RealPage applies AI to resident screening

339

At RealWorld 2019, RealPage introduced its first application of artificial intelligence/machine learning (AI) to its product line. The function that they chose for this purpose is resident screening.

Expanding the view

RealPage says that traditional resident screening techniques focus on a resident’s ability to pay the rent. However, this may lead to losses for the landlord if the resident chooses to spend their available funds elsewhere and to skip out on the rent. To avoid this, the new RealPage software considers the resident’s willingness to pay in addition to his ability to pay.

So, how do you identify a potential resident who has a high willingness to pay the rent? RealPage does it by examining the actual outcomes of past rental engagements. They have a historical database covering over 30 million past rental agreements and whether or not they ended satisfactorily. In addition to the rental outcome data, RealPage pulled in other available resident financial data in order to allow them to make the best possible assessment of the factors leading to the outcomes observed.

Checking it out

Over the last 6 months, RealPage has been beta testing the new capability with several large property managers. The properties who have been using the AI capability have a total of more than 100,000 units under management. RealPage says that their results indicate that properties using their AI screening can expect an average savings of $31 per unit per year without negative impact to occupancy or revenue.

As with any machine learning system, the efficacy of RealPage’s AI resident screening depends on the accuracy of the data upon which it is based. As more users adopt the system, as more accurate data is entered into the system, and as more actual outcomes are recorded, the quality of the assessments made by the system should improve. Because the inputs use in the AI screening algorithms are blind to demographic data identifying protected groups, RealPage says that AI screening ensures Fair Housing compliance.

Choosing your target

When Yardi introduced machine learning to their product line, the application that they selected was one to optimize marketing spending on lead generation. RealPage has picked one to improve resident selection in order to reduce losses due to resident’s failure to pay. Both should pay dividends for adopters. Multifamily property managers should expect machine learning to continue improving all areas of their operations as software vendors continue to roll out product enhancements.

The RealPage press release on AI resident screening is available here.