Why do 85% of AI efforts in real estate fail?
Emily Wright - CREtech
Thanks for keeping up with the CREtech Community in the News! For more news about members of the CREtech Community, follow us on Linkedin.
This was one of the big questions posed last month at CREtech London where delegates heard that over three quarters of attempts to implement AI within the real estate sector are not currently meeting the desired outcomes.
Not the most inspiring statistic on the face of it, but the fact it was raised as a problem to be scrutinised is actually extremely encouraging. And when Phillip McCorkle, chief executive of management consulting at RealFoundations put that figure out there on day one of the event, it felt like a collective plaster had been ripped off. AI’s exceptionally high failure rate - a problem that many people in the room were either all too aware of or feared would be the case - had been addressed head-on. And so, all eyes were firmly on what he had to say next as the crowd waited for some workable solutions. So, what were they?
For McCorkle, the biggest problem is a lack of understanding. Once that has been addressed, he said, the sector should start to see the failure rate start to fall. “There are more acronyms within AI than there are in the military and consulting professions combined,” he said. “So when a poor executive within a real estate company gets through reading all of this new language the only acronym they can think of is WTF. They don’t understand it. And the only thing guaranteed in this industry is that real estate executives and professionals do not invest in what they do not understand.” The upshot, he added, is a slow learning curve across the sector. When AI is deployed, it is currently showing an extraordinarily low hit-rate. “Anywhere between 70% and 85% of AI efforts within the sector fail to meet the desired outcomes,” he said. Why? Because swathes of the sector are looking for quick-fix solutions from external providers before making sure they are AI-ready. Then, and only then, said McCorkle, can a targeted approach be executed. “AI is not the ends to a mean, it’s the means to an end. Businesses are feeling pressured to just do something so they are throwing darts at random and that’s why efforts are failing. We need to go back to the internal platform like a game of golf. The most sophisticated golfers in the world know it is about competing against the course, not the other players. And the course is the market.”
He added that real estate companies need to go back to “grassroots business fundamentals” and ask themselves questions like what opportunities do we have to automate repeatedly well-defined work? How do we take that work away from human touch, put it in the hands of a machine and elevate the purpose of the individual who surrounds that work? How do we actually allow asset managers to focus on managing the assets rather than creating volumes of data? “Talk about the most grossly misused capital on all of these high-paid asset managers who are nothing more than data grunts,” he said.
As you might expect, it wasn’t just McCorkle delving into AI over the course of the two days. From JLL chief executive Christian Ulbrich warning that the metaphorical AI trains are leaving the station to Head of Google Maps’ Ed Parsons on why we are still in the, albeit rapidly moving, early stages of AI development, conversations across the board were firmly underpinned by the impact of machine learning on real estate.
So far, so unsurprising. What is worth clarifying though, is that the discussions at CREtech were not an exercise in supposition and conjecture. These were conversations anchored by hard truths on why AI is, so far, being relatively unsuccessfully deployed and cemented by strategic updates from some of the biggest companies in the sector around how they are grappling with this, as yet unfamiliar, technological game-changer. JLL, alongside it’s venture capital arm JLL Spark, set the scene with an event-opening fireside chat focused not just on the business’s all-in approach to innovation and technology investment - something Ulbrich has championed since 2017 - but the company’s current strategy around AI and machine learning more specifically. Referring to the arrival of this digital advance as a tidal wave, JLL Spark managing partner Raj Singh said the company was focussed on riding it, rather than becoming submerged. Ulbrich added that the business has deployed an AI-powered tool to analyse 20 years’ worth of transactional data to help drive leads before adding that JLL’s Q1 earnings results showed that 27% of the company’s capital markets revenues were impacted by the product.
This kind of targeted application in a business where there is already a large data lake ripe for analysis is where AI use is proving particularly successful but CREtech London was also anchored with words of warning around the extent to which the sector should be relying on machine learning and AI. Hanif Kara, the founding director of AKT II and one of the world’s best-known engineers behind buildings including Broadgate, Bloomberg’s European HQ and both Apple and Google’s HQs made it clear that while the technology will make certain things possible, that doesn’t always mean it’s a good idea. “The emergence of AI is an exciting time for engineers as long as we don’t end up designing the world,” he said. “You do not want buildings designed by engineers, because they are quite ugly usually. A whole street designed according to what I have in my head would be awful. We have to be very cautious about not stepping into the heads of architects and developers.”
Pros and cons, challenges and opportunities… it felt like we did more than just scratch the surface when it came to machine learning and AI at CREtech London. But this is a huge, ever-evolving technology within a huge, ever-evolving sector so those conversations will, by their very nature, also continue to be huge and ever-evolving.
A good reason, if ever I have heard one, to book your ticket for CREtech New York on November 13-14, where those discussions will be picked back up and put in force once again.
Register now for 25% off your CREtech New York pass with code EMILYNY25. This offer is valid for one week.
- Emily
Emily Wright
Head of Content
CREtech