The Rise of AI in Self Storage - Self Storage Explained with swivl ai

You service more self storage sites than even exist in the entire of the uk.

That is wild.

Welcome everybody. Thanks for joining our latest edition of our podcast. We're joined today, very exciting episode. We're joined by Mason Levy from Swivel. Mason is the CEO and co founder of Swivel in the US and we've met a good few times in the US things. So delighted to have you here. Thanks very much for joining us, man.

Stoked to be here, Gavin. It's good to see you even if it's over a digital screen. Yeah. And excited for today's conversation.

Cool. Mason, you set up swivel in 2017. I'd love to hear. Because when we've talked in the past, we've talked a little bit about startups.

And.

The challenges of that. I'd love to hear how Swivel started.

Yeah, I appreciate that. It's always fun to talk about the origin story. I, I think everybody's origin story is kind of a roundabout way that they became who they were supposed to become at some point in time. For me, it probably started when I was like 13, you know, I was obsessing over, well, even probably before that. My father owned his own architecture business and ran his architecture business out of a home office for as long as I can remember. Growing up, like work was going to your home office and working in front of a computer, uh, you know that, that was the, the example that I had from my dad and you know, I fell in love with technology, specifically consumer technology, and loved watching, you know, Steve Jobs, you know, get on stage and launch whatever the latest greatest was at Apple events. And so coming out of school, you know, I was actually working for Apple and realized like, even though I was working for a company that I absolutely loved, that wasn't the, the path for me. Ultimately I really like enjoyed building, as Elon would quote, the machine behind the machine. So like the actual thing that allowed for the machines to grow faster or grow smarter, respond better. And you know, when Rodolfo, Kyle and I kind of put our heads together, we were like, what are we going to build? I think there was two big hairy problems that we wanted to go solve. One of them was how do you make these machines continually learn? Another one was how do you control the response mechanisms? And those two kind of core problems are still huge problems today. So, you know, eight years later, as much as changed in AI, the same problems are still there. Although, you know, there's a lot more tooling in the market to allow for these things to be A little easier and you know, fast forward, you know, ultimately we got connected with a marketing firm. We were selling to everybody that would buy the service and a marketing company actually introduced us to extra space storage. And so we worked really, really closely with Extra space in the early days to build a purpose built machine learning model that helped get people from their marketing website to their conversion pages. That was the big KPI and it worked really, really well. Fast forward six years and we're powering 4700 facilities in the self storage industry across the the world. You know, we did seven and a half million conversations in 2024. We're already surpassing 10 million this year. So we're, you know, set to triple that number.

You service more self storage sites than even exist in the entire uk.

That's wild.

It's an incredible stat. So congratulations. And what's been the hardest? Like, so like we'll get into swivel in a second. But like, what would you say has been the, the biggest challenge for, I don't know, setting up a company? You're starting from 0 in 20, 20, 17, 28, 2018, getting your first 12 customers. What's the biggest, looking back now on kind of seven or eight years, what has been the toughest part of it?

I love that. Let's, you know, rack my, my brain, I think the interest. So one of my favorite quotes, this is a Reid Hoffman quote. The, the CEO of LinkedIn or the founder of LinkedIn, he's not CEO anymore. As the greatest entrepreneurs know how to let fires burn. And I've really like leaned into that my entire entrepreneurial career. Like there's going to be challenges that always do. And I think, you know, the idea is that like as long as the problems are changing, you know, and you're progressing forward, that's a positive thing. And so, you know, early on, I mean, it seems crazy today in the world of ChatGPT and everybody talking to chatbots all the time. But like early on, I mean, every turn around the corner when I tell people that we're building, you know, the way that our AI manifests itself in the early days, we obviously power multiple channels today, including voice, email, SMS was a web chat bot. And everyone and their mom would tell me that nobody wants to talk to a bot, you know, like nobody wants to talk to these things. And so I think like that was the biggest hurdle early on was just convincing people that this was a something that was a value add in this business. Right. Like you have customers across the globe that, you know, they Want instant answers. They don't care if it's slightly robotic at the time. Like they're going to go to this chat ecosystem because, and again, you know, I'm not as familiar with the, the UK market, but you know, at least in the States, traditionally these operators are, you know, real estate investors. We're going to put a billboard up right next to our facility or there's going to be a phone number on the, on the facility as job buy it. And their, you know, acquisition strategy was that location based thing. They're gonna, they're banking on somebody walking in or making a phone call. Right. And so the idea that like, oh, they're gonna go to a website and talk to this AI system on a website in order to purchase was just kind of like you had to get people over the hump there. I'd say that's like probably the biggest like market, market infiltration challenge that we had early on was just convincing people that this was a viable thing, that that was going to happen. Obviously there's a lot of technology challenges, you know, obviously there's always sales challenges that, that arise. But I'd say that was kind of the big one, was just like, is this something that is going to have a stay, is going to have staying power not just in the self storage industry, but across other verticals as well too.

