Our COO & Co-Founder, Robert Turnbull, had the opportunity to be a guest on the Apartments On The Go Podcast. On this episode you’ll hear discussion about how it is we are building the best conversational leasing bot in the real estate rental market. To catch the full episode, CLICK HERE!
Apartments On The Go
Episode 30: It’s a Bot Time – Robert Turnbull
You’ve joined the Apartments On The Go podcast, with your hosts, Courtney Smith and Matt Ruedlinger. We’ve got expert advice from the industry’s biggest stars, with insights into trends, best practices, and manageable actions you can use for your careers today. This is the Apartments On The Go podcast.
CS: Welcome to the show! How’s it going, Matt?
MR: It’s going great – excited today, I love this topic, so.
CS: I know, I’ve been really, really jazzed about this one. So I will just kick it off, and introduce Robert Turnbull. Robert has been creating and starting businesses for the last 25 years. And 21 of those years have been in multifamily. I just learned that you started ApartmentGuide.com in 1999 and then Rentals.com right after, I mean the list goes on and on. And he made a comment that I just love, I think we’ve heard this time and time again: “Once you’re in multifamily, you can check in but you can never leave”. I love that! But today we’re going to talk about BetterBot, which is a company that Robert started. They are a chatbot that personally, I just have become the biggest advocate for this year. I have been testing it and implementing it in any community I can – my co-workers probably think I’m crazy. And that’s why I’m so glad that Robert is here. Welcome to the show!
RT: Yeah, thank you so much, and happy Friday. This is uh – thanks for having me on. I’m looking forward to this.
CS: So first, I want to talk about – your bot was named by your clients. I think that’s super cool.
RT: Yeah, that’s a funny story because when we first built it, we tested it with a few clients that I had known for some years, and they’re testing it against a few other bots – this is a couple years ago, and my business partner and I, his name is Zlatko Bogoevski for those who those who – you can call him “Z”. We were testing it, and we were talking with 3 clients in 3 days and every time we talked to them they said, “Well this is such a better bot!” So by the third time we heard it I said, “I think we just need to call this BetterBot!” and that’s how we came up with the name, and thanks to the clients that helped us name it!
CS: I just think it’s so cool.
MR: It’s a great name too. Easy to remember.
RT: It’s got alliteration, too right? BetterBot. BetterBot. We see people actually saying that at trade shows and events. They say, “BetterBot, Betterbot”. So, rolls off the tongue.
CS: So, I’d love to just jump into kind of, you know, what brought you to the bot world. And I want to just kind of talk about – you obviously have been in multifamily for a while, so kind of how you got to the bot part.
RT: Yeah, so I was fortunate to work with one of the largest chat companies in the world, and in so doing I got introduced to the bot technology about 5 years ago. So I traveled around, really four countries. And I loved some of the technologies I’d see on [companies] and I loved what it could do for industry, but there wasn’t one that I saw that was a good fit for our industry. I happened across Zlatko, my business partner (“Z”) again and he had built a proprietary bot platform and I said, “That’s it – that’s the one we’re looking for”. He had built it perfectly for our industry and so that’s the beginning of the story. I can certainly get into some of the consumer behaviors and industry behaviors if you’d like that propelled it.
CS: Please, yes!
RT: Well, in that case. So what was interesting was really two big drivers we started noticing and people often ask, “What is the deal with all these bots? Everywhere I turn and all over the world and websites and even in our industry…” And really that’s coming from two places – the first being from prospective renters. And consumers and, you know, renters want real-time, accurate conversations. They want them, wherever they are in the digital marketplace. Whether it’s on a property website, whether it’s on Yelp, or Google My Business or Facebook.
The other thing we started noticing with consumers is, they want control over the leasing process, and you know what – why not? They have control over buying an airline ticket and selecting their seat. They can even select the meal they want. If you go to a hotel website, you can upgrade your room and you have, people can order from any restaurant and have it delivered to their house. And the third thing we saw with respect to renters is they really don’t want to fill out lead forms any more. Nobody does. They want fast, simple, easy access to the information.
