little bit about what you guys have have solved? You talked about a million orders a million orders that you guys processed in the month. But what else?

BB

Ben Brown

36:46

Yeah, so we also break it down into three different categories. And we’ll start with the one that probably the majority of your listeners are most interested in. And that’s bottom line impact. So average ticket is a huge thing that we that we like to focus on. And I’m happy to say that we increase average ticket upwards of 25%. With a number of the stores that we work with some it’s higher than 30. And there’s a couple things that happen there. Namely, the AI will upsell 100% of the time. So it doesn’t matter how busy the restaurant is, AI has all the time in the world. It takes limitless calls at once it’s programmed to upsell. It’s programmed to upsell based on whatever the brand specifies. And it has a dynamic upsell. So it’ll upsell you on the most appropriate item based on your original order. So if you’re ordering, you know an ala carte entree, then it might upsell you on a side or a drink. If you’re ordering a combo, it might upsell you on a dessert. If you’re ordering, you know, an item that doesn’t necessarily pair with either then it might upsell you on, you know, like an akuna mall like, you know, a dipping cup or something like that, where it will upsell tastefully so that people don’t feel you know, turned off by it.

JJ

Jeremy Julian

38:11

Or you’re not you’re not a vegetarian, and you’ve ordered nothing but vegetarian. And now it’s like do you want the 16 ounce ribeye, it’s like no, come on. Yeah, right. I don’t want the 16 ounce ribeye, I just wondered, you know, three salads. But

BB

Ben Brown

38:25

yeah, so it’ll it will do it will employ this upsell tastefully. And we have recently seen a very significant number of people converting on those upsell offerings, which is great, and it’s driving ticket or driving the average ticket up pretty substantially. So that’s one item where we’re really driving results is then that average ticket. The other of course, is with regard to order counts. So because the AI takes limitless orders at once, number one, it’s taking orders that might have otherwise gone missed. So in the in a phone environment, it’s those calls that were you know, that guests got fed up, they didn’t call back or they hung up early before somebody could take their order. And now those orders are getting converted in the drive thru. You know, maybe speed of service is faster. So you’re getting through, you’re getting more people that are going into a line that would otherwise have appeared too long when they were, you know, driving past the the restaurant. So during those critical peak hours, we’ve had restaurant operators tell us that their orders per labor hour are up nearly 2x Wow, ouch. Yeah, it’s it’s big. Because you look at the efficiency that drives all of a sudden, you have people who are just making orders, there’s no more. There’s no more taking orders. They’re focused solely on fulfillment. So someone who would otherwise you have to spend five minutes on the phone with a really picky customer. Now they’ve just cranked out half a dozen orders, you know, just throwing them in the bag to just drive that volume. And you multiply that on a national scale. And those numbers add up pretty substantially. So we’ve we’ve seen that our AI is redeploying a significant amount of labor. And when I say redeploying, what I’m getting at, is that they’re taking PPE, most restaurants don’t have like, a lot of them don’t have dedicated order takers, right? These are people that are juggling multiple tasks. So now they’re juggling one less item. And now they’re be they’re able to devote that time to more productivity, so reallocating that labor in a way that’s really allowing restaurants to drive their sales.

JJ

Jeremy Julian

40:50

Yeah, well, and I think, you know, one of the things I’ll mention, and I think a lot of people start here, oh, you’re gonna throw chaos, you’re gonna put a bunch of people out of jobs? And the answer is, is one, all of our restaurant operators know, it’s hard to hire people right now. And number two, most of them don’t want to do this job. You know, they’re not in the business of doing this job, they’re happy to go deliver the food, they’re happy to make the food, but they don’t necessarily want to deal with this portion of it. And so figuring out how to solve that is is paramount, whether you’re doing it through a digital ordering platform, or you’re doing it through an AI platform, or, or the combination of both because your statistics about the increase in revenue is commensurate with other digital ordering platforms that that better there. When you have that at all, whether it’s on a website, or it’s on, you know, it’s on the kiosk, those same statistics happen, which is, which is awesome.

