We pulled off because the tank was down far enough to make it sensible and the back seat had gone quiet in the way that means the next hour is going to cost somebody something. Two stations, one on each side of the road. I took the one on my side, filled up, went in for a drink and something for the kids, paid, and got back on the road.
Nothing about the place asked to be remembered. I could not tell you now which brand it was, and I have driven past both since without a flicker of preference. The pumps were the reason we stopped and the reason the site exists at all, and the pumps were also the part of the visit I paid the least attention to. Fuel is fuel. Whatever the stop was worth to anybody happened inside, in the short walk between the door and the counter.
Petrol is a commodity
The obvious reading agrees with the industry. Petrol is a commodity. A litre is a litre, you cannot differentiate a thing that is identical everywhere, and so stations compete on price and location and nothing else. The two boards face each other across the road and move within a cent, because each one is watching the other.
That description is not paranoia. The Commerce Commission’s 2019 market study into retail fuel found that competition was not working well: importer margins had more than doubled over the previous decade, fuel-company returns were roughly double the Commission’s estimate of a reasonable return, and firms had been making persistently higher profits than you would expect in a workably competitive market. (Commerce Commission 2019)
So the reading holds, on the fuel. It just assumes the fuel is the business.
What a forecourt actually earns
This is where I had it wrong, and the correction comes from an industry that publishes its own accounts. In the United States, NACS splits a forecourt’s economics in two, and the halves do not match. Fuel is about two-thirds of the money that goes through a site and only around 39% of its gross profit. Net margin on a gallon runs to roughly 13 cents, about 1%. And about 57% of the people who stop to refuel also come inside. (NACS 2025)
Inside is where the money gets made. Food service on its own accounted for something close to 40% of in-store gross margin in 2024, and it is the fastest-growing part of the shop. (NACS 2025)
That is American data, and a New Zealand forecourt is not an American one. The local nuance sharpens it. The Commerce Commission located most of the excess profit in New Zealand fuel upstream, at the importer, not at the site where you stand with the nozzle in your hand. (Commerce Commission 2019) So the chain is not thin end to end, and it would be wrong to say there is no money in fuel anywhere. What is true at the forecourt is narrower and more useful: the litre is the biggest number on the property and one of the smallest sources of its profit, and the operator cannot move it in any case, because the board across the road is doing the pricing.
A forecourt does not sell fuel. It sells the decision to pull in.
The reason can be a fruit ice cream
On a state highway in summer, with the tank still half full and the back seat going quiet, where to stop is not a fuel-price calculation. It is human, and it turns on one question: is there a reason to stop here rather than push on for another twenty minutes? The reason can be absurdly small.
The reason can be real fruit ice cream. Not the pale, too-white kind that is more frozen air than fruit, but the sort so thick with berries the whole thing goes purple, the kind you can see is good before you taste it. Done properly and done every single time, so that a family learns it can trust the one on that stretch of road and starts planning the drive around it. That is not a petrol product. It is a reason to choose a petrol station, which is a different and far more valuable thing, and it costs close to nothing beside a fuel margin nobody at the site can move.
I had this filed as empty ground, and it is not empty. BP has run Wild Bean Cafe on its forecourts here for years. Loyalty schemes and supermarket fuel dockets have been the industry’s standing answer for two decades. Food to go is precisely where the big networks have been putting their effort, for precisely the reason the profit split gives them. The space across the road is occupied, and by people who can read a margin report.
What has not happened, as far as I can find, is narrower and harder. Nobody has made a trusted, every-time real-fruit ice cream the reason to stop: not a line in the freezer beside the ice blocks, but the thing a site is known for, good enough and consistent enough that a driver will plan a trip around it. That is a much smaller claim than the one I started with, and it is the one that survives contact with the industry.
Why the freezer is harder than the idea
The gap is not sitting there because nobody thought of it. It is sitting there because consistency is the whole product, and consistency is expensive.
An ice cream worth planning a drive around has to be the same at every site: a supplier who can hold quality through a summer, freezers that hold temperature, and staff trained to make it the same way on a wet Tuesday in Waiouru as on the Saturday of a long weekend. Get it wrong once and the damage is lopsided. One watery scoop at one station does more harm than a dozen good ones repay, because the thing being bought was the confidence to plan around it. That is an argument, not a finding, and it is also why forecourt food service tends to be run by a handful of scaled operators instead of licensed in as somebody else’s brand.
The belief is doing work as well. People running a site think of themselves as being in the fuel business, or at a stretch the convenience business, so a freezer with a supplier and a training problem attached reads as a distraction from the real job. The profit split says the freezer is the real job.
The version of this I keep sketching is called Peachy, and it is a concept, not a company: one freezer, one supplier, and a promise never to skimp, attached to whichever forecourt wants a reason to be picked.
The board and the counter
The split explains the behaviour better than any story about tired management. The one number a driver can read from the road is the number the site cannot move, because the board opposite moves with it. Every number a site can move sits behind the door, on the side of the business that earns most of the profit and gets none of the signage. So the forecourt goes on competing over the litre, which was settled before anyone arrived, and leaves the short walk from the door to the counter, where the choice is actually earned, to a pie under a lamp.