50% of prospective renter leads go unanswered by the property. 20 years later, today, it’s the same stat.
And if you don’t mind, I’ll just jump over to the multifamily marketers. What we discovered with them – and it’s funny because you brought up that I started ApartmentGuide.com (or helped started) in 1999 – back then there was this statistic: 50% of prospective renter leads go unanswered by the property. 20 years later, today, it’s the same stat. All these leads are being generated and people aren’t responding because they’re overwhelmed. That means marketing dollars are being wasted. Multifamily has a 34% leasing staff attrition rate. That’s higher than retail, and the number one reason given is they’re overwhelmed at the property. And the final thing we kept hearing is, “Robert, please do not introduce another technology into our industry that’s going to increase our cost and increase our work. If you’re going to send me something, reduce our cost, reduce our workload, or both”. And that is what led us to bot technology, what we call, “Conversational Leasing”.
MR: I think that’s great, because you hit a point that we hear often, that everyone is overwhelmed and they don’t have time to follow up on things like that. So well, I mean, the world of bots, I want to go into that just a little bit. I know some people are… they want that actual conversation, you know, versus just an auto reply. Can you kind of decipher what you’re saying versus just doing that, you know, “Hey, here’s the question that’s been programmed and this is the reply”.
RT: Right. So let’s start with the concept of “chat”. So chat is human to human. The problem is our industry is, there’s none around – they’re not available. That was always the big miss with chat – how do you have someone on the other end. Even call centers that do have chats still are having lag times and simply don’t know the property as well as the leasing staff at the property. So, chats have always been a struggle in our industry, but bots provide the other side of it.
They are really two kinds of bot technologies. One of them is what we call, “Natural Language Processing”, so NLP for short. That means that people can type in free stuff, free text, and the bot will try to figure out what it’s saying, and it will respond. So it kind of emulates, or tries to emulate, a human on the other end. The other side of that is guided conversation methodology, or “Guided Conversation”. That’s where all the text is all pre-selected. So natural language processing seems really cool. It’s a great, shiny toy, it’s got a big ‘wow’ factor, but it’s got a big problem – it doesn’t work. When I worked with the largest NLP in the world, 3 out of 10 times it would fail – the bot would just get confused. So, cool, but not cool when you have 30% of the time the bot’s getting confused, creating a very frustrating user experience. At BetterBot, we actually created the largest proprietary natural language processing engine in all of multifamily – and we walked away from it in favor of guided conversation. And that’s because with guided conversation, you can actually preselect the intents, the answers are half a second, there’s no confusion, and it guides them down a certain path to get the answers they want very quickly.
So just to kind of bifurcate the fact that there’s two different kinds of bots: one’s NLP, one’s guided conversation. We went the GCM route and our results have been exponentially better as a result.
CS: Yeah, I want to continue to talk about results. Sorry, Matt – did I interrupt you?
MR: No, I just wanted to thank you for clarifying that, because I know that’s the perception with bots, that there are 2 different avenues on those, so.
CS: Yeah, I want to dive in more about conversion, because I think that’s really what attracts me so much to the bot, that, not only the fact that when I’m searching for anything, I would much prefer talking to a bot personally, so I know there has to be people out there like me, who you know, are looking for that as well. But what I also love so much is that, it really gets that touchpoint to where you’re almost – you know it’s a bot, but it’s almost like you’re talking to a person. And so, when you guys were developing that, and kind of talking to those multifamily [clients], what were some of those conversion statistics that they were wanting to see, or that you were hoping to achieve?
