It began decades ago. First as a trickle, then it turned into a flood. Before billings became automated and before computers were able to capture it, the amount of vehicle data that fleet managers managed was fairly small. How much did fuel or repairs cost, what was the spend for, when, and for which vehicle? That was it — at the beginning.

Then programs became more sophisticated, technology raced forward, and pretty soon fleet managers were told they had access to every bit of excruciating detail on every purchase made in the operation of their fleets. The torrent of information continues today unabated, and fleet managers can now slice and dice fleet cost information in ways their predecessors could not imagine. 
Is this a good thing? Is there such a thing as too much information?  

What exactly do fleet managers manage? Are they really managing vehicles or are they managing information? Fleet managers, for the most part, manage vehicle information. Most fleet managers see few of the vehicles for which they’re responsible, and depend almost entirely on the information and data the operation and possession of those vehicles generate. 
Therefore, the more data a fleet manager has access to, the better job the fleet manager will do, right? After all, when you have minute details on a vehicle’s costs, “drilling down” will result in the greatest savings. Or so it is thought.

Creating the Monster

Fleet management began to evolve hand-in-hand with the remarkable technological revolution that began in earnest in the 1970s. Prior to the widespread advent of computers and other types of automation, the kind of data detail that fleet managers needed to track vehicle cost and performance was difficult to capture. Fleet lessors and other service companies were able to provide it, but only on paper via voluminous reports, and exception reporting was cumbersome and difficult. (Some may recall the days when every new report had to go through the full IT process: requests, project planning, programming, testing — all done on mainframe computers). Then came PCs, which could be networked to share data, and desktop software into which data could be fed by suppliers, with reporting run from there. 
When the Internet burst onto the scene, suppliers created websites where customers could log in and view their data quickly; create, schedule, and run exception reports; and, overall, manage the fleet quickly and easily.

Some suppliers will admit privately that they’ve created a “data monster,” the sometimes insatiable need for the most arcane minutiae that can be captured.
“We can provide you Level III data on all fleet spend,” fleet managers have been told. “There are dozens of data elements that you can mine 24/7 to find savings,” suppliers boast. And sometimes, fleet managers are simply overwhelmed by all that information, with little time to use it despite the tools at their disposal.

What’s Important?

Let’s start from the beginning. Back when most everything was manual or when “computer” brought about science fiction visions of a HAL 9000, a huge machine that needed to be programmed to perform the simplest of tasks, fleet managers not only managed data, but staff as well. The old rule of thumb that a fleet needed three people per thousand vehicles held sway, and it wasn’t unusual to see a fleet department with a half dozen full-time staff. There were thus ample people to perform the manual gathering and manipulation of data, backed by expert programmers whose favorite question was, “What do you want?” whenever a report was needed.
Then came the outsourcing trend begun ironically in those very IT departments by IBM, as routine tasks were outsourced to suppliers who may have had the expertise in the form of the process, but little in the substance. Companies outsourced their entire IT departments to IBM (and other such firms), only to find that this critical function wasn’t something you left to others. Similarly, the outsourcing trend soon enough began to move into fleet management, as suppliers told their customers that they didn’t need a fleet department; just turn it all over to them, and all the fleet’s ills would be cured (and money saved).
And all along, there was that increasing wave of data. More sophisticated means by which data would be captured arose every year: swipe cards for universal fuel purchases; maintenance management programs with automated data capture; accident reporting with a single, toll-free call. Fleet managers would know not only how much, for which car, and what driver, but what grade of fuel, whether the maintenance was an oil change or power steering fluid, and have photos of accident damage sent to them electronically. 

Next came the start of shrinking the fleet department staff. That staff of six became three, then one, and finally went away altogether. Yes, data was captured in waves automatically, but what to do with it? Fleet managers had only 24 hours in the day, and no one to help mine all that detail for savings. The result? Fewer (or nonexistent) staff to manage and study a rapidly increasing set of information, and that is what leads to the need to determine what is important, and what is just “interesting, but uselesss” data.

Further, the data flood isn’t just in vehicle expense categories. In 2010, some 1.9 billion people sent more than 107 trillion e-mails. Yes, trillion. The technological revolution, now in full swing, has made communication from cell phones to e-mail, from Twitter to Facebook to texting easier than ever before. A quick test: count the number of e-mails you receive in a typical work day, sort them by relevance to the job, and you’ll find that you receive literally hundreds of useless, irrelevant e-mails every week. 

[PAGEBREAK]Where the Money Is

Resource-strapped fleet managers began to prioritize to determine what data was important, where to focus limited time and resources, and how to obtain the best results possible. The first step in such a process is to look around and see where the money is. 