So, so for, for anyone listening to this who doesn't know what Swivel does, what, what's the, what's the elevator pitch for?

Swivel?

Swivel. What does it, what does it do?

Yeah, I mean we're on a podcast so I'll be a little bit longer than an elevator pitch, but you know, very long elevator. Yeah, you know, we're an AI platform for self storage operators. You know, I like to think of us as an intelligence layer for your entire business. Right. So we're going to power and be the first line of defense for answering the phones, doing outbound phone calls, sending SMS messages, that's for collections for abandoned carts, for review or referral solicitation. You know, we're obviously going to send out emails and as we kind of already touched on, we're going to have a web chat bot there that's going to answer your most frequent questions. So somebody forgot their gate code. You've got an AI agent that's going to verify that this is the person they say they are, that they're in good standing and provides that gate code. And then we also have a few other parts of our platform. You know, we have A progressive web app powered by QR codes or sent out as a link via text message that allows for your most frequent actions that your customers might take to be there. A lot of our operators post signage across their facilities for this, that, that's called Spot Spot also has a review piece of the puzzle so you can land on a really nice landing page, allow for somebody to give you a one to five star review. It does referral mechanisms, whether that be B2C or B or C2C. So Gavin referring me or B2B2C referrals. So a business down the street referring you to it and then it also takes payment, right. So we integrate into the facility management systems payment gateway, give them a really nice interface, do a two factor authentication and then take that payment and push it through the payment gateway. And then on the back end of this we've got a ton of data sources, right. We're integrating into the facility management systems, we're integrating into CRMs. Although we have our own CRM, we're pulling information from Google or pulling information from the website and all of that is training our AI agents on the information that's necessary. And then kind of the last little component is, you know, since day zero when we were trying to figure out who we wanted to be when we grew up, you know, we always believed that AI should be human in the loop at any point in time, whether it be a customer, whether it be an operator setting business objectives or the AI system itself saying hey, I want to get somebody involved in it. There should be a stop or a pause button. And so we've got an entire part of our platform that allows for the human to human connection to happen. And then machine learning models that do agent assist in that as well too. So you know, we're a next generation or a modern CRM, a next generation contact center with you know, kind of reputation management listings features kind of on there as well too.

Wow. So one of the first things you said earlier on was that you know, you were always interested in kind of consumer products interest in Apple and you know, from the way you've described that it sounds like you're, you're taking a huge part of the consumer interaction with the self storage business and kind of taking that on and applying AI and applying your tools and your system to that. Would that be, is that a, is that a fair assessment?

Totally. You know we're a B2B2C platform, right. So we're obviously selling into operators, providing operators the tools that allow for them to do their job more efficiently. But ultimately, the people that are the highest users of our platform are customers. And so we're always thinking about the end customer. Right. How do we make this easier for somebody that's renting and storing? You know, me, myself, I'm a very big consumer of the product. Literally, today, everything in my life outside of a couple of suitcases is in three storage units. It's been there for five months. It'll be there for another nine months. This is not the first time in my life that I've done this. I've moved across the country multiple times, so I'm pretty nomadic. But I'm a big consumer, and storage has enabled me to do that. And so, you know, I build technology that. Or at least experiences that, like, me as a consumer would love to utilize. Right. Like, not that I forgot my gate code because it's my phone number, but, like, you know, it makes it really easy. So if I did, you know, forget my gate code or I needed to make a payment, like, I don't have to sit here and wait on hold for 15 minutes to get an operator to take my credit card.