RT: Uh, you know, wow – great question. There’s a couple ways to answer that. Let me answer it this way: so I’ll give you some stats and kind of back into that a little bit. When you drop the bot today – cause look, I’ve seen some shiny toys come and go in this industry – and at the end of the day, it’s all about the metrics. Do the metrics work? And what’s the math? So when you take a bot like a BetterBot and you drop it on a property website, and maybe you put it on Google My Business, cause our bot can go on all these channels, and Facebook and Yelp, what happens is, the bot on a particular property per month, has about 200 unique conversations. So, some of them may talk to the bot and then, you know, 3 days later come back and talk again. Well, we only count that as 1 conversation. So the average bot on a property website has about 200 conversations answering about 150 questions. So even if they never become a lead or an appointment, or anything else, sometimes the best renter is the renter you never had to talk to. So it’s already deflecting all these – because they simply don’t qualify, they don’t have what you’re looking for.
But we do see on average, of those 200 conversations, what we call 26 handoffs. That’s where we’re handing it off to a human. Half of those are actually appointments actually being scheduled. So the bot is scheduling on average 13 appointments per property per month with an astounding 90% show rate. I didn’t even know what a show rate was before five months ago, and I’ve been in this industry 20 years, so that’s obviously where people are, for those of you who are in the property management business, that’s where someone schedules an appointment and they actually show up. So we’re seeing a 90% where call centers get 50% and properties get 65. That’s because the bot’s just letting them select their own and reselect their own times for coming. We did, to answer the last part of that, the metrics. So we drilled all that down and did some lease matching with a lot of companies. We said, “Ok, how many leases on average is the bot actually helping to facilitate in a given month for your property?” And the answer was 4, on average. 4 leases per property per month, and you know that 70% of those are after hours, between the hours of 6pm and 8am, so the bot is quite literally working for you while your staff is at home watching tv or sleeping.
CS: Yep, that’s so important.
MR: Yeah, and I think the, you bring up a good point about qualifying ones that aren’t a good fit, because I can imagine how much time that properties spend on that alone, there’s a value that you probably don’t even look at a lot of the time. You’re doing the budgeting and things like that, and all that time, I won’t say you wasted, but you didn’t work.
RT: Well, and you could say ‘wasted’, because if it’s not a good fit, it’s not a good fit.
RT: Don’t you want to know right away before you start getting your very valuable time involved? Let’s figure that out from the start.
MR: Yeah, I definitely agree with that. So, I have a question: we’re kind of talking about, bots, you know obviously deciphering the right candidate and things like that. And with the AI technology, does it look at everything, and if I ask something that’s too much, is it looking at all the steps or the patterns and things like that with everyone, or is it specific to the property, you know, based on its community and things like that?
RT: So, let me throw this up. And look, it took me a while to figure this out – the AI is the term we hear a lot in our industry, “artificial intelligence”. It actually really doesn’t exist anywhere in multifamily today. That’s machine learning algorithms that teach it over time. So we actually had BetterBot.AI once upon a time, and we were like, “We don’t have any AI in it!” Nobody does, there’s no real AI machine learning. There are rules-based, there’s NLP, but that’s in itself not AI either. So, just want to throw that out there, because I get asked that a lot, and you’re not alone. I mean that comes up all the time. What we do is we create each bot and we train it, so some bot tech companies actually have the property train it – I’ve never fully understood that because properties are not adept at training a bot, so we train the bot, we configure it to the property, add in all the calendars, and so forth.
What’s interesting is, we’ve begun to see patterns and behaviors, it’s a natural click-streaming. We’re noticing people don’t always start at the same place – they start at different parts of the conversation, so we’ve optimized against where do people start, and after doing 2 million conversations, we’ve gotten really darn good at figuring out where they start, and where they want to go. So we call that “circular directed logic”. We know where you’ve been, we think we know where you’re going at the next intent, which makes the conversation much more seamless. We’ve had a few folks say, “Wow, this bot almost seems too simple”. Do you know how much time went into making that bot look too simple? I equate it to when people saw AOL and Microsoft and all these crazy sites, and along comes Google and all this white space. And it was just so simple to use, and that’s really what we’re going for at BetterBot is just to make it a very simple experience that you can get through (on average) in 90 seconds if you’re a consumer.
