It’s an exciting time for the Real Estate industry. While the benefits of AI and Machine Learning are well acknowledged across dozens of verticals, CRE has traditionally been a few steps behind in terms of digital transformation. According to the Forbes Real Estate Council, this is all about to change, as traditional business models are usurped by the newest innovation and smart technology.
Commercial real estate [stakeholders] who are open to new methods and who evolve with the latest disruptive technologies should remain market leaders. Innovation will often produce very good results if you’re willing to embrace it. If not, you are likely to be left behind.
These disruptive technologies are diverse, and include brokers using software to give potential tenants 3D tours of their properties in advance, or chatbots that can negotiate a fair deal for a property with a tenant directly. In particular, the use of Data Analytics and Machine Learning is changing the way that many CRE businesses operate, allowing stakeholders to act with certainty rather than guesswork.
How Can I Benefit from a Predictive Model?
- On a basic level, businesses are now able to answer many of the questions that were traditionally complex or impossible to answer using manual analytics. Traditionally, a client who wants to know how they are measuring up in terms of tenant renewal may have been able to find some Open Data on the topic, or loosely compare themselves to the market more widely. With a room full of engineers and months of sifting through datasets, manual analytics will never be able to accomplish what a well-trained AI model can achieve at the click of a button. Comparisons with industry or market averages are just the beginning. With Machine Learning technology, businesses can understand past mistakes using root cause analysis, recognize what’s happening in real-time, and even look to the future to make accurate predictions about what will
happen next. - The first hurdle for many companies is recognizing that the data you need is likely to be already in your possession. Despite popular belief, you don’t need thousands or hundreds of thousands of datasets in order to benefit from Machine Learning algorithms. In fact, a few hundred is enough to start gleaning information from. At Okapi, we take the customer data and then enrich it further with third-party information from disparate sources. This could be market information, social media, governmental data or more. By giving the computer this data, we can train it to learn which features are predictive features, and which factors really make a difference.
- Identifying which variables are significant can have a powerful effect on your business strategy. Let’s say for example that you are a landlord who puts a lot of your messaging and branding behind keeping your price point low. You pride yourself on being affordable and you believe that this is what keeps your tenants loyal and happy. Through our predictive model, this gut feeling can be either reinforced, or debunked. Underneath your instinct, it could be that the classification of the building, its attributes or location are actually what encourages your tenants to renew their leases time and time again. Raising your price, and thereby your bottom line, might not cause as much change as you fear, and your empty units could be filled by focusing on the real variables that make a difference. For the first time, CRE will have answers to the questions that make a real difference to business optimization. It’s no surprise that the Deloitte 2019 CRE Industry Outlook report proves more than 80% of stakeholders believe that predictive analytics and business intelligence should be a priority for CRE. In fact,
over the next 18 months, nearly two-fifths [of companies] plan to increase the use of these two technologies in their business decisions.
How Can I Be Sure it’s Reliable?
Training a model is a long process. By testing enough variations of existing data, we know that new information will fit seamlessly and accurately into the model that we have created. This needs to cover multiple datasets and types of information. Take a company that leases 10,000 buildings for example.
We build our predictive model able to handle multiple types of data and give accurate results without adding complexity. Our goal is to take the noise of all this data, and turn it down so you can hear the music underneath. By paring it back and extracting only the relevant actionable insights, you have widely applicable information to use in your unique business context.
In order to do this, we use simulations. By taking the data from 2016-2017 for example, we can enrich this, and predict what will happen in 2017-2018. We can then check the results against the real-life data from 2017-2018, and measure how accurate our model is compared to what actually happened in that year. In this way, we get a good level of assurance that when the machine says something is going to happen at a high level, it will.
Defining your Questions from a Business Perspective
- The capabilities of Machine Learning and AI for your company will depend heavily on the questions you are looking to answer, and the outcomes you want as a business. These are not always one and the same.
- Often, a client might approach us wanting to know whether they can raise rental prices without losing tenants. This is the question they want to answer. Underneath that question is the outcome they are looking for – most likely to increase their profit margin. Using Machine Learning we can look at meeting the outcome in the best way possible. This may well be to raise rental prices, but it could also be to reduce maintenance costs or to streamline operations behind the scenes.
- Using our Root Cause algorithm, we can also learn from existing data to understand why an event occurred. Let’s say that in 2016, 75% of leases were renewed, while in 2017, this fell to just 60%. We can tell you why. Understanding how and where you performed badly as a business can help you recognize where you should be focusing your attention, and which parameters affect performance overall. These actionable insights give you a good sense of changes you should be making as a business, such as shifting maintenance from one place to another, or what specific groups you are not speaking to effectively at the moment.
Where Else Can AI Make a Difference?
Tenant renewals are a great focus point for understanding how AI can revolutionize your business process, but they are far from the only benefit. Our powerful algorithms provide a competitive advantage across the board.
Predictive Maintenance
Managers can now stay two steps ahead, with alerts into the machines and systems that are about to fail. Enriched data through building sensors, similar machines, and even the amount of people using the equipment takes you out of crisis mode, for good.
Resource Management
You can plan ahead with the right teams at the right time for maintenance, cleaning, supplies and more. Never overspend preparing for a doomsday scenario that doesn’t arrive, and don’t allow yourself to be caught short without sufficient workforce for the load.
Vendor Relationships
By establishing accurate benchmarks, you know exactly what vendors are essential to your bottom line. Identifying trends can give you new ways to partner with relevant suppliers and resources, building your network within the industry.
Cashflow Planning
Monitor and track customer behavior, allowing you to predict who will pay on time and who won’t. With insight into cashflow problems before they become a reality, you can prepare in advance and make smart financial decisions for your business overall.
By: OKAPI (Globe St.)
Click here to view source article.