Yeah, yeah. One of the things you mentioned that I'm really interested to hear a little bit more about is you talked about phone calls, so you talked about outbound calls, and so handling calls outbound, but also handing calls inbound. Is that AI or is that. Is that some other part of a CRM that involves humans?

It is. Everything in our platform is powered, at least as a first line of defense for AI. Now, we also give operators the ability to control this. Right. I think that's one of our big differentiators is we've got a lot of bells and whistles and. And toggles and levers that operators can pull. And so, you know, if. If an operator, you know, wanted to during business hours, when they have humans sitting around or store managers sitting around, wanted to be 100% live human chat, you know, we would bypass our entire AI agent system, but, you know, out of the box, you know, our AI agents do answer the phone. Hey, welcome to the self storage lab. How can I help you today? And, you know, we'll go through and we've got a number of different agents. One's a sales agent that helps with, you know, finding the right unit for you and then passes it off to our, you know, workflow platform to actually perform the move in and the rental part of our process. We've got a billing agent. So when we're doing outbound billing calls, you know, it'll say, you know, I'm here to confirm that. I'm talking to Mason Levy. Yes, this is Mason. You know, hey, did you know you have an outstanding balance of $112? I didn't know that. I thought I was on autopay. And it'll go through the entire nine yards, hopefully ultimately getting you to a place where you're, you know, going to the payment portal to. To make your payment as well, too.

And so I'm interested because I've been listening to a lot of stuff recently about AI call agents and the quality of them and how much they've come on over the past six or 12 months. How have you seen that? Has that been good enough for, in terms of providing that service for years for Swivel, or is this a recent development, the ability for AI to handle that at a high enough quality level?

Yeah, I think there's a lot of nuance to this because, like, the technology has been there for a really long time, right? For a really long time. We've been able to do, you know, text to speech or speech to text and then, you know, do processing on it. Obviously, with GPTs coming into the market over the last couple years, there's just been a boom in these technologies getting better and better and better. And so we've been answering the phone for operators for almost over a year and a half. We've seen vast improvements, you know, and there's multiple parts of this layer, right? Your voice layer is just one piece of the puzzle. The voices that you can go and plug into there are really great, you know, and 18 months ago, it was very, very robotic. You could pay for some premium voices, but it'd be really, really expensive and not probably worth the money. Today that price has come down. Not as much as some of the cheaper voices, but it's come down a lot. The text to speech piece of the puzzle, the, you know, the ability to stream responses. So in the past, you'd have to send an entire packet to, say, the voice. The voice layer, the, the speech or the text to speech layer. And it would have to synthesize all of that at once. And so there's a latency problem. Now you can like literally stream, you know, sentences and words continuously to that layer. And so there's been a lot of improvements even just over the last six months. A lot of what you're seeing in the market is people just wrapping ChatGPT, right, or wrapping Claude or, you know, OpenAI's API. I think the big piece of our puzzle is we were because we're deeply integrated and we've built this huge platform that allows for operators to control this. We've got a lot more ability than just like prompting the voice AI to respond appropriately. Right. Like we're able to contextually prompt it, we're able to cut it off and give it guide rails. So if you start talking about the weather or you know, you want to know what kind of discounts are available, like if it's outside of the purview of what you're trying to accomplish, it will 100% k it out of, of that workflow or of that agent and push it over into like very structured, what we would consider deterministic pieces of the puzzle. And so this is a long winded answer to say like the technology has come a long way recently that makes it a lot easier to do a lot of things, but you still need to have a lot of, you know, infrastructure in place to do it correctly.

Yeah, it's always from when we've spoke in the past, it's always sounded like you've been building swivel with like you kind of the law, like the long term in mind kind of, you know, you're not just kind of wrapping something up, you're, you're actually thinking and it's interesting again just to think about. Early on you talked about your interest in consumer products and consumer experience. You're thinking about the consumer experience. So when you're talking about some of this stuff, you know, you know, if as a customer phoning up a self storage business that's using your swivel.

