To Predict and Serve

CRM may more often be found in the boardroom than in the courtroom, but one local police department’s investment in predictive analytics—and a subsequent drop in crime—may herald a drastic expansion of CRM’s capabilities. 

In the past, the Prince George’s County Police Department (PGCPD), in Landover, Md., would have to manually prepare data on active crimes each morning. This labor-intensive and time-consuming process required the participation of specially trained officers, and often delayed the delivery of critical information.

The PGCPD has implemented some new software, however—an application called Active Crime Report (ACR) and WebFOCUS Rstat from the vendor Information Builders. ACR allows police officers to get information on demand, with data automatically updated every 15 minutes. Officers of any rank, regardless of technical skill, can run reports and look for information, monitor incoming calls, learn where crimes are occurring—and not only respond more quickly, but anticipate when specific districts require a shift in policing. 

“ACR allows us to be proactive instead of reactive,” says PGCPD Lt. Henry White. Investigators and crime analysts use ACR as an internal search engine, preparing officers to anticipate circumstances. They can query the system for the details needed to solve any type of offense: the person who fielded the original call; potential patterns; and information on the suspects, victims, or vehicles involved. 

The optimized data entry and access are a boon, of course—and ACR also enables officers to answer citizens’ questions more quickly—but the true benefit for PGCPD is the improvement in resource allocation that the system has made possible. Officers no longer burdened by manual data entry are able to spend more time patrolling areas where crime is actually occurring.

Soon after ACR was implemented, for example, a local community leader asked for information on a high-priority case involving a missing person. In the pre-ACR era, simply identifying the officer who had taken the original call or been assigned the case would have taken an unpredictable number of phone calls. This time, however, an officer was able to access ACR and locate the relevant information within seconds. 

PGCPD officials credit the quick and and effective implementation of ACR for helping produce some startling successes: Total violent crime in the county dropped by 11 percent from 2008 to 2009, with 21 percent fewer homicides and a 25 percent reduction in auto theft. There were also 73 percent fewer carjackings in 2009 than there were in 2005.  

PGCPD Chief Roberto Hylton says he checks his BlackBerry before dawn every morning to track crime statistics on ACR—and that he expects his commanders to do the same throughout the day. When crime rises above an established norm in any area in the county, the chief connects with the appropriate commander, who will likely have already deployed additional officers to that area. ACR is also a staple in upper-level PGCPD meetings: On his laptop, Hylton can pull up information on any district’s crimes to examine the real data and discuss strategy with fellow officers.

The entire department now gets real-time information from the always-on application. “I use the system all the time—first thing in the morning, last thing at night, and throughout the day,” Hylton says. “It literally makes it possible for me to know what’s going on in the field and the department at every minute of the day, and to plan our strategy accordingly. It’s definitely helping us reduce crime in the county.”

In other words, smart software makes smart cops even smarter. “Our approach is ‘intelligence-led policing,’” Hylton says, “and ACR is essential. It gives us confidence. It keeps us from being blind. I need to know what occurs throughout the day, and the officer needs to see what’s happening on his beat. Now we know how to protect ourselves and the community.”

PGCPD is now looking to push its predictive analysis even further, using data and resources to anticipate criminal activity before a crime is even reported. At present, the department is using ACR to focus the manner in which it handles car thefts, analyzing trends such as the types of cars stolen, where they’re stolen from, and where they’re found.

 ACR has fundamentally changed how the officers of PGCPD are performing tasks, automating some of the department’s slowest, costliest, and most manual processes. By getting better, more timely information, presented in an easy-to-grasp, accessible style, the officers of Landover, Md., are definitely taking a byte out of crime.

THE PAYOFF

With predictive analytics provided by Information Builders, the Prince George’s County Police Department has seen:

• an 11 percent reduction in total violent crime between 2008 and 2009;
• a 21 percent reduction in homicide;
• a 25 percent reduction in auto theft; and 
• a longer-term reduction in carjackings (down 73 percent from 2005) and auto theft (down 60 percent).