CS: Yeah. I also love the reporting. So, what makes BetterBot so valuable for, just personally in marketing and for my teams is the fact that they really don’t have to do any work until that appointment is made, or until you know, that conversation is actually fed into. So, we use an ILM system, or we use a lead-manager system, so all of the bot appointments and conversations are fed directly into the community’s software instantly. It’s seamless, and I know you hit on this, how it connects to my Google My Business, which is not something I’d say many people are versed on connecting with and figuring out. We’ve watched many conversations get closed just by having that link to make an appointment. We’ve also watched senior citizen communities – this is one thing I love, is that this bot is actually used by a ton of their kids helping their parents find housing.
RT: Yup, yup.
CS: And so there’s just, I just want you to talk about all the ways the bot can be used for kind of our listeners who aren’t in the chatbot world yet, ‘cause I know I could talk about it for years.
So when someone goes to ILS and they click on ‘website’, they can actually go over to, have the bot pop up and schedule those appointments and do all that.
RT: Hey, go for it – you’re doing a great job so far! Go for it! Yes, so different ways the bot can be used, certainly is…Let me just start with channels. I know we often say, “Oh it’s a chatbot and it goes on our website!” Well, yeah, it can do that. It can also actually go on internet listing sites under the, you know, ‘click on property website’. So when someone goes to ILS and they click on ‘website’, they can actually go over to, have the bot pop up and schedule those appointments and do all that. Yes, it can be on Google My Business. I gotta tell you, Google My Business is killing it, I mean, they’re getting a lot of traffic and it’s typical Google, right? And so, if you’re on Google My Business, a lot of people want to go ahead and schedule that appointment right there – well, the bot can do that. It can actually be launched from GMB – don’t even have to go to the website. We work with Facebook again, so you don’t have to go to the site, and you can have the agent 24/7, 365. We’ve gotten quite a few on Yelp and a few here and there on Instagram, so those are growing a little bit.
One other thing I’ll throw out is, we can intercept text messages. So if a consumer texts and says, “hey, I’m interested in the property”, the bot responding can go, “oh hey, I forwarded this to my human friends but I can help you now. Click here”. We actually take them out of text and into our mobile version, our web-enabled mobile version, which is my favorite, and so, they’re having a conversation on their mobile, and a lot of them actually think they’re still on text. And so, it’s answering it so the property isn’t having to do this. And the last thing is email. So we can intercept emails coming in from various sources, and the bot can respond and separate the tire-kickers and prospects. So you’ve got channels and you’ve got mediums, and we can do kind of all of that anywhere, 24/7.
MR: That’s great, and I think it’s also important for listeners to understand that we have changed the way we interact online. So I think, like, you know, some of the hesitation may be that people don’t want to talk to bots, but I would disagree. Just like with the old saying with, on ads. People used to not want to click ads because they thought they weren’t trustworthy – well that’s a thing of the past, so you know, I think it’s important that we realize that we have changed on how we view things online. And so, have you seen anything that stands out, patterns of behavior from the pandemic? Are you seeing any change before/after?
People don’t want to talk to bad bots. But they do like to talk to good bots. If a bot has a successful conversation, it has upwards of a 95% consumer satisfaction score – CSat score.
RT: Oh, wow. Boy is that ever timely. And let me agree with you. People don’t want to talk to bad bots. But they do like to talk to good bots. If a bot has a successful conversation, it has upwards of a 95% consumer satisfaction score – CSat score. The average human chatting with another human has a 65% at best. So the CSat score is way higher with the bot, and why is because it’s so responsive. Pandemic, yes. Wow. Come January 15th we had a client in California say, “hey, is there anyway you can throw virtual tours and videos in the bot?” And we said, “that’s a great idea!”, so we listen to our clients, and so we built it and on March 15th, we launched virtual tours and videos inside the bot. March 15th, when the whole world blew up. So March 16th, we had everyone asking us if we could put virtual tours and videos inside the bot and we were fortunate that we had already helped that [client]. But with everyone scrambling and figuring out how to work from a distance, we saw an increase of 60% of our business grew from March to May of this year because people wanted to figure out, “how do we service in today’s pandemic environment?”