Service.

The experience that they're going to have on a, on a call there. What, what's the, can you explain the, what's the difference in the consumer experience on, on a swivel product versus maybe something else that's not been built sort of from the ground up in the same way?

Yeah. I think one, from a consumer perspective, it's going to be the depth, right. The ability for this to go actually perform action. And so if I give a couple of examples, right. Because we're fully integrated into facility management systems, somebody that's asking for an account balance, like in real time, we're going to go out and do an account balance. We're going to actually take that phone number, we're going to send that phone number to the facility management system and say hey, is this a phone number on file? Okay, great. Oh, we saw Gavin. We already have all the mechanisms to pull all of Gavin's information into our platform now. Oh, Gavin's past due and his bill is X. Well like now we have the ability to go out and do a workflow. And I think again from a consumer perspective, one of our big differentiators, because we are thinking about the full engagement is we're not just voice. So now we're able to send a text message. That text message links to either the website or to a spot landing page. And our platform is built to be completely multi channel and this is super powerful. We can correlate that phone call to the user that's on your website, engaging with the chatbot and actually engage between the voice and the in the chatbot on the website. So you know, if you take a slightly different example and you say we're doing a sales call, we answered, somebody says hey, you know, I want a 10 by 10. Well, we believe that like actually renting over the phone is not the best experience. Even if you're talking to a human, there's still things that you want to see and make sure that are happening. And so we could send a text message, open up the website. You've got this beautiful website that has a great checkout flow that you've been working really, really hard on. Well, we just want the website chat widget to be the thing that controls the website. So like maybe we pull you to a size visual and you know, from there we confirm over the phone that this 10 by 10 is the size that you want. Well, we might automatically change the website without the customer needing to click a button to that unit mix. Right. Showing them immediate pricing. And so you're really having this true multichannel experience. So there's just a lot more depth to a full engagement platform than just like hey, we wrapped ChatGPT or we're going out to 11 labs or bland AI and you know, we prompted it with you know, a paragraph or six paragraphs worth of like, you know, content around. Hey, you're a self storage manager and here's the rules that you're having to follow. Yeah, the thing that I'm going to really hit on and I don't know how important this is in the UK but like in the states, you know, state to state have different regulations. So like in California how you do collections is different than how you would do it. I'm in Texas right now than in Texas. And so another big piece of this is like making sure that the grouping of your locations across your portfolio also have variants to it. Right. So one, one collection AI agent needs to know the nuances of not only the locations, but the state that you're in so that you're matching regulatory policy too. So again, we go to like reliability is kind of a big word that we go to. And this is not just reliability, the technology, but reliability that the answers that we're providing are factual, that the answers that we're providing are legal, etc. And so I think that those are big differentiators that we provide because I mean, I've been doing this for over a decade, you know, and the 10,000 hour rule, rule. I know the limitations of conversational AI. Us as a team, we've been thinking about conversational AI problems for a very, very, very long time. And so, you know, that multi channel thing is very complex. It's something that maybe doesn't immediately come to the mind. You know, the flashy, shiny new object that is conversational AI and using ChatGPT's voice technology, you know, is not the thing that we're getting distracted by. It's the problem that we're trying to solve. And at the end of the day, to go back to, you know, point that you made, you know, we're obsessing over what that end customer's experience is and not going, hey, you know what, Sometimes the flashy new sexy thing is not always the thing that needs to be used to solve the problem. Like let's solve this problem in the right way and if something comes along that's better to help us solve that problem, then we'll go and do that as well too.

So I'm really interested in your answer to this next question because I think you, of everyone I've spoken to, you're going to give me the most interesting answer. If you think forward three years.

What.

Is the role of AI and self storage and what's left for humans?