Fleet expense is generally divided into two broad categories: fixed or holding costs, and variable or operating costs. The former are those costs related to leasing or ownership, insurance, and license/title/tax expense, and the latter are any costs related to the operation of the vehicle (e.g., fuel, maintenance/repair, tires, and oil). Fixed costs are dominated by depreciation (or, in the case of leased vehicles, depreciation reserve), which can encompass 70 percent or more of fixed costs. Variable costs are, in turn, dominated by fuel, which similarly will be 70 percent or so of operating expense. Both depreciation and fuel can fluctuate; actual depreciation fluctuates depending on the used-vehicle market, and fuel expense on prices at the pump. 

Now you have a fleet manager, trying to juggle responsibilities for hundreds, perhaps thousands, of vehicles without staff, and with limited departmental resources. He or she also has access to a huge store of detailed data, easy to access via a supplier’s Web tool. At the same time, there are the inevitable meetings, conference calls, webcasts, industry functions, travel, and emergencies that demand immediate attention.

Is it critically important that, for example, tire expense, which amounts to about a penny a mile (out of a dime or more in variable expense), be monitored closely, particularly if that time spent will inevitably draw from time spent doing other things? Fuel can be as much as seven to ten times tire expense. Should a fleet manager, thus, ignore tire expense to focus on fuel or depreciation? The short answer is no. Fleet managers should not “ignore” any expense category to focus on others. But, with time and resources limited, it is clear that more time should be spent managing the larger expense categories than the smaller ones. Where the money is, and the payback for time spent is much greater.

Drowning in Data — Thirsty for Information

The data deluge isn’t limited to the fleet profession; industries from banking to manufacturing have more data available to them than ever before, to the point where they simply save everything that comes in the door, to decide what to do with it all later. And, “later” seldom — if ever — arrives. A typical smartphone today can capture and store more data than a roomful of 1970s mainframe computers. 

Data, however, isn’t necessarily useful information; indeed, much of it can comfortably be classified as “interesting, but useless.” What kind of information, then, is a fleet manager looking for or should a fleet manager be looking for? Simply put, any information that can be used to either avoid or reduce expense.

First off, data must be accurate to be of any use at all. Ensuring the accuracy of data captured is a project unto itself. A simple example: fleet fuel cards can capture literally dozens of individual data elements — what was pumped, how much, cost per unit, date, time of day, self vs. full service, etc. One of those elements is the type of fuel: regular unleaded, premium, diesel, ethanol, etc. The problem that arises is that the data captured when the card is swiped at the pump is at the mercy of the merchant’s efforts to program those devices properly.

Most fleet managers can tell a story of a transaction that came through showing diesel fuel pumped into a four-door, gasoline-powered sedan, only to find out upon follow up that the point of sale device was coded incorrectly, and that it was regular unleaded that was purchased. There are many other examples of data that is dependent upon the source, where human error inevitably comes into play. 

Unfortunately, it is impossible for fleet managers to ensure the accuracy of such data; they must depend on the source to scrub it clean, so that what they get is accurate. With so much data flowing in, the caveat that there may be errors in that flow becomes part of the management process. “Drilling down” becomes more important here. Looking at all the detail from the outset can overwhelm a fleet manager. It is much easier, and more efficient, to begin with a “big picture.”

Take an overview of the forest before getting into the trees. If, for example, a fleet manager runs an exception report to see what kinds of fuel the drivers are buying and discovers that there is an entry showing “diesel” in a fleet whose vehicles are all gasoline powered (or vice versa), the next step is to drill down and find out why diesel is appearing, and take action to correct the errors. (If a driver actually is, for whatever reason, using diesel in a gas engine, the drilling down won’t likely be necessary, as an engine failure/replacement will soon follow). 

The point here is that fleet managers need to start at 30,000 feet, so to speak, before diving down to treetop level. Start with the big picture. Data doesn’t become information until and unless it has relevance and use. Knowing what brand power steering fluid was added to your vehicles is of no practical value. Knowing how much full- and how much self-service fuel was pumped is very useful, since full-service fuel is nearly always priced higher than self-service.

Learn what data is truly actionable information, how to separate the “wheat from the chaff” in the data harvest, and know how to put that information to good use in cost avoidance/reduction.

[PAGEBREAK]Slow Decision Making

Here is another problem that the data flood causes: decision making becomes excruciatingly slow. Why? Here’s a scenario to which any businessperson can relate. You’re at a meeting, the lights are down, and the presenter is on slide five of a 40-slide Microsoft PowerPoint presentation. The presentation is crammed with numbers, graphs, charts, bullet points — data. As you, and your fellow audience members, try to keep from nodding off, you wonder “is anyone in this room absorbing what this guy is saying?” Is the presenter hoping that the sheer volume of the data in the presentation will help to stifle any challenge to whatever point he’s trying to make? Sound familiar? 

