“PropTech AI won’t be able to immediately fill the shoes of an estate agent.”
Starline Art / 123rf
PropTech is hardly a new industry; estate agents and other businesses in the property industry have been digitally transforming for decades.
Search portals, such as Rightmove in the UK, started amalgamating available properties and recommending homes over 18 years ago when Amazon was just expanding from being a book reseller to a full-blown reseller site for third-party consumer goods.
In 2019, it’s not surprising that some PropTech companies have already been touting claims of artificial intelligence (AI) and machine learning so sophisticated it can replace a real estate agent outright.
This is stretching the truth. While many PropTech companies might possess highly sophisticated automation software with some AI elements, none so far have established meaningful AI that can replace the human component.
What is AI?
AI in its most basic terms is the simulation of human intelligence processes by software. What it does and how it works can vary depending on the task it’s aimed at. However, AI is much more than automation.
Automation is programming that completes a task based on set rules. This type of programming, while hugely helpful, has been around for quite some time, and is responsive rather than analytical. What makes automation intelligent is AI.
AI requires data, and lots of it. But more than this, AI is about the ability of a given piece of software to analyze and interpret that data to the point where it can then automate processes in more intelligent, creative and flexible ways.
In the truest sense of the word, AI should therefore act like human intelligence, rather than simply executing a prescribed set of inflexible rules.
The progress of Google
Google Search started as a series of automated processes. If you entered a word or group of words into the search field, Google would look for articles using that word or group of words. Over the years, the vast amount of data and recorded interactions has given Google the ability to go beyond word matches.
Fast forward 20 years, and the search engine can now intelligently predict what you are looking for in more complex terms and can recommend more useful content. Beyond that, it can also now guess what other items you might be interested in that are similar to your search request. That’s AI.
In the property market, this kind of advancement has yet to fully take place. There’s plenty of software automation, that’s for sure. There are automated emails that respond to property requests, and chat boxes that can answer specific questions with specific programmed answers.
You can add hundreds if not thousands of questions and answers to a chatbot, but it doesn’t change how it performs; one question leads you down a singular determined path with determined answers.
If, on the other hand, you have robust software that has learned through patterns of interaction what data a new customer might really want to know when asking a question about a property, then the bot can start to ask questions back and make intelligent recommendations beyond the original requests.
Within PropTech, AI could analyze the substantial body of historic housing data, along with estate agents’ knowledge bank of customer queries, to analyze patterns and extrapolate information that would appeal to a customer based on the questions they ask on a property.
A false prophet?
While automation can supplement a property sales process, it has yet to achieve anything close to removing the need for human intervention when it comes to value-based recommendations and assistance with selling or buying a home.
When PropTech AI does become more sophisticated, it will require access to robust databases. Connectivity to a CRM (or a service tied to a CRM), where the entire customer journey from applicant to sold has been recorded over thousands of interactions, will be integral.
This isn’t a simple task. PropTech AI won’t be able to immediately fill the shoes of an estate agent – it will have to first learn how to parse complex data and automatically generate new results that accurately match customer requirements without the need for human input.
The future of PropTech AI
AI can improve and enhance almost all digital consumer experiences delivered by property agencies: it can help consumers get to the right information more quickly and can add significant value in recommendations based on their choices.
But we’re not quite at that stage yet. PropTech vendors are yet to achieve this meaningful level of AI (most are not even close); and automation, no matter how sophisticated, is not intelligence.
Even when the AI revolution does kick into action, its effectiveness will depend on how it gets implemented and whether it works cohesively alongside business models and strategies.
It’s also worth remembering that, for the foreseeable future, the role of the estate agent won’t be constricted by the growth of AI. The UK sales and lettings process is still complex and requires the expertise and personal approach that only an agent can provide.
The foundation of AI is already in place with CRM systems collecting more data along the customer journey. And automation is already helping agencies save time and deliver a better customer experience. As we move more and more to meaningful AI, we will see PropTech solutions start to deliver new and improved agency services for both buyers, sellers and tenants.