Technology

What the rise of machine learning means for real estate sales

Potential repercussions of automating different parts of the industry
  • Tech and software companies are battling to create AI that will begin to automate parts of sales.
  • A recommendation engine will implicitly know the best time to contact a consumer, the buyer's criteria, their preferred outreach medium and the best strategy and messaging to supply that customer with information.
  • The salesperson whose job can be most automated will be the first to go.

Future-Proof: Navigate Threats, Seize Opportunities at ICNY 2018 | Jan 22-26 at the Marriott Marquis, Times Square, New York

Will artificial intelligence (AI) replace real estate salespeople?

Last year, I sat in the audience at Inman Connect New York at the Marriott Marquis in Midtown Manhattan, listening to Brad Inman propose that we’ve passed the era of big data, into “an age of predictive analytics.

Descriptive analysis (historical data) has certainly given way to predictive inquiry (insight into the future) — but I would also suggest that we have already even begun to surpass both, into an age of prescriptive decision engines that use machine learning to recommend action to achieve a specific outcome.

Don’t miss this year’s Inman Connect New York — register now!

At the forefront of technology, tech and software companies are battling to create AI that will begin to not only automate parts of sales, but also allow businesses to make better decisions than people — and real estate is just one of the industries poised for disruption.

Elegran's Analytical Journey: Descriptive, Predictive and Prescriptive Analytics

IBM first structured the data analysis landscape, from first descriptive to predictive to finally prescriptive analytics.

Where we are now

By late 2016, most people are aware that we are dependent on AI optimizing the efficiency of our personal lives.

3 essential tools that will 10X your real estate marketing
Smart landing pages, a synchronized database and automation generate results READ MORE

We now ask questions to search engines expecting personalized results, and Google’s RankBrain now uses not only our search terms, but past search data and semantic context, as well as existing algorithms built on top of machine learning.

Facebook identifies our friends using facial recognition and feeds us tailored information based on what we’re likely to click.

Siri and Alexa learn not only our voices but also our buying habits, and act on more and more complex commands every day.

The thrust of technology is focused on not only predicting what has or might happen anymore, but upon building an algorithm to provide us with actionable steps to achieve an outcome we want.

Where we are going

At the same time, machine learning is much bigger than gratifying personal desires. AI has expanded rapidly, especially in the enterprise business market.

Self-driving cars are already on the road (and safer than us already), and Uber has no illusions that it is the cars that are the future, not the drivers they currently employ.

Why, then, do we assume that Uber drivers are expendable, but not real estate agents? Can AI really apply to sales?

Earlier this year I had the opportunity to speak about marketing automation at Dreamforce, where all the buzz in the tech community centered around Salesforce’s newest product announcement, an AI product called Einstein, which will be woven into their sales, marketing and customer service products and will make prescriptive recommendations about how to improve the sales cycle.

Einstein now competes with Microsoft Azure, IBM’s Watson Analytics and Oracle’s Adaptive Intelligence in a race to automate decision-making capabilities of businesses built on a foundation of data architecture.

Machine Learning and Real Estate: Tigh Loughhead and Einstein AI

Analytics, automation and real estate

Experienced real estate agents and analytical marketers dive into their customers’ journeys and data to make descriptive assumptions about their clientele and the market.

We then use this information to try to predict the average time to close on a home, seasonal or market cycles, the best time to call or send an email, where to target particular marketing media for optimal performance, or what tone or language to use with a particular customer.

All of this information is immensely valuable; it takes 10,000 hours for one person to master. But buyer behavior and user input are also just data points, which can be ingested into a machine.

If data is managed properly, a recommendation engine will implicitly also know the best time to contact a consumer, the buyer’s criteria, their preferred outreach medium and the best strategy and messaging to supply that customer with information until they move forward with a deal.

More and more of the tedious or regulated parts of real estate are now taking place online, where numerous products are disintermediating an inefficient value chain (from electronic signature management to automated CMAs to online loan origination). A platform will be less self-interested than any salesperson, and thus ultimately more efficient.

I don’t think I’d be some sort of utopian futurist to suggest that someday a service could initiate an automated process to interact with the consumer from a website inquiry all the way to a closing.

However, if information becomes truly transparent, do real estate professionals become little more than tech support?

The robots are coming for you(r jobs)

While my (and my company’s) livelihood depends on sales and marketing, these platforms not only have the capacity to enhance the efficiency of agents, they could also be beginning to recommend better decisions than a real estate agent is capable of doing — in fact, taking part of the job away.

Lead scoring, anticipation of customer sentiment, identification of sales opportunities and predictive customer journeys — and targeted, automated follow-up, providing dynamically generated content of predictive comparable properties — all of this would actually remove the need for some of the intuition and responsibilities expected in a traditional real estate salesperson.

And the salesperson whose job can be most automated will be the first to go.

Futurists like Elon Musk warn of the potential moral hazard of AI to humanity, but also believe that robots will inevitably take over nearly everyone’s jobs. Musk was just quoted claiming “…people will have time to do other things, more complex things, more interesting things.”

Ultimately, all businesses will be disrupted by AI — though real estate agency might be more resistant to automation than other jobs, as the value-add agents provide is so tied to a personal relationship.

But the free flow of information isn’t evaporating, and machine learning isn’t happening any slower (in fact, it’s speeding up).

One thing is certain: Although the rise of machine learning might be inevitable, agents who don’t learn how to leverage automation and AI to grow their business will find themselves out of a job, while salespeople who do adopt technology are poised to inherit the future of real estate.

My team is one of the first companies using Einstein. If it works as advertised, it will leverage the large amount of data we have carefully organized in our CRM and make prescriptive recommendations to our agents as to how and when to reach out to a prospect.

Machine learning will actually utilize our prospects’ past buying criteria and learn from the data gathered throughout the sales cycle, to advise our agents to be more efficient and become better salespeople.

The rise of AI is just beginning, and could hypothetically be a threat to any business, but in the race against the machine, the only sure losers in the long-term will be Luddites ignorant of technology altogether.

Tigh Loughhead is the marketing director for Elegran.

Email Tigh Loughhead