Splitting the Atom: Using Data Science to Drive the Right Kind of Residents
Today’s quest for renter leads begs the question whether sales and marketing are working from the same recipe. Quantity mostly trumps quality as metrics tend to focus on activity and outcome. The industry has a wealth of ways to cultivate leads – everything from guest cards to website chats. And according to marketing experts at JVM Realty Corp., marketing’s job is to attract, make people aware and compare.
But a pile of leads, although it may look like a good thing, can crowd the kitchen. The lease conversion funnel, the culling of quality applicants who are most likely to become renters, can get clogged by unworthy leads. This is the fine line the multifamily industry now walks in an age when so many more opportunities to attract residents are at play than they were 20 years ago.
Lead scoring helps properties identify leads likely to become leases
During RealPage’s RealWorld 2017 conference, marketing leaders discussed why traditional leasing measurements like lead volume and lease conversion are facing extinction, and how data science is changing the landscape. “You need volume but you need the right customer. You can pop a lot of junk leads into system that aren’t going to do you any good in the long run, but this is about the right customer at the right place at the right time,” says RealPage Senior Director Leads and Leasing Product Management Caleb Winn.
According to JVM, leads are being supplanted by new technology that shifts the leasing paradigm toward more granular, actionable metrics so property management can focus sales on prospects who are most likely to rent.
Data science of revenue management is the future of multifamily. The systematic approach involves measuring the quality of inbound leads and comparing them to leasing outcomes. There, marketing and sales effectiveness can be accurately measured to improve the chances for a lease.
“We believe it’s possible to get more granular and actionable information by splitting the atom,” says Winn. “We do this by measuring the quality of inbound leads when they hit the system. It’s using data science to algorithmically score each lead as it comes in.”
“By measuring the quality of inbound leads and comparing that to leasing outcomes, Winn and team have found a way to isolate marketing and sales and measure them separately.”
Predictive analytics are transforming multifamily
Through data science, property managers can measure lead quality, see how predictive analytics are transforming the way multifamily understands business, and understand that crafting a quality-adjusted demand-optimization strategy can deliver dollar-measurable return on investment.
The ability to evaluate lead quality based on statistical evidence − and forecast future behavior based on that evaluation − improves business intelligence and decision-making, Winn said. While lead volume may seem effective, quality leads which result in leases ultimately determine the true effectiveness of marketing efforts.
JVM marketing says that marketing and leasing apartments today require a more analytical approach. The days of skimming the surface by canvassing a phone list to find renters are long gone. It’s messy and time consuming. It’s even difficult to get on same page of what a lead is. There is a lot of cleanup required.
A JVM marketing leader recalled years ago trying to lease up a student housing community that hadn’t been built. Armed with a list of 500 prospects, a leasing associate called each during business hours. Naturally, she got a lot of answering machines and very few callbacks.
“That’s an example of focusing too much on action and activity without accomplishment,” Winn remarked. “You have to deliver to the right people at right place at right time and have the right people there to receive.”
What did we learn? Lowering the price isn’t always the right solution, either. Multifamily is in the business of creating value and slashing rents isn’t the way to do it. Leasing agents need to do their homework and find the kind of renters that add quality to the business at market rate.
Data science is the answer.
Algorithms predict future behavior based on past performance
Rich Hughes, RealPage’s head of data science, said that splitting the atom of lease conversion requires robust scoring of lead quality at entry, continued tracking throughout the leasing lifecycle, reconciliation to leasing outcomes and comparison against benchmarks.
The process works using a complex algorithm based on Bayesian statistical theories around predicting future behavior resulting from past performance of similar datasets. Lead scoring factors in various predictive factors like phone information, email information, guest card content, rent needs, need-by date, work and apartment geo-mapping and call duration time.
Each lead is compared to millions of records to find a match. The comparisons consider how similar leads performed in the past to predict behavior, or how likely the prospect is to lease.
“If you look at enough observations, you manage to distill micro drivers of causations,” Hughes said. “These would not be accessible if you did not have very large repositories of data of actual outcomes.”
Leads are sorted into stages and assigned a score based on probability of becoming a lease. The lead is constantly rescored as new information is obtained.
Rescoring, Winn said, is done from every bit if information the property receives during interaction with the prospect. Every bit of new information the leasing agent receives offers more insight about the prospect and how likely a lease will be the result.
‘This is really the future of bringing together the entire leasing equation’
“The future of leasing depends on advanced forecasting, targeted demand generation and workflow optimization through automation,” Winn said.
The impact data science has on forecasting within revenue management is huge. Knowing the traits of a quality lead will help forecasters get a more accurate look at the future based on what’s in the prospect funnel at the time.
Also, the information can identify targeted marketing toward specific inventory based on demand so property managers can focus on certain product rather than employ a broad-brush campaign. If a property needs to fill two-bedroom units, for example, leads from prospects who are looking for large units should get prioritized.
The new model, Winn says, is taking data science statistics and building out behavior that will maximize the leasing outcomes for the quality demand that has been generated.
“We think there are implications over the next several years to take this data science forward to more impactful levels,” he said. “This is really the future when it comes to understanding the effectiveness of marketing, the effectiveness of sales and pricing, and bringing that entire leasing equation together.”