Saying that a company is “data driven” has become a pledge of sorts, a commitment to business excellence and integrity. “We are not flying by the seat of our pants here. No sir,” it implies. “This is a company with a plan and a program based upon actual data you can trust.”
The challenge has been turning all that raw data into something understandable and usable. Fortunately for fleets, the same technologies that helped create those mountains of information are now being deployed to make sense of it. Thank goodness for business intelligence and business analytics.
Business intelligence answers questions such as: What happened? Where did it happen? Tools like dashboards, scorecards and key performance indicators help fleets see that information in an immediately useful and usable way. Business analytics answers other, often longer-term questions such as, Why did this happen? What should we do about it going forward? It may include activities like data mining, predictive modeling and statistical analysis.
“The challenge folks are faced with is that they now have lots and lots of data. In the case of very large and complex organizations, it may be from many different locations and in many different forms,” says Christopher Shaffer, Utilimarc partner. “The question is how do we get all this information back in a way that allows us to use it — to track and trend both internally and externally in order to make both tactical and strategic business decisions?”
For Utilimarc, part of the solution has been to have customers provide raw data so that it is all in the same format and follows the same model. “We have [our customers] provide us with their raw data and then we build the metrics,” says Shaffer. “That way, everything fits our model so people can be sure they are really seeing apples to apples. That is extremely important. Then we put it all online so that they can query their data in all kinds of ways.”
Queries include comparing overall cost-per-mile with or without fuel, cost per engine, cost per vehicle make and model, total ownership costs and more. “We essentially divide data into three financial buckets — ownership costs, operating costs and support costs,” he notes.
According to Shaffer, the company “takes everything down to the vehicle equipment level. Utilities and municipalities are our history,” he says. “Ironically enough, they are some of the most complex companies to profile and benchmark. They have everything from police vehicles, to lawn mowers, refuse trucks, and alternative fuel vehicles. We track it all.
“As part of the benchmarking process, we make recommendations when we present a customer's data to them,” Shaffer adds. “They often ask us to help with process improvements. The great thing is that the metrics are already in place to do trend analysis, so they can see if they are really making progress [toward their goals.] That closes the loop.”
Shaffer says that seeing data in this way changes things for companies. “They get a new awareness of the value of their data,” he notes. “They really start to see their business differently, to get clarity. The data does not lie.”
SEEING WITH FRESH EYES
Tom McLeod, president & CEO of McLeod Software, observes the same sort of changes taking place among McLeod customers as they begin to look at their data in new ways. “Our users almost always find surprises,” he says. “They see things they did not expect to see. For instance, customers who they did not think were profitable customers sometimes really are.”
Like Utilimarc, McLeod says the company is very mindful of the importance of creating good models, solid business intelligence and analytic tools that take into account all the appropriate data for use in day-to-day operations or in longer-range planning. “We do make a distinction [between business intelligence and analytics],” he says. “Especially in operations, there are some things that individuals can make better if they know about a situation right now. You have to be very careful with the data you use, though.
“There are also longer-term things that you can't change moment-by-moment. These are more complex considerations that take more analysis and may even require changing the organization to implement,” McLeod adds. “In this case, we look at entire groups of sets of information and then drill down.”
Automating business analytics, however, is a painstakingly careful process. “We developed an MBA (Masters of Business Analysis) program,” he says. “There are a lot of decisions to be made and models to be set up. Once that is done, you have to prove the business model in order for people to gain the confidence to use it. There are several ways to do that. For example, we come in and make sure the revenue cost models are set up and then we compare them to the company's financial statements to make sure we really are accounting for everything. When a customer queries the system, they need to know they are not seeing artificially inflated or deflated results.”
According to McLeod, there is much to be gained for fleets, shippers and brokers at the business intelligence level. “Lots of people are getting great results just from seeing more,” he says. “[We can provide] a good look at what is really happening and then there are almost limitless ways to consider that information. Fleets can look at loads under 200 mi. only, at the profitability of loads on Monday versus loads on other days, at when to accept new business and when to turn it down.”
