Due to the emergency nature of the profession, incident commanders often are forced to make decisions with incomplete and sometimes inaccurate information — there's little time to get caught up in “analysis paralysis.”

But as these on-scene commanders progress upward into management positions, they often carry the same method of problem resolution into strategic decision-making. Quick decision-making may be the hallmark of an effective incident commander, but it is not necessarily the sign of a high-performing upper-management team member.

To combat this from-the-gut management, the Austin (Texas) Fire Department used a data-analysis program to determine apparatus placement, allocate limited resources and enhance the decision-making process.

Initial Problem

Encompassing 256 square miles, the city of Austin is home to 746,000 residents and the hub of a metropolitan area of 1.6 million people. The Austin Fire Department operates 44 fire stations with a force of 1,079 sworn personnel and 67 civilian support personnel. With an annual budget of $121 million, the department responded to almost 76,000 calls in 2008 (71% of them medical calls).

Fire Station No. 27 is located in the southwest area of the city. In late 2000, the department replaced the station's ladder and engine units with a quint as a cost-savings measure. Station No. 27's territory is primarily residential and is generally in an area of town with low to moderate call volumes. During the last two to three years, the quint captains raised concerns that they desperately needed another response unit. They were concerned that the lack of an additional apparatus was leading to longer response times.

At first glance, their request seemed well justified. The geographic response area for Station No. 27 is substantially larger than that of most of the other stations. The call volume was not overwhelming (1,435 calls from April 2007 to March 2008), but there was concern regarding the response times in the area. The station captains prepared a well-organized presentation explaining the critical situation, and those tasked with leading operations were in general agreement that another unit probably would be in the city's best interest.

But even prior to the recent downturn in the economy, the department did not have the resources to create a new engine company. Management would have to reassign another engine company to Station No. 27.

I scheduled a discussion of the issue for a chief officers' workday in fall 2008. Because a battalion was going to lose an engine to fix the problem, the goal was to prepare a defense for the decision. Executive management knew they needed to get their ducks in a row and, if possible, let the story tell itself.

The Austin Fire Department is fortunate to have a planning-and-research section with a principal planner, a second planner and a senior research analyst. I asked this group for data related to response times and backup responses. Using incident-reporting data, the section developed several reports and maps for review, and one explained the situation perfectly.

Early Misconceptions

The report indicated how many times a backup unit had a response time greater than five minutes. This report revealed the historical impact of a unit not being available for a call (because it was assigned to another call, training out of the area, or other reason).

As my staff and I reviewed the data, we realized that Station No. 27 indeed had an issue, but that other territories had even greater problems related to backup response times. For Station No. 27's 1,435 emergency calls, 287 required backup response that took more than five minutes to arrive. Investigation of the data revealed that nine other station territories surpassed this number. Management now believed that the most critical problem was not at Station No. 27, but elsewhere. But they needed independent verification to be completely comfortable with the counter-intuitive direction the data was taking them. Fortunately, the department recently began using a business intelligence (BI) tool that could help with the data analysis.

With an initial investment of less than $50,000, the department began a pilot application of the BI application in 2007 using its existing technical personnel — a database administrator and a project manager. The application used the data in its existing format which made implementation fast and simple; there was no need to create special data warehouses. Ease of manipulation of the data by the end-user was a key concern; if the business managers could not easily operate the tool to query the data, the software would be an overall failure. Fortunately, non-technical personnel discovered they could quickly learn to use the system to query and analyze data. The business users, including combat operations chiefs, the human resources manager, and the financial manager played an important part in the implementation because they were the individuals most familiar with the data.

BI programs allow the user to evaluate a situation from a 30,000-foot perspective, or drill down to a very granular view of the data. The department already was using BI to analyze leave use and demographics. Because of this familiarity, I was comfortable with the quest to analyze response numbers and times. Rather than using information from incident reports, the data came from the city's computer-aided dispatch system. Using the BI software, I was able to validate the data provided by planning and research. This second level of supporting documentation provided the information that needed to be presented to the other chief officers.

