In Sync or Not

by By Bob Levins, Angus Analytics

As our desire for incremental improvement appears to be accelerating in all areas of the business and delivery in particular, I have been spending much of my time evaluating delivery data, as it relates to operational efficiency. Admittedly, I am part of the old school and was around when many of our old processes — some still in use today — were developed. Degree day accounting has served us well for so many years. The current back-office accounting systems continue to enhance this forecasting methodology in any way they can. A key component of projecting deliveries, other than the obvious use of degree days to measure temperature over time, is the dealer’s determination of the “Ideal” or “Optimal” drop size value to be assigned to his/her customer’s tank(s). This number, as we all know, represents a threshold at which a customer’s product usage will or should yield an efficient delivery, while providing a cushion of gallons in reserve to cover any miscalculation in forecasting, manual adjustment of delivery schedules or slight changes in burning habits. It was and still continues to be our “best guess” at when a customer is due for product, trusting that their burning rate has remained constant.

With this said, and of the many process changes or lack of changes that I have noticed over the years, the ideal drop threshold has been continuously decreasing, thus affecting the overall delivery efficiency/costs of many dealer operations. Back in the day, the ideal drop default value assigned to a 275-gallon heating oil tank used to be approximately 80% of the usable fuel (20% reserve). Translated, a 275 tank with a 256- gallon usable would be assigned an ideal drop of 204 in the tank record. I have witnessed targets decreasing down through 200 to an average of 175, and lately I have been seeing some companies as low as 150. When I challenge the dealers about this, they are accepting of the shorter deliveries, reacting to their continued fear of runouts as well as their concern of customers’ complaints that their bills are too high. There are certainly exceptions to this trend but few in number. I was visiting with one of my clients recently and pleased to notice that his ideal drop for 275 tanks was set to 200, something that I haven’t seen for years. His approach, along with his impressive delivery performance results, encouraged me to pursue something I have been thinking about for some time, formed around this hypothesis.

Increasing every tank’s current ideal drop by 10% will not represent the perceived corresponding linear increase in runouts on which the current drop reduction is being justified.

I decided to get together with our Senior Database Analyst (Dave) and set out to find the truth. We took a sample size of approximately 30,000 deliveries for calendar year 2018. The count represented 17,000 heating oil and 13,000 propane delivery transactions. We took every transaction (automatic and scheduled), multiplied each tank’s ideal drop by 1.1, and then added the resulting increase to the actual units delivered, within their respective transactions, imitating as if the ideal drop was higher than currently set. We achieved the results shown above.

Referencing the Heating Oil line in the example above, review the definitions for each column below, for our logic:

Run Out Count (A) – This is the number of runout transactions stated by the dealer via declaration in their back-office system, keeping in mind that many runouts can occur for reasons other than miscalculations (credit holds, late deliveries, etc.).

True Run Out Count (B) – Number of runouts declared by dealer via their back-office system, plus the number of transactions for which the units delivered actually exceeded the tank’s usable units on record, where the transaction was not marked as a runout.

Run Out Using Adjusted Target Count (C) – By adding our 10% increase of ideal drop to the units delivered, this column represents the number of transactions that would exceed the tank’s usable units if the additional units were actually realized on the delivery, keeping in mind that the count is influenced by very late deliveries that were increased by the 10% as well.

Optimal Delivery Units Additional Units Sum (D) – This is the total units that would be delivered across all deliveries (the Delivery Count) — in this case, this equates to an average of an additional 19.6 units per delivery (340,386/17,328).

It was an interesting exercise that yielded results that will, moving forward, hopefully make us question our previous motives for reducing ideal drop to offset runouts. In the case of heating oil (propane followed a similar path), we found a less than 1% increase in runouts across all deliveries while achieving a potential increase of 340,000 gallons that would have been delivered had we increased the ideal drops. This could also lead to fewer deliveries over time. Of course, our objective now will be to better qualify the accuracy and contributing conditions of each runout contained within the dataset in an effort to further reduce the reported minimal increase in runouts. This exercise is just one simple but organized example of how your data can be used to either confirm, disqualify or tweak the effectiveness of your existing common practices that have been developed over the years.

Related Posts