Do gas stations intentionally throttle pumping rates to improve profits?
I used to drive a lot. Almost 600 miles every or every other weekend, plus a 14-mile round trip commute to work three to five times a week. Even the most fuel-efficient car would be gasping for fuel, and my Camry was no exception. That's when I noticed that the gas pumping rate seemed dramatically lower than what I remember. I may have been more patient back in the day or at least didn't stare at the gallon counter going up trying to figure out when it would be done so I could get on with my day, but what's definitely new is the presence of screens that play a variety of videos: sometimes they're funny, sometimes they're ads, sometimes it's even a vocabulary lesson.
That got me thinking: have companies deliberately slowed pumping rates to expose their customers to more ads per pump? In theory, it makes sense. There aren't enough customers throughout the day to merit the fastest safe pumping rate possible, and intuition tells us more ads means the consumer is more likely to see something that interests him and/or possibly become more convinced that they should purchase the advertised product. Throttling flow is done simply by placing a flow restrictor somewhere before the nozzle. The smaller the orifice, the less flow at a given pressure (here's an unbranded example calculation I found on the internet that I am going to blindly trust).
In checking to see if this has been reported before, I found there is an upper limit on pumping rate as decreed by the EPA:
(j) After July 1, 1996 every retailer and wholesale purchaser-consumer handling over 10,000 gallons (37,854 liters) of fuel per month shall limit each nozzle from which gasoline or methanol is introduced into motor vehicles to a maximum fuel flow rate not to exceed 10 gallons per minute (37.9 liters per minute). The flow rate may be controlled through any means in the pump/dispenser system, provided the nozzle flow rate does not exceed 10 gallons per minute (37.9 liters per minute).
No mention is made of a lower limit for pumping rate, meaning it's possible rates can be throttled.
Total volume of gasoline pumped is measured by using a metering pump.
The data collection was simple. Every time I filled up my gas tank I noted the following:
I set the pump trigger to automatically pump and left it until it automatically shut off.
I see a few confounding variables that can affect pumping speed, causing some issues with my data. Unless I find a way to understand these, I will assume they are the same across stations and companies.
First, a fuel filter is present in the gas line to prevent direct contamination of the vehicle's fuel tank. Clogging can occur, impeding the flow of gas and slowing the pumping rate. I have no way of checking the filter status, short of asking the attendant, which I will not do.
Second, a poor or old pump can provide sub-optimal performance, regardless of the rate the station is calling for. There's no way convenient way to verify the pump model and status.
Third, overall station design. Does each nozzle have its own pump, or are they shared across multiple nozzles? Too much pump sharing reduces the pumping capability and thus pump rate per nozzle. How wide are the delivery lines?
Fourth, some pumps at the same station may differ in speed. If using the same station more than once, I will attempt to use a different pump.
The complete set of data can be found here.
The difference in averages is seemingly well-accounted for in the standard deviations:
| Ads Present | Average (gpm) | Standard Deviation (gpm) | Number of Samples |
|---|---|---|---|
| Yes | 8.603 | 1.457 | 42 |
| No | 8.491 | 1.133 | 19 |
An unpaired t-test gives the following results:
No major differences were found between different brands of gas:
| Brand | Average (gpm) | Standard Deviation (gpm) | Number of Samples |
|---|---|---|---|
| Chevron | 8.231 | 0.603 | 8 |
| Exxon | 8.588 | 1.055 | 25 |
| QT | 9.038 | N/A | 1 |
| Racetrac | 7.461 | N/A | 1 |
| Shell | 8.393 | 2.146 | 12 |
| Valero | 8.738 | 0.787 | 14 |
No major differences were noted between YA and NA gas stations.
The amount of data collected is meager and more collection should be done to improve the trustworthiness of the results.
Zachary Crockett's Why most gas stations don’t make money from selling gas examines the economics of running a gas station and where the money does—and doesn't—come from. A bit surprisingly, "gas stations make an average net margin of just 1.4% on their fuel" while "the goods inside these stores ... account for ~30% of the average gas station’s revenue, yet bring in 70% of the profit".
I could not find any literature on the effectiveness of gas station ads, so general ad effectiveness will have to be looked at.
Regarding exposure time and purchasing intentions, Burton et al.'s Revisiting the Relationship between Ad Frequency and Purchase Intentions found that "consumers who saw an advertisement 10 or more times had greater purchase intentions than consumers with less exposure". While frequency is obviously different from duration, frequency is cumulative duration in a sense, so it's not entirely unreasonable to extrapolate from one to the other.
Slide 7 of Does Every Second Count? Understanding New Video Ad Lengths shows that purchase intent continues to increase past the 15-second mark all the way up to 45-60 seconds, which well past the upper bound for possible gas station ads.
Overall, throttling pumping rates is probably worthwhile for the owner for a few reasons:
- 15 second spots: range of $1,195 to $1,395 per gas station per 4 week period (includes all screens in station).
- 30 second spots: range of $1,495 to $1,795 per gas station per 4 week period (includes all screens in station)
All this being said, throttling flow rates seems like a good idea if the owner is entirely profit-driven and doesn't care about the customer wasting their life sitting at a gas station waiting for their gas to pump.