Coverage is not linear.

The relationship between how many tire sizes you carry and how much demand you actually capture is steeply concave. Almost nobody has plotted the curve for the market they serve.

Every operator in this business knows intuitively that a small number of sizes drives most of the volume. It is the oldest observation in inventory management. What almost nobody has done is plot the actual curve for the market they serve and look hard at what the shape implies.

The curve is a property of the fleet, not of any company. Take the vehicles registered in a given geography, map them to their fitments, weight them by where they sit in the replacement cycle, rank by projected volume, and plot cumulatively. What you get describes the market itself — before anyone makes a single purchasing decision.

The shape

Using Ohio’s registered fleet as the worked example from our earlier analysis, 33 sizes reach 80% of projected replacement demand. That is the steep part of the curve. Then it flattens, hard.

The first thirty-three sizes buy you eighty points of coverage. Getting from 90% to 95% costs another seventy sizes for five points.

Coverage is not something you buy at a constant price. Every additional point costs more than the one before it, and the last several points cost more than most operators have ever calculated.

What the tail actually is

The long end of the curve is not random noise. It is a specific and predictable population, and it contains at least four different things that look identical on a slow-mover report:

• Aging fitments. Sizes whose OE placement ended a decade or more ago, attached to a shrinking and increasingly price-sensitive vehicle population
• Trim-level fragments. A size that exists because one trim package on one nameplate for two model years specified something slightly different
• Emerging fitments. Sizes on current OE that have not yet reached first replacement — real demand that has not arrived yet
• Genuine specialty. Applications that belong in an entirely different stocking logic

Those four demand opposite responses. The aging fitment should be managed down toward exit. The emerging fitment should be built up ahead of the wave. Treat them the same and you will systematically stock the past and miss the future.

The case for deliberately not carrying something

Most organizations treat a stockout as a failure. In the tail of the curve, that framing is expensive.

A size in the 96th percentile of market demand might turn once or twice a year. Against that you are carrying acquisition cost, a physical slot, handling, obsolescence risk, and the opportunity cost of capital that could sit in a size turning twelve times. If the practical alternative is a next-day source or a special order the customer accepts without much friction, that size may be actively destroying value by occupying space.

This is not an argument for thin inventory. It is an argument for knowing precisely where on the curve each size sits and making a deliberate decision, rather than inheriting one from a replenishment minimum somebody set years ago and nobody has revisited.

Where you sit is a strategy, not a constant

Different positions on the curve are correct for different businesses.

• A high-velocity operation serving price-driven volume may be right at 85%, running lean and fast
• An operation competing on service level may need 95% and should know exactly what those last points cost
• A commercial or specialty operation may need deep tail coverage, because the tail is the business

All three are defensible. What is not defensible is arriving at a coverage level by accident, through years of accumulated settings, and never having priced the decision at all.

Why the obvious version of this fails

The sketch above sounds straightforward enough that most analysts would assume they could build it in an afternoon. Count the vehicles, map them to sizes, rank, plot. That version produces a curve. It also produces a curve that is wrong in a specific and predictable direction.

The failure is not in the counting. It is in the assumption that a vehicle in operation converts to replacement demand at a uniform rate.

• Vehicle age is not replacement position. Two vehicles of identical age can sit at opposite ends of their tire cycle depending on annual mileage, prior replacement history, and whether the vehicle is still on original equipment
• Vehicles do not consume tires at the same rate. Replacement intervals vary widely across the fleet, and a single blended rate applied uniformly averages away the thing that matters most
• Scrappage is not uniform either. Older vehicles exit the fleet at rates that vary by segment and region, and a raw count treats a vehicle about to be retired the same as one with a decade of service left
• Emerging fitments are invisible to volume ranking. A size on current OE that has not reached first replacement ranks near zero today and near the top in thirty-six months

Run the naive version and you get a curve that systematically overstates the tail and understates what is about to arrive — which is the same error a sales-history model makes, arrived at by a more expensive route.

The curve is only ever as good as the replacement model underneath it. That model is the actual work, and it is not something you derive from a registration file.

Converting a vehicle population into a tire population is where TreadSignal's method departs from the standard approach, and it is the part we do not publish. What we will do is show you the curve for your market and let the output make the argument.