How real estate can capitalise on big data, machine learning and AI

Grace Ormsby, REB: AI is obviously a massive part about what is going on in the world at the moment. Let’s start with a simple question, what exactly is artificial intelligence?

Jeff Gray, Propic: I think probably the better question is, what isn’t it? It’s almost obligatory now. If you’re selling tech, in your sales presentation, you’re going to have something to do with AI. But when you scratch the surface, quite often, there’s less intelligence and more artificial.

What it isn’t is a big database or fancy dashboards. You don’t need artificial intelligence to build products around that; you can use it, but you don’t need to. It’s not simply a matter of augmenting or automating workflow, again, you can use AI for certain aspects of it, but you don’t need to use it in those sorts of contexts.

So we need to realise that artificial intelligence isn’t a silver bullet and isn’t in every piece of technology. So then you get to what is it, right?

And AI really is an umbrella term. It’s actually describing a whole host of different technology, and it’s technology we’re using every single day: machine learning, which enables a machine to improve and get better over time; neural networks that are replicating how our brains operate; speech-to-text, like “Hey, Google” and “Hey, Siri”; robotics; the internet of things, even recommendation engines.

It’s a whole bunch of different technology, but ultimately, what AI does is it tries to replicate or mimic certain aspects of human capability, how we learn, how we speak, how we interact with the world.

A real simple, practical example is anyone that’s watching Amazon Prime, it’s equivalent to Netflix, if you touch the screen, it’s going to tell you who’s acting in that screen and what their bio is. That’s quite a humanistic sort of capability to understand who the actors are and a little bit about them. So that’s a really simple practical example of it. But the reality is, it’s all around us, everyone’s using it, every vertical in the world is pursuing an AI strategy, and it’s here in real estate.

GO: So how do you then see it impacting the real estate vertical?

JG: Well, it’s occurring right now. It’s already augmenting customer journeys, and I think Daniel said it very, very well; it’s not there to replace, it is there to augment.

If you think about the buyer inquiries that are happening at 9pm, 10pm, which is when, with our conversational AI, that’s the peak time that people actually want to talk about property, is when the industry’s closed. So it has the ability to take away some of that workload, the inquiry anxiety, from agents, and augment customer journeys across the buy cycle, helping vendors, potential landlords, and so forth. So that’s a really practical example. We know that at least 40 per cent of what you do in managing properties can be automated today.

Think about maintenance and the process of managing maintenance, or chasing rental arrears, or complaint management, all of those sorts of mundane tasks can be taken and automated through artificial intelligence.

And then I think the other big sort of frontier is growing businesses. We, like other companies, are starting to mine data on a massive scale and building algorithms and artificial intelligence to try and anticipate which properties and consumers are most likely going to transact in the future, to tell an agent who to call, when to call, and most importantly, why they need to make the call. So there [are] some practical examples of it playing out in real estate today.

GO: I don’t know about for anyone else, but AI to me does have a lot of connotations of being expensive. What would you say to that?

JG: It is, and it is if you try and do it yourself and build it yourself, and I know. You’re talking, for a senior data scientist, $250,000 a year. You need data engineers, they’re $170,000 [to] $180,000 a year. And trust me, I’ve got a team of 30, we’re a start-up, and we’ve already spent well over $15 million, and spending millions and millions a year. But that’s nothing compared to the Netflix, the Disneys, the Amazons of the world. Think about Netflix, and think about how much money they’re spending on technology, you subscribe for $14. So, companies like ours are making those massive investments, hiring all these really expensive talented people, but then able to bring that sort of Netflix business model to the industry.

If you think about it really practically, today, you can have AI serving your buyers 24/7 for roughly what it costs for a floor plan, and then your vendors pay for it, so you don’t even buy the technology. Or in property management: we did a time in motion study on 5,000 properties on management, and we tracked every single person in that business every single day and every single task that they did. On 5,000 properties, it costs three-quarters of a million dollars to coordinate maintenance if you add up every single task of every single person. So for $1 of property, would you use AI to automate 80 per cent of that? So 5,000 properties, $60,000 a year to save $600,000. It starts to make a lot of sense.

GO: Absolutely.

JG: That’s the cost side, but then you got to start thinking about the flip side: what is the commercial benefit? Yes, there are cost savings, but the other way to look at it is, the human capital that you have in your business is the most valuable resource. What is the benefit of taking away the mundane that AI can do, and enable them to focus on the critical things, which is the people-to-people relationships? It is a people-driven business. We’re not going to suddenly automate and take away agents, or property managers, out of the equation. But we can enable them to focus on the things where they add the most value, which is stuff that AI just certainly can’t do.

GO: I’d love to elaborate a little bit more on that people aspect; because people make up the real estate industry, we’ve already sort of got to the point where we realised that robots aren’t going to take over the world or do any of that. But where do you see people actually fitting in when it comes to not only artificial intelligence, but machine learning as well?