BB

Ben Brown

41:39

Oh, absolutely. And you know, going into that staff experience, right? One, you’re looking at a lot of people in the younger generations that are working these jobs, and that generation, for better or for worse, really doesn’t enjoy a lot of human interaction as much as maybe people are otherwise used to. So you know, having the AI take care of that. So they don’t have to, you know, it makes their job more enjoyable. Not to mention, just having to juggle one less thing in a hectic restaurant environment, it reduces stress, and it dramatically increases job satisfaction. So we’ve had operators tell us that their staff retention is substantially longer than it was before they got AI. And some have gone so far as to say that team members are leaving other franchises and joining, there’s because they have AI, and they know that it’s just going to be an easier job for the same pay. So it’s a great recruiting tool. And then as far as the guest experience goes, we measure customer satisfaction. And we’re happy to share that. Look, our CSAT scores are exceeding the QSR industry standard, in many cases by a very large margin. So it’s showing us that the guests are really happy with their experience. You could say that they’re happier with the experience than they were with regular team members. So you know, right now, between the guest, the staff, and the bottom line, all signs are, are pointing in a positive direction.

JJ

Jeremy Julian

43:09

All right, I’m sold, Ben. Now, how do I get engaged? Tell me Tell me a little bit more like what? So I’m now a restaurant, I’ve sat on this call, I’ve sat and watched us sit here and talk for 45 minutes. How do I engage with your team? What does it look like? What does engagement look like? And what can I expect as a restaurant operator? You know, whether I’m a franchise of a national brand, or I’m an owner, I’m an IT guy. What does that what does that look like to start the dialogue with you guys? And what can they expect?

BB

Ben Brown

43:35

Yeah, we’re you know, we’re always eager to support more more brands, we want nothing more than to just serve the larger restaurant community and really shape the future of restaurants. And for those interested, you know, just contact us our website is Converse now.ai, you’re welcome to contact me directly, if you would, like I’m Benjamin, dark brown@converse.ai. I can put you in touch with the right people. And it’s a pretty easy hands off discussion where, all right, we talk about the way that you measure success, we want to make sure that we’re aligned, and then you just share us your menu, we, our tech team, you know, make sure the program the AI to get to know your menu. There’s a relatively hands off onboarding process where, you know, we just we test out the AI we make sure it works well. You know, we choose a couple of pilot locations where it goes live in a controlled environment. We hold your hand every step of the way, right. And then, you know, once it’s live, it’s live. There’s really no staff training required. Depending on whether your phone whether your revenue is driven through phone through drive through through both, right, there’s just some setup that takes place there. But it doesn’t disrupt your day to day operations. Aside from the fact that AI is taking your order you know If any additional hardware is something that we take care of, it’s not like, you’re gonna have to install a new base station or get a new telephony provider, we work with virtually all of them. So it’s really as simple as saying, Hi, you know, learning more about our platform, our training the AI on your menu, you know, testing it out, and then it’s off to the races.

JJ

Jeremy Julian

45:21

Yeah, and I know, I, in our pre show, before we jumped on, I was playing around on the website, and there’s a lot of examples, a lot of the stuff that we talked about today is represented on the website. So good job on the marketing and customer experience side, letting that letting that happen, you know, I mean, and again, it for a lot of people they want to go, they want to go check that stuff out. And so, I would implore you guys, if you guys are interested in this. And in general, these guys are amazing. And they focus on this vertical, there are other competitors of what these guys do, if you’re not thinking about it, or already in the middle of implementation, I promise you, your competitors are, so figure it out. Because you know what, at the end of the day, these you know, whether it’s order accuracy, guest guest satisfaction, it’s the the growth in revenue, it’s going to continue to be something that is going to be a point of differentiation for your brand, if you have it. And if you don’t, you are going to be left behind and ultimately they are going to win. And I’m not just saying that because I just I know what’s happening. I every time I talk to a Restaurant Brands, I asked them what are they doing to solve these problems? Because if you are not solving it, somebody else is solving it. And they’re getting that order, oftentimes where you wouldn’t be.

