Everyone has car data.

Vehicle registration data is a commodity. Anyone with a budget can buy it. What almost nobody has is the thing that turns a vehicle count into a tire number — and that gap is wider, and more expensive, than the industry assumes.

Walk into any serious tire business and ask what data they have. You will hear about registration counts, vehicles in operation, garage data, market sizing by metro. All of it real, all of it purchased, most of it from the same two or three sources everyone else uses.

Then ask a harder question: how many tires will the vehicles in that market buy next year?

The room goes quiet, or somebody offers a rule of thumb. One tire per vehicle per year, roughly. Adjust for the market. That is the state of the art in an industry that moves hundreds of millions of units.

The number is wrong, and the method is worse

Start with the arithmetic. Take the published count of light vehicles in operation in the United States. Take published replacement tire shipments for passenger and light truck, excluding medium truck because those units belong to a commercial fleet that is not in the light-vehicle count.

Divide one by the other and you get roughly 0.90 tires per vehicle per year. Not 1.0. Anyone can reproduce this in five minutes from public sources, and almost nobody has, which is why the rule of thumb persists about ten percent above reality.

But the level is the smaller problem. The uniformity is the real one.

Vehicles do not consume tires at the same rate

The spread between segments is not a rounding difference. Battery electric vehicles and performance-fitted vehicles consume tires at multiples of the fleet average, driven by weight, torque, and compound. High-trim touring fitments — large wheels, high speed rating, driven normally — consume below it, because the tires that replace their factory rubber are built for long life rather than grip.

Top to bottom, that spread exceeds a factor of two. Apply one number across it and you will overstock the segments that consume slowly and run short in the ones that consume fast, in the same market, in the same season, using data you paid a great deal of money for.

The error hiding in plain sight

Here is a specific one, and it is worth checking against whatever model you currently run.

Federal Highway Administration data is frequently cited at roughly 13,476 annual miles. That figure is per licensed driver. There are approximately 238 million licensed drivers in the United States and 289 million light vehicles.

There are more vehicles than drivers. Apply a per-driver mileage figure to a vehicle count and you overstate demand by about a third.

Second cars. Seasonal vehicles. The truck that only tows. They sit in the vehicle count and contribute little to the mileage. Tires are consumed per vehicle, not per driver, and the distinction is worth 35% on a forecast that otherwise looks perfectly reasonable.

Why nobody sells the conversion

The registration providers do not publish a tire replacement rate. Neither do the trade associations, which report shipments rather than consumption behaviour. Neither does the government. This is not an oversight — it is simply not their business. They count vehicles, or they count units shipped. The bridge between the two requires knowing how tires actually wear on the vehicles that carry them, and that is field knowledge, not a dataset.

Which means the conversion cannot be bought. It has to be built, validated, and maintained — and it has to be checked against something. A demand model that cannot reproduce a known published total is an opinion with decimal places.

What a real conversion has to survive

• It must reconcile. Applied nationally, it has to reproduce published replacement volume. If it does not, the segment logic is wrong somewhere and the aggregate is hiding it.
• It must be physically grounded. Derived from how far vehicles travel and how long tires last, not fitted to a target. A curve fit cannot be defended line by line; physical wear can.
• It must separate the original-equipment cycle. Factory tires are specified for showroom ride and noise, not wear, and they are consumed markedly faster than the replacements that follow. Collapsing the two into one figure understates young segments badly.
• It must publish its errors. A method that only reports its successes cannot be evaluated by anyone.

The practical consequence

Vehicle data tells you the size of the opportunity. It does not tell you the shape of it — which sizes, which tiers, which months, which locations. Two markets with identical vehicle counts can differ substantially in replacement demand because of what those vehicles are and how they are used.

Everyone buys the first half. Almost nobody builds the second. And the second half is where every stocking decision, every assortment plan, and every forecast that actually matters is made.