There's so much left for humans. I think that's like, let's take it. I'm going to answer the question about self storage specifically, but I do want to kind of like nerd out just for a half second because the. So GPTs were invented by Google in 2017, right? The We've been engaging with GPTs for a very long time. If anybody here that's listening or watching has used Google Docs for more than five years, you know, we were used to pre writing words. It might only been one or two words that it predicted, but ultimately that was a GPT that was predicting the next word in the sentence. The thing that changed from the invent of this in 2017 and like OpenAI coming out with ChatGPT in 2022, I think late 2022 was really just the size of data centers and the size of these models being trained, right? And so not much has changed. And I give that context to be like, and it's actually really timely because of the OpenAI just launched GPT5, you know, within the last seven to 10 days of the recording of this podcast. You know, the idea with a lot of these businesses, if we continue to scale the size of our models, the size of the data centers, the number of parameters that we're pushing into there, at some point in time we might achieve AGI, right? Artificial general intelligence. And for a long time, you know, the last couple of years, the size from the first model at OpenAI to 4o that all of us have fallen in love with over the last year, the increase in size have given drastic improvements to the model. They become less hallucinate, hallucinating, they become able to go out and search the web. They've been adding all these things, but we've kind of plateaued, right, in a way. And you've got two schools of thought. One school of thought is that if we continue to make these things bigger, there's going to be some inflection point where it just becomes artificial general intelligence. And the other school of thought that is ultimately we need a new type of model. So GPT being the type of model, we need an Einstein level mathematical breakthrough that really changes these things in order to, to be there. And I'm more in the like Einstein level breakthrough. I think GPTs are amazing. They have a lot of strengths and a lot of weaknesses and they're going to be a really great tool in our tool belt as data scientists and people that are building technology. But they're only going to take us so far. And so I'd say that to say like, you know, in five years, I don't know that much is going to change. Like, yes, Maybe we're on GPT8 by that time and it's going to be really smart and do some really cool things. But I don't think it's like, hey, reasoning and taking over the world or anything like that. And then to go back to like, okay, well how does this impact operators? Well, like one, I think I've always thought of AI in the self storage space as this is taking the most repetitive task off of an operator's plate so that you can have the most human moments. Right. And I'm a little bit of an outlier. We can use me as an example Since I'm a large self storage customer. But I think, you know, historically most self storage customers are finding storage because there's some big life event that's happening. I mean for me it's just a move, right? I'm, it's not like this, it's an exciting thing and maybe I can that the operator that I'm storing with could spend the time to congratulate me on my next chapter. But like maybe I'm storing stuff because I'm having my first or second child and I'm needing to take that extra bedroom or that extra office and convert it into a nursery. Well, like that human moment, now we have the ability to slow down, smell the roses and like congratulate that individual. And I guarantee you if you do that and you take that time to have that human moment, one, that's what we all freaking crave. Like that's what we want. We don't want a robot doing that, we want a human doing that with us. And two, I'm never going to go store anywhere else for the rest of my life. Right. And you know, you could take that to the, you have a problem, right? Somebody's having a payment issue, somebody lost their job. Like, you know, they're struggling to just get into their unit. Maybe it's not a financial thing, they're just having some problem in your business and like you taking the time to have a human moment. You know, again, if I go back to Apple, I worked for Apple. Align, acknowledge, Align, assure. Say, hey, I acknowledge that you're having this problem line. I understand that having this problem is extremely frustrating. When I've had this problem in the past, you know, I've been frustrated and then assured, you know what, I'm going to do everything that I can do to resolve this problem for you and the outcome is going to be there. Taking that human moment, doing that simple thing, you just, you just created a moment of delight from something that, that was frustrating to somebody. And again that impact is that they're going to tell their friends about the experience that they had, they're going to tell their family and they're going to always come back and rent. So I think there's a lot of human moments that are here. We crave that. Right? I don't necessarily need a human to give me my gate code or to help me make my bill to pay my bill. But if I'm having an issue that can't be resolved really quickly, like let me get somebody, someone that can have empathy for the thing that I'M going through.

Yeah. So AI will free up the time from their petit tasks and make time for the important things which you know, especially as more and more goes into kind of as less things become about human connection, kind of the more you crave it. So you know, you know, in a few years time we'll have more time hopefully because we're not doing these other tasks. More time to create that connection, that bond. I think that's, that's a nice vision for where this goes.