Reassessed Plans

It was time to make the presentation, and test the data and preparation efforts. In hindsight, the presentation was almost easy. Having the data to present in a clear format left virtually no room for dissension. There was very little discussion once the reports and maps were revealed. It was obvious to all of the chiefs that there were larger problems with response times than what Station No. 27 was experiencing — that location may have been the “squeaky wheel,” but it didn't need the grease this time.

Fire Station No. 22 is located in southeast Austin. Historically one of the busiest companies in town, Engine 22 services a diverse cross-section of Austinites. Station No. 22 had 3,881 emergency calls in the same reporting period. Of these calls, 552 were handled by a backup company with a response time of more than five minutes. To exacerbate the problem, Station No. 22 is in an area of the city that historically has high call volumes. The fire stations in this area have enough of their own calls to respond to without assisting Engine 22.

The fact that the data indicated that Station No. 22 needed some assistance made sense to the chiefs. Everyone knew that it was one of the department's busiest units, but the personnel assigned there hadn't been concerned about the need for a second unit; firefighters assigned to Station No. 22 take a certain level of pride in the high number of calls they respond to every month.

Again, the effort invested in analyzing data paid dividends; the crews assigned to Station No. 22 could understand management's concerns. To their credit, the crews did not allow their pride in being one of the busiest units in town to overshadow their desire and commitment to best serve the citizens.

The department initiated a six-month test in which Engine 35 was relocated to Station No. 22. With Station No. 35 being the closest station just south of Station No. 22, the engine already was making some of the backup responses into Station No. 22's territory. Serving a territory that is primarily commercial and industrial, Station No. 35 is one of the only areas in southeast Austin with a low call volume. At least for the test period, it appeared that Quint 35 could handle the primary response duties for its territory.

Engine 22 has received some needed relief. In the past, the engine often would respond to more than 20 calls per 24-hour shift; now they make it only into the low teens on busy shifts. The crews assigned to Engine 35 are enjoying the opportunity to make more runs than they were previously doing. Average response times for Station No. 22's territory have dropped by approximately 15 seconds per call. With an average of almost 230 calls per month, that is a substantial improvement in this important performance measure. Also important, the other surrounding units now are rarely dispatched as a backup unit into Station No. 22's territory and can remain available for emergency calls in their home areas.

The results of this test will allow the department to analyze the impact that placing a second unit at a station can make and apply the results to various areas of the city. There are at least three other station territories in Austin with similar response patterns to those of Station No. 22.

By evaluating the available data, the department was able to address the issue at hand properly. The fact that the data presented management with information that was counter to their instincts was not initially comfortable and met with some resistance. But the use of analytical software tools to validate the initial data encouraged management to overcome any remaining hesitations and concerns. The members at Station No. 27 now understand other areas of the city need assistance even more than they do. The department has achieved its top priority of providing effective service to the community.

Informed Decisions

As our officers promote up the ranks, we need to be sure that we are properly developing their abilities to address the new issues they will face. Intuition and instinct are valuable assets, but for non-emergency decisions, understanding that there are data and tools available to properly assess and interpret the situation at hand is tremendously valuable. The department is incorporating BI familiarization into its professional development curriculum for mid- and upper-management personnel. Fire-service managers have to learn the right questions to ask. Being blissfully ignorant of the data is not providing the responsible leadership that our personnel deserve and that our citizens demand. We already are gathering a lot of information. Why not use it to make informed decisions?

The fire service needs to further explore mechanisms to better allocate our resources. Law enforcement has made strides by using COMPSTAT and similar methodologies to address their problem areas. While not directly applicable to the fire service, we can learn from these systems and adapt them to address our situations.

The fire service needs to learn how valuable data in a reportable and analyzable format can be. The benefit is not really validated until you use it to prove a reality or disprove a myth. Once you “get it,” you realize that you cannot live without it.

Div. Chief Ken Crooks is a 21-year veteran of the Austin (Texas) Fire Department. Currently a shift commander over Operations, he also has experience in fire prevention and communications. Crooks holds an associate's degree in fire science from Austin Community College and also attended the University of Texas.

Bttn. Chief Harry Evans and Business Analyst Supervisor Elizabeth Gray contributed to this article.

Related Stories

• Take Statistics One Step Further
• Data in the Driver's Seat
• Deployment Matters