JG: MRI, the software company, they put out a voice of property manager report, and I thought that was really telling. That research is saying that one in four people in property management are considering leaving the industry, one in four. Fifty-three per cent of those surveyed were struggling with some form of mental health issue. And that’s just been published, so it’s real, burnout is real, it is happening.

AI is not a silver bullet, but it can add some value in dealing with these sorts of issues. Because if you can take away the mundane from somebody’s job, and empower them to focus on more meaningful interaction, building relationships with landlords, working with tenants, growing the business, and not doing stuff that AI is really good at, then you’re starting to create more meaningful, more valuable jobs across the entire industry. I think the question is: can we use artificial intelligence to maximise the value of our human capital? And I think the last point I’d want to make is that instead of using the word artificial intelligence, I think the better description is augmented intelligence. How can we augment and free up our people to do the things that they do best? And it just so happens that AI can assist.

GO: Something that has come up a few times in the work that I’m doing here at REB is that machine learning and artificial intelligence, as well as big data, feels like a repetitive process. But we don’t really think about them actually being able to offer bespoke experiences.

JG: I think it sounds like an oxymoron. AI is the enabler of mass personalisation, and again, think about your own context, who’s using Facebook, or TikTok, or the social media platforms? Think about it, the ability for these companies, on a global scale, across hundreds of millions of people to personalise your individual experience on these platforms, that is only possible through artificial intelligence. Yes, there are some dangers in that, and I think they’re taking it so far that it becomes so myopic that you don’t see what’s really happening across the whole world, but that the application of artificial intelligence actually enables mass personalisation.

We currently are punching out tons and tons of emails; think about your new, just-listed emails that go to everyone in your database. Why does everyone get the same email? I’m not the same, I have different intent, and probably a different requirement to you, but everyone is treated the same, because we haven’t had the tools to go, “We understand your intent, we understand your behaviours, we automatically can classify you, and we can communicate in a way that makes sense to you, but do it automatically, and do it on a national scale.” And so, I think that’s the role that AI will have in creating, for the first time, true sophistication in mass personalisation.

GO: Some very good points there. Got some good food for thought, especially thinking as we are heading into the rest of 2022. But I want to move on a little bit now, Jeff, to big data. We’ve already thrown it around a little bit today, but what actually makes data big data, and why is it referred to as the new oil?

JG: So there [are] two concepts with big data; the opposite of big data is small data. And small data would be the average size of a property management system, the data in a property management system, or CRM system, or even across an entire network, the amount of data that’s sitting in those systems would still be relatively small. The way to think about small data is: you could get it into an Excel spreadsheet, you could probably do a pivot table, humans can still comprehend that sort of data set, and be able to analyse it, and do something with it. It’s very valuable, it has lots of use cases, in very practical terms, but the way to think about it is, as a human, it’s small enough that I can actually get my head around it.

Then you get into big data, and there [are] three Vs. The first is volume, that’s just purely the billions and trillions of data points that sit in these massive data sets, so the scale of the volume of data is well beyond any human’s capability to comprehend or manually sort of analyse it and understand it. The other aspect of big data is velocity. Today, literally today, we’ve created more data than the entire time of humanity that has existed, the entire time of the world. Tomorrow, that same thing is going to be true.

The amount of data that we are individually and globally collating every single day, the velocity of that is beyond human comprehension. So velocity matters in big data. And then last one is variety of data. And that is the ability for platforms, such as the ones we’ve got, to harvest data from all these different data sources, join it all together, transform it automatically, and then make something magical happen out of the back of it.

The reason why it’s being called the new oil is, you think about the industrial evolution, you think about electricity, you think about oil itself, the combustion engine, all of those moments in time. And the reason why this term has been around is, data is doing the same thing. It is completely and will fundamentally change how we exist as humans, in every aspect, medicine, property, banking, finance, telecommunications.

GO: How are we already using both big and small data within real estate? Because I feel like as well, as you’ve mentioned previously, it’s already in use, we don’t even probably realise when it is in use.

JG: This is all happening in the industry. And people are already using these technologies and using data to do these things. Developers are using big data and artificial intelligence to identify development opportunities and work out return on investment. The early adopters are already using this sort of technology.

GO: Are there any areas where we haven’t even worked out how we can use this data? Where do you see some of the potential uses coming in the next, even maybe 10 to 15 years? There’s obviously potential there. We don’t know where that’s going.

JG: I don’t think it’s the next 10, 15 years. I think you’re talking 36 months, 48 months.

GO: That’s quick.

JG: That’s quick. And I’ll talk about 2022 shortly. A heavy focus for this industry is going to be on understanding consumer intent and the ability to harness data to do that. And what we’re talking about is anticipating a consumer’s intent to get the right service to the right consumer at the right time. And that’ll be artificial intelligence and people working hand in hand to do that.