Yeah, there should always, again human in the loop. There should always be an option for this. I think this is going to allow for not just in self storage, but operators across different verticals to really differentiate themselves. Hey, we're not automating this and this and this. We're still human at it. Or it's just going to allow for people to differentiate and have those different experiences and to choose. And I think that's a really great way to think about this. Let's be glass half full and think about how this improves our quality of life and not the opposite way in the sense that we don't want to be doomsday that the robots are just going to come over and like Elon Musk might say, to continue to quote him, that we're just like cats to the robots as they take over the world.

Yeah, let's hope not. Let's hope not.

Yeah, exactly.

So if you're an operator listening to this and kind of wondering where to get started with AI, what's your advice to them?

Yeah, I mean I'd say like number one is, is in a day to day, I mean I'm using you know, both chatgpt and Claude, depending on, on the, you know, tasks that I'm trying to do. Like number one, the easiest way to get started is go open a free account on one of these things and like instead of going to Google, like you've got to retrain your reflex. So like the next time you have a question or a problem, instead of going to Google or opening up a blank word document, like you know, train that reflex to go to whatever that preferred AI is and like just type it in there. Right? You know, a simple example is like maybe you're reviewing a legal document or you're trying to come up with ideas for promotions to get people to move in. Well, like just go have a 25 minute conversation. And I think, you know, in this, like don't just one shot prompt it, right. Like go back and forth with it just like you would if like me and Gavin are having a conversation, right, like, let's dig in, let's poke holes in it, ask it to revise things and just see what happens, right. I think you're going to be really surprised at the output from that piece, Right. I think that's one of the biggest benefits of these things is that blank state problem is almost solved. Like you never need to start with a blank page, right? There's always going to be something there because the systems can help you get there. When it comes to AI in the business operations, again, maybe take your financials. I would pay probably for a pro plan so that those financials are not being, you know, just sent everywhere across the Internet, you know, and upload the financials to one of these systems and ask the questions about how do you increase revenue, you know, how do you get more people onto autopay, you know, and get some ideas from that. Obviously they're going to be partners like swivel in the market to help with like, you know, communications problems. So if you're looking for a modern contact center and a next generation, you know, CRM, you know, we're here to help, you know, answer questions, kind of show you what that is. But there are other tools out there too. So like maybe, you know, this is again, not using even generative AI. This is tabular models, but pricing, dynamic pricing models, right, that are pulling in information from your competitors across the market, looking at what your inventory mix is, what the occupancy's rates are and dynamically changing, you know, the, the rates. I think that's a really great place to start as well too, if that's a problem that you're trying to solve. And so I'd always, I just would start like, hey, what's the thing that is the biggest headache for me and my business? Let's go to ChatGPT. Let's start having a conversation with ChatGPT about it. If it comes to communications or it comes to pricing and increasing pricing, like, let's go find a vendor and a partner, you know, and, and start to talk to them about what best practices are there.

Jason, thank you so much. That's been absolutely fascinating. Your knowledge in AI is amazing. I could learn from you for hours. So I appreciate you taking the time to talk to us with this and share some of your knowledge. And if anybody's listening wants to find out more about Swivel, where do they go?

Yeah, you can find us online. TrySwivel.com We're a startup, so we couldn't pay for one of the vowels. So that's t r y s W-I-V l.com you can email me directly. That's just masonryswivel.com you can find me on LinkedIn. We have presence as a business on Instagram and Facebook as well too, so wherever you're wanting to engage, we're going to have some type of engagement there.

Great.

Thank you Mason Gavin I appreciate the time and thank you so much.

At Stora, our mission is to help our customers build the best self storage businesses. Self Storage Explained is here to try and bring you the best from inside and outside dentistry to help you grow your business. If you want to find out about more resources, go to Stora Co to sign up for a demo or find out more. Or you can go to Stora Co Academy for lots more resources including our financial model and other in depth guides to setting up and growing your business.