These kinds of meetings, along with memos, documents, and other forms of modern communication, are all too common, as managers at all levels of the company try mightily to make use of the reams of data to which they now have access. Carry this further, and you can now see how the decision-making process can become a victim of the data flood. You study a problem, come up with a solution, and present it to management. Somebody along the way sends it back with a note, “what about _____?” or “run the numbers again, but build ______ into the model,” adding their own piece of arcane data to the process. The further up the decision-making tree the solution gets, the more such comments and requests come back, until first it becomes completely unrecognizable, and to whatever problem or challenge the solution was addressed has now mutated into something different or larger.
There’s nothing innately wrong with using relevant information (the more, the better) to develop solutions. The data flood reaches every corner of the organization, from the mail room to the executive suite — and everywhere in between. If decisions require input from other levels of the organization, a flood of data can do much to slow the process, as each individual who has input will surely provide it, selecting from that ocean of data, and thus making the decision-making process unwieldy, slow, and unresponsive to the customer. Attention spans being what they are today, the more data available, the worse the problem becomes.

What to Do?

First off, there is really little a fleet manager can do to “dam up” the data flood. The nature of the fleet programs most fleet managers use (leasing, maintenance management, collision management, fuel cards) is such that they grab and send in as much data detail as possible (this is inevitable when developing programs and services to attract the widest possible market). The data comes in, and the fleet manager can’t stop it. 

Thus, as described above, once a fleet manager comes to terms with the fact that the data is coming — like it or not — the next step is to determine what data is or can be developed into information useful to the fleet mission. One way to do this is to determine what information will help the primary responsibilities fleet managers have: to acquire and manage the operation of a fleet of business vehicles in the safest, most cost-efficient manner possible. Going back to the issue of relevance, recall that the large majority of any fleet’s costs are generated in only two expense categories, depreciation and fuel. Thus, information relevant to either of these will be of primary interest.
● Original vehicle cost.
● Resale performance.
● Time in service.
● Maintenance/repair tire spend.
● Used-vehicle market data.

● Fuel type.
● Cost per fuel unit.
● Fuel grade.
● Self or full service.
● Non-fuel spend under fuel program.
● Cost/use ratios such as mpg and cpm.

There are more, but the picture is becoming clearer. Initial focus should be where the most cost is, and data that can be turned into information used to avoid and reduce these costs. 

Beyond these two categories, a fleet manager’s ability to deal with other data will depend upon the time and resources available, and the relative impact the data will have on the mission. For example, one can make the case that tracking just preventive maintenance scheduling and performance is a critical function, not only in and of itself, but to the extent that proper preventive care will impact both depreciation and fuel performance. (This is indeed true, to some extent, for most expense categories; however, maintenance and repair are both “next in line” in relevance to fuel in variable expense.) If a vehicle is well maintained, it will likely get good fuel mileage and bring higher resale proceeds as well. 

[PAGEBREAK]Drinking From the Fire Hose

There is little question that as it pertains to data, fleet managers often feel as though they’re trying to drink from a fire hose. It has always been true that a fleet of company vehicles produces a great deal of data; it is only recently true, however, that fleet managers have access to nearly all of it, some in real time, and fewer resources and time to do anything with it. An embarrassment of riches? Not quite, but there’s “gold in them thar hills,” as the saying goes. It’s just a matter of breaking enough rock, and digging deep enough, to find it.

Given the choice of returning to the “good old days,” when capturing data was difficult, time consuming, and often inaccurate, and learning to deal with the data flood that technology has created, it is reasonable to assume that most fleet managers would choose the latter, provided they have learned or are willing to learn how to deal with it. You’ll find there are even fleet suppliers who rue the “data monster” they’ve created, where customers demand more and more data, in real time, than ever before. But dealing with it takes careful planning, choices, and patience. 

Decide, first, what data is important to carrying out the fleet manager’s mission. What will help keep drivers safe, vehicles on the road, and expenses manageable? In other words, learn how to turn raw data into critical information.

Know what resources you have and don’t have, and plan to apply those resources to the task. If time and resources are dear, as they are with most fleet managers, go where the money is — focus most efforts on managing the two major fleet expense categories, depreciation and fuel.

Understand the difference between critical information and “interesting, but uselesss” data. The more a fleet manager has to work and think about how the data can be applied to the mission, the less relevant it likely is.

Automate, automate, and then automate some more. The technology that created that flood of data can be used to corral it and use it to do the job. Concentrate on exception reporting (knowing what isn’t right, as opposed to what is).

Watch data trends, as opposed to just the data itself. What is going up that shouldn’t, going down that should, and the effect of decisions over time.

Use data to benchmark. Benchmark internally, that is, performance today versus previous periods. Benchmark externally versus other fleets of similar size, vehicle type, and the same industry. Make certain that external benchmarking data matches your own before drawing conclusions.

Data, in and of itself, isn’t bad or good. It is either useful or it isn’t, and it is up to the fleet manager to determine whether it is or not. Grab control of that data flow, learn how to use it, and fleet performance will benefit.