Google has got a mantra, which I think is absolutely critical for this industry. And that is: intent beats identity. To understand that, think about the current process if I inquire on a property. “What’s your name? What’s your mobile number? What’s your email address?” Well, guess what? People don’t want to hand over that information. They’re bypassing and going straight to the open for inspection. So the ability to actually start to understand, in a very, very detailed level, the intent of a consumer is more valuable. And Google knows this. It is actually more valuable than their identity.

And the second part of it is that immediacy trumps brand or immediacy trumps loyalty. Think about it right now. If I inquire on a property at 10pm tonight, the probability of me getting a response at 10:01 pm is nil. I may be lucky and get a response the next day. Yet Google is saying, immediacy trumps brand, right? That’s what every vertical in the world is pursuing. So big data and artificial intelligence will enable the industry to start to measure intent and operate in a 24/7 world.

GO: Pretty crazy stuff. Bringing it back a little bit, what are you then expecting within the next 12 months to three years, considering so much is happening at the moment?

JG: Well, let’s just start with this year 2022. You’re going to have more augmented customer journeys: whether for potential vendors, tenants or landlords. So artificial intelligence is already and will become more and more sophisticated.

The automation of big chunks of property management will happen this year. So using AI to automate 80 per cent of what happens in maintenance, complete automation of rental arrears so you’re only dealing with the edge cases. Complaint management, all of that stuff is here and now. And we’ll be using it for identifying which consumers and properties are most likely to transact in the future.

Right now, we can still get an agent to make a call. Well, what’s going to happen in 2022 is an autonomous platform that will contact the consumer on behalf of the agent, instantly create a website about their home, enable the consumer to interact with AI. It will enable a consumer to talk about their own home and book the appraisal for the agent. So 2022 will be the launch of an autonomous, intelligent prospecting platform that’s operating on a national scale, 365 days a year, 24/7.

GO: Pretty cool stuff there.

JG: That’s ’22.

GO: Yeah.

JG: So then when we start to fast forward further into the future, and now I don’t obviously have a crystal ball and no exact timelines. Think about commercial real estate. They already have digital twins. Why don’t we have digital twins of every single property in Australia? I, as an agency, why do I list a property on my website, but it only stays up for when I’m selling it, and it disappears?

Why has the digital representation disappeared? If you think about this industry, the property is the constant, the relative constant. Properties will largely stay the same. Yes, they get knocked down and subdivided, but the property is the constant. The variable is people and their relationship with properties. What you’re trying to do is serve those people and their relationship.

The concept of a digital twin will come in. If we can tell an agency, there are this number of properties in your area, you have relationships with X per cent of them. You don’t know these properties, you need to be targeting these properties. Why can’t you have a digital twin that enables a home owner to get services from you? I’m about to sell my house. Yes, I’ve got a vendor view now. I’m looking to buy a house. I’ve got a buy a view of the property. I’ve sold it to a landlord. They now own it, but I’m also a landlord. Then there’s a tenant view.

Digital twins are coming. It will happen. It’s already happened in real estate, in commercial. Then the other thing is the portals of the future are not going to be what we’ve got today. Conversational AI essentially has the power to become the portal. If you think about a portal today where you get REA or Domain or wherever, that’s just a channel. What if I want to use Google Voice to talk about property, or I’m in Facebook and I want to use Facebook Messenger. Think about conversational AI that can talk about any single property, in any channel, to any consumer 24/7. I think that’s where it’s going over the next few years.

GO: A very exciting time to be in real estate.

What are your key takeaways from all of this?

JG: What I’ve just described sounds scary. I think first is: don’t be scared. I think there are really exciting opportunities. We’re interacting with it every day. My four-year-old is already using Voice for search, and we’re interacting with AI in every aspect of our lives, and the same is going to be true in real estate.

You can learn about it and you can embrace it; I think [that] is probably the key thing. I always say to customers, start small, but move fast. Don’t try and take on the whole world. Just start in one part of your business where it can have an impact, and then progressively move on.

The second thing I think is, it’s not about the future. What we’ve just described and talked about today is already happening. You think about the bell curve of adoption. The innovators, the early adopters are already using this type of technology in their businesses in Australia today. It’s here. They will get a disproportionate benefit. The early adopters will always get a competitive advantage. The laggards, by the time it gets adopted across the entire industry, get no competitive advantage. That’s probably the second thing.

Then lastly is, I think we owe it to our people. I’ve just described the results from MRI. I mean, these are pretty profound facts. One in four people in property management are wanting to leave. Fifty-three per cent saying they’ve got some form of mental health struggle. If we can adopt cost-effective technology, they can take away the mundane, improve the customer experience, make jobs more rewarding and sustainable for the industry, then why wouldn’t we be considering it?