TMW Systems is likewise focused on giving its customers a better look at what is actually taking place, at turning data into true business intelligence that can be used to help guide day-to-day operations or to make forward-looking decisions. “Our Freight Analyst system was really done to help provide a part of what an actual analyst would do for a company,” explains David McKinny, vice president & general manager-optimization and lean management for TMW. “For example, you can run what-if scenarios with the system: What if I add this customer? What would happen? If I acquire another company, what would happen?
“Business intelligence is much broader [and more useful] than a good database. It can tell you more than just what is going on now or what will be happening in three days. Capacity balance is a good example,” McKinny says. “[With the system,] I can see three to seven days ahead at what the capacity versus the demand will be [and plan accordingly.]”
True business intelligence integrates revenue, costs, the impact of timing in areas like service performance and driver hours of service, he adds, along with the basic concept of network balance — how and where your trucks are moving in relation to freight flows, market opportunities and deadhead miles.
Jason Koch, president of Telogis Fleet, shares this perspective on the value of business intelligence and sees it as part and parcel of business analytics. “We don't distinguish [between business intelligence and analytics,”] he says. “Business intelligence is a tool for business analytics. Any good business intelligence tools will do trend lines, for example, so that you can look to the future.”
At Telogis, business intelligence means merging real-time, GPS-derived data with historical views and plan data. “For us, it is all about the single, integrated platform,” says Koch. You can get much more clarity if you have job information as well as GPS information. If you only have GPS data without access to the plan, you have to make a lot of assumptions, and your resolution may not be as tight as it should be. You need to be able to compare the plan to the actual.
Like other software solution suppliers, Telogis sees the market for business intelligence growing and changing. “The market is definitely changing,” Koch says. “Everybody knows what GPS is now. Beyond that, every customer wants routing, business intelligence and the ability to compare the plan to the actual.” Awareness is up and needs and expectations are also up. Tom McLeod also sees a changing business intelligence/analytics marketplace, including expanding capabilities on the supplier side. “Finally [software solution suppliers] are able to deliver things that customers have been asking about for years,” he says, “such as round-trip profitability analysis, for instance. The tools we provide are being used better than they ever have been and we can do more than we ever could before.”
BI / BA / KPI — A spoonful of alphabet soup
Successfully managing a fleet operation is a math story problem to beat all story problems and the lexicon of terms developed to describe the effort is enough to make anyone decide to take up poetry or maybe house painting instead. Still, the terms are useful pegs on which to hang key concepts. So here are a few common ones you may want to keep within easy reach:
Ad hoc query: A little function that describes a lot of hard programming work. Software systems may allow users to access specific types of information or solve problems by automatically analyzing a fixed collection of data points via a “query” or question. For example, think about how many factors are involved in the question, “What is the optimum way to handle this unexpected load?” A useful query would have to consider available drivers and equipment, delivery windows, rates, etc., and a change to any one variable could impact all the others.
Business analytics (BA): The process of examining internal and external data along with other information in order to make informed business decisions often about the longer-term direction of the company. May include activities such as data mining, predictive modeling and statistical analysis. Answers questions such as, Why did this happen? What should we do for the future?
Business intelligence (BI): Answers questions such as: What happened? When did it happen? Gathering and reporting good business intelligence is really at the heart of the fleet management process today. Tools like dashboards, scorecards and key performance indicators help fleets to see what is actually happening in order to make better informed daily and even longer-term decisions.
Data mining: “Sifting through very large amounts of data for useful information. Data mining uses advanced statistical tools…to reveal tends, patterns, and relationships which might otherwise have remained undetected,” is the definition provided by BusinessDictionary.com. In other words, data mining is an automated way of looking for the aha moment.
Dashboards: A user interface that is designed to present various types of business information at once in a standardized, graphical way that is easy to see and quick to understand. Dashboards are generally interactive and permit users to access information in real-time (like the moving gauges on a vehicle's dashboard) so that decisions can be made on the fly.
Heuristics: Wikipedia calls this “Experience-based techniques for problem solving, learning and discovery.” Heuristic methods are used to speed up the process of finding a good enough solution, where an exhaustive search is impractical. Examples of this method include using a rule of thumb, an educated guess, intuitive judgment, or common sense. Fleet owners and managers know all about heuristics.
Key performance indicators: The routine achievement of some agreed-upon operational goal (on-time delivery, no shortages/damage, cost-per-mile, inventory turns, accidents per week, etc.).