BB

Ben Brown

46:33

Yeah, Jeremy, very well spoken, I really don’t have anything to add to that don’t get left behind. Just on the note of competition, you know, whether it’s Converse now or another group that does voice for restaurants, you know, we all aim to lift up the restaurant business and make restaurants, you know, more operationally efficient and financially sustainable than they’ve ever been. The only thing I would I would want to instill on anyone thinking about AI, are the our understanding, essentially real AI versus fake AI? Because with the explosion of chat, TPT yeah, there’s 1,000,001 groups out there that call themselves AI experts in the same way that, you know, anyone who has an Instagram account is a self proclaimed social media expert. What we’ve found is, look, this is coming from a group that spent, you know, five years building our own AI from scratch. What we found is a couple of groups are using call centers, you know, like overseas call centers based in the Philippines, what have you, with live human agents monitoring every call, and they’re entering the order manually. And then that order is shared with the guest through an AI voice. So the guest feels like they’re talking to AI. But it’s more of this wizard of oz kind of effect, where there’s a man behind the curtain, who’s actually pulling all the strings. Now, at the end of the day, the guest gets served. And that’s great. The reason that a lot of groups do this, or one, they don’t have the AI technology, but to it shows very well. In other words, during that really critical pilot phase, when you’re live and maybe like a couple of stores instantaneously, a quote unquote fake AI company will demonstrate immediate results. Whereas real AI, look, there’s a training process that takes place, it’s going to be good, but it’s not going to be perfect, out of the gate. So, you know, when it’s all it’s all excellent, until you say, Okay, next week, I want this in 10,000 stores. And if you’re with a vendor that doesn’t have the real technology that will simply be impossible to scale.

JJ

Jeremy Julian

48:50

And how do they find that? I mean, I know I know, we’re getting close to time because I figure that out if I’m if I’m a restaurant brand, and I you know, because, again, nothing against what you guys do, because I personally have played with it. I know it’s awesome. I’m glad to have you on the show and share the good news of what you guys are doing. And at the same time, I promise you, a director of it is going to have to go praying in two or three options to a CEO before they move forward with it. He’s not gonna be able to just go, Oh, these guys are the best and you should go with them. How do they figure that out? That people are doing this? Because, yeah, well, it makes sense. People are lying about this crap.

BB

Ben Brown

49:25

And I only bring it up at the end because you know if someone is interested, right, it’s a really important thing that you need to flag and the easiest way to flag quote unquote, fake AI is consistently long lag times. And that means that almost every time after you say something to the virtual assistant, that it will take, you know, five plus seconds for the ordering assistant to get back to, which means that’s the time it’s taking an agent to type everything appropriately in the background. So every once in a while real AI will take some time to process an order especially if it’s a really long order. But if it’s consistently happening, then you know if it feels fishy, it probably is

JJ

Jeremy Julian

50:09

probably, I would have never even considered that I know that was in my notes and we never got to it because I was so geeked out on the other conversation. But that all being said, Ben, I’m grateful that you came on. If you haven’t already subscribed to the to the podcast, please do. So if you haven’t gotten the check these guys out. I’m excited about what they’re doing. I do think it’s going to continue I shouldn’t say think I know it’s going to continue to change the way that we order within restaurants. And so I’m here cheering you guys on I can’t wait. So we have our first mutual customer where we can go go do that stuff together. That’s a totally different deal than that than the podcast but but again, to our to our listeners out there guys, we know that you guys have got lots of choices on how you spend your time how you spend your energy. So we appreciate you guys spending time with us to Ben, thank you so much and to our audience make it a great day. Thank you.

I

Intro

50:59

Thanks for listening to the restaurant technology guys podcast. Visit restaurant technology guys.com For tips, Industry Insights and more to help you run your restaurant better