May 2018

Are K-Factor deliveries your biggest impediment to profit growth?


What do the following items all have in common?

  • On-board computers
  • Route optimization software
  • Back-Office Systems
  • Company websites
  • Email
  • Voicemail

The most common thing is they are all core technologies that are embedded in almost every distribution company, and most companies couldn’t compete or survive without them. The other common theme is they weren’t around when your grandfathers started the company.

However important and obvious the value of these necessities—and other “basic technologies”—the truth is that each was purchased for the company with a degree of insecurity as to the understanding of the technology, the need, process changes required, the potential impact to the company’s operations and whether it was worth the cost. More simply put, the uncertainty about new technology’s impact, combined with the out-of-pocket costs delayed implementation of these items for weeks, months and often times for years.

People rationalize consciously and subconsciously. When you are not certain as to the benefits of something, and you know that you need to write out a check, it’s pretty easy to dismiss the “thing” on the basis of not seeing the benefits. Not surprisingly, as prices come down, people somehow become smarter and more understanding very quickly!

My suggestion is to separate out the “understanding” from the “cost.” If someone understands the benefits of certain types of technology, the only logical obstacle should be the cost (return on investment).

Deliveries are absolutely unpredictable

Whether we understand why or not, the correlation between the Heating Degree Day (HDD) and the gallons consumed—commonly known as the K-Factor—simply do not correlate enough to allow dealers to maximize delivery efficiency. To be clear, the mathematical formulas that make up a K-Factor are correct. However, the implicit assumption that each home consumes fuel at the same rate per HDD ignores customer’s awareness of the cost of fuel, visitors to the home, travel of the homeowners, wind, humidity, etc. In other words, it ignores the real-world items that do impact the consumption.

FILE - In this Jan. 5, 2010 file photo, Jason Kilpatrick of Wholesale Fuel hauls a hose across a snow covered yard while delivering home heating oil in Framingham, Mass. Natural gas prices climbed Monday, Feb. 8 as another winter storm was expected to dump even more snow on the East Coast. (AP Photo/Charles Krupa, File)

(AP Photo/Charles Krupa, File)

It is clear that you will consume more fuel in a colder winter than a normal winter, with all else being equal. It also makes sense that a normal January should require more fuel than a normal October, with all else being equal. However, the numbers do not lie and they show us that consumption, even in the same house in the same winter, is not formulaic. Weather and consumption are not linear, and we see that every time there is a run-out or a very small delivery for a K-factor customer. On average you are doing “okay”… perhaps. However, if your sense of “okay” is delivering only 55% of a tanks’ capacity (150 gallons into a 275 gallon tank), then you are either comfortable with not optimizing your deliveries or the deliveries simply cannot be optimized using the current delivery forecasting tools.

K-Factors do tell a story

K-factors let you know how much was consumed between deliveries. That time span might be 25 days during the winter and might be 150 days from the summer to the fall. In either case, it is not tracking consumption as it happens, but using it to report what has happened in the past—usually about six times per year. The math, as mentioned earlier, works…mathematically. However, it is not capable of accurately predicting forward consumption in a way that you can use to increase your profits.

How do we know this? We know this because if you look at individual deliveries, not average deliveries, we see numbers that are all over the map. Yes, they will concentrate between 150 and 170 gallons, but there will also be a number of deliveries above 200 gallons, a bunch closer to 100 gallons, and of course the dreaded run-outs. The “standard deviation”—most easily explained as what would be deemed “normal” mathematically—was about 16% for the delivery-to-delivery period this winter in tracking over 160,000 daily reads from tank monitors. If you are averaging 170 gallons, it would mean that a range from around 140 to 200 would be deemed to be normal. That being the case—and it is the case—it is understandable that, if a 200 gallon delivery were to be considered a normal delivery, you likely wouldn’t want to raise your target, lest you see more run-outs. It also should be noted that a quarter of deliveries were outside of the “normal” range, meaning that there was a reasonable likelihood that even with a target of 170 gallons, you might still have a good number of +200 gallon deliveries.

That is the reality of the dealer who is using K-Factor forecasts to make deliveries: Low targeted delivery sizes, wide variances in the actual individual deliveries and a few run outs that just rationalize the small targeted delivery size. It’s a cycle that you really can’t break out of. Yes, we are all aware that “Ks” need to be adjusted and accounts need to be looked at. On the other side, it can take about two winters before finding a “good K-Factor” for a new customer, delaying optimization even more.

The upshot is that K-Factor forecasting causes you to target small deliveries so that you can manage your run-outs. It does a good job, but so would deliveries to every customer every week—hardly a run-out to be seen there! K-Factor forecasting, if the only tool available, does a competent job of helping you avoid run-outs, but when your single biggest operational expense is the cost to deliver fuel, isn’t it time to embrace the technology that is now available?

It might have taken you a few years to understand and to buy on-board computers. It might have taken your father a few years to move from manila file folders to a back office system. Both cost a bunch of money and that cost made the “need” seem to be less of a need. For all you know, your great-great-great grandfather had a hard time buying a clock when his sundial was “okay.”

The cost factor of tank monitors still needs to be considered. However, the impact—larger and more predictable deliveries, smoother operational logistics, lowering staffing constraints, fewer run-outs, improved customer engagement—is undeniable. All this comes with costs that are far lower than when you first looked at them 10 years ago. In the future, all customers will have remote monitors—just like every other delayed technology that you now embrace. ICM

See other similar Articles in the Information management Category.