The good news is, I think we’re seeing the end of the pandemic on the horizon, but the good news also is, it really hasn’t changed that we’ve learned that bot technology can help – pandemic or not, as we’re seeing after hours, so. Yes it has helped navigate through this troubling time, but I think we’re learning a lot about what we’ll keep with us after the pandemic’s over, so I think bots are here to stay.
MR: Yeah, I agree, and I think that’s a great thing to point out, because you know, as more properties are going to evolve, it’s going to be important that, you know, for listeners who don’t have that understanding that consumers aren’t going to alter their expectations. They’re going to be expecting and want to see and tour anytime that they want to, and have communication anytime they want to, and if you’re not there, that could, you know, definitely hurt. So I think it’s good to make [people] aware of the changes that are happening.
RT: Yeah, yeah. And a lot of them are going to stay, no doubt.
CS: Well I want to just ask one last question. I feel like you kind of already hit on where this is going, I mean, just the whole intercepting a text message and kind of changing I see that really coming out in the future. I’ve watched texts really take a leap in the past couple years as a method that people do want to use. Do you think there’s anything that is going to come about here soon, or do you have any future things that you can talk about with the bot?
RT: Yes, I do! So a couple of things: I’ll speak on broad terms because we’re going to launch some stuff 1st quarter. One of the things I will throw out is, when we first built this, I don’t think we fully understood (just to be honest) all that bots could really solve. So, we have learned a lot, and in the process we’ve learned a lot of problems that the industry is facing. I alluded to this earlier, which is the overwhelming volume of leads coming from internet listing sites, the ILSs will tell you, “Hey, these are great leads!” and the properties are like, “yeah, I don’t know – I may disagree with that”. Well, how do you know if you’re not responding? “Well we don’t respond!” And so you have that back and forth, and both ILSs and properties are very interested in seeing how bot technology can become an intermediary between the two, and separate the tire kickers from the prospects that I alluded to earlier. That is something we’re calling our “pro-product”, we’re coming out with it in 1st quarter called “Nurture” – so it will have ILS intercepts, it will have follow-ups. We’re going to dive much deeper into the resident portion to help service the resident. We’re launching student today actually. We’ve launched our student bot and we’re going to be launching the industry’s first affordable bot in January. So these are some trends and things we’re going to solve for the industry.
CS: I can’t wait!
MR: You’re going to come back and tell us all about this later, right?
RT: I’d love to.
CS: Oh this is so great. I always hate when we run out of time.
MR: I know, but Robert has an offer for our listeners.
RT: That’s right! Hey, so – and I really appreciate you having us on, and just to thank you and thank all of your listeners we’re going extend our, we’re calling it our “Bye to 2020 special”, so we’ve said, “hey, if you want to bring it on, we won’t charge you any of our bot build fees”. Normally $150, we’re waiving it in December, but to your listeners, we’re going to extend that all throughout January. So, if any of your listeners, just let us know, “hey I heard you on the podcast, I’d like to do it”. Take advantage of it, and we’ll build as many bots as you’d like in January. We’ll waive that set-up fee for you.
CS: I’m telling you guys, you gotta get these bots.
RT: I agree.
CS: Thank you so much, Robert.
RT: You bet, thank you!
MR: And remember I was telling you last time, Courtney, if you like our show, leave us a comment. And we’re on all your favorite streaming podcast channels: iTunes, Spotify, I Heart Radio, and now, Pandora, so hey – we’re getting up there, right? Alright, till next time.