Siding color visualizer: 2,848 US siding colors, and how few of them are actually different

Every siding manufacturer runs a color visualizer, and every one of them shows you the same thing: their colors, on their houses, arranged so the palette looks generous. What none of them can show you is the only number that matters when you are three brands into a decision and starting to lose track. How many genuinely different colors are you choosing between?

We can answer that, because the whole shelf sits in our catalog as measured values. As of today, the US siding market in our data is 2,848 active colors, across 95 product lines, from 10 manufacturers. That is the number the industry effectively advertises.

It is not the number of choices you have.

2,848 listings, about 710 distinguishable colors

Of those 2,848 listings, 2,797 carry a published color value. We set aside 132 variegated blends, whose tone shifts board to board so a single value would be a fiction, and 117 whose value we derived from a swatch image rather than a published spec, which is accurate enough to sort by and not accurate enough to publish. That leaves 2,587 measured, solid colors to compare properly.

Converted to Lab and compared by color difference, they collapse hard:

at ΔE < 1, which is a difference almost nobody sees on a wall: about 965 distinct colors
at ΔE < 2, roughly the point where two large painted surfaces stop reading as the same color: about 710
at ΔE < 3, a difference you can see when the two are touching and not when they are on different houses: about 485

So the working figure for "colors you could actually tell apart on a facade" is around 710 out of 2,848. Roughly three out of four listings on the American siding shelf are a color that already exists somewhere else on that same shelf.

A note on method, because a number this convenient deserves it. Counting distinguishable colors means grouping ones that sit closer together than the threshold, and the grouping depends slightly on which color you start from. We ran it in several orders: the ΔE < 2 answer moved between 707 and 716. That is why we write "about 710" and not a false-precision figure. The counts that do not depend on ordering, and are exact, are the ones below.

Half of the palette has a twin at another brand

Here is the exact version of the same finding, measured per color rather than per group. Of the 2,587 measured colors, 1,273 have a near-identical match at a different manufacturer at ΔE < 2. That is 49.2%. Loosen the threshold to ΔE < 3, and it climbs to 1,988 colors, or 76.8%.

Some of the pairs are not "close." They are the same value. Alside's Glacier White and Westlake's Aspen White are both published as #FFFFFF. Alside's Vintage Wicker and CertainTeed's Savannah Wicker are both #CCBFA2, under two different names, in two different catalogs, from two companies that do not mention each other.

This is the single most useful thing to know before a quote. When a contractor tells you the color you fell in love with is only available in a line that pushes the job past your budget, that is a real constraint about the line, and often not a constraint about the color. Half the time, the same color is sitting in somebody else's catalog under a different name.

The same house rendered with a fresh white facade, black windows and dark trim — What a swatch cannot tell you: how the color behaves across a whole elevation, next to your own roof and windows.

Two-story American house photographed from the street before any exterior work — What a swatch cannot tell you: how the color behaves across a whole elevation, next to your own roof and windows.

The catch: a color is not sold on its own

If 710 colors were simply available, this would be a short guide. They are not. A siding color exists only inside a product line, and the lines behave in two completely different ways depending on what the color physically is.

When color is a coating, it travels across the model range. At Allura, Diamond Kote and Norandex, every single color name appears in more than one product line, one of them in as many as 14. Mastic repeats 92% of its names across lines, Alside 76%. For those brands, choosing the color and choosing the profile are two separate decisions, and you can make them in either order.

When color is the material, it does not travel at all. At James Hardie, not one color name appears in more than one line: 721 named colors, four collections, zero overlap. LP Building Solutions is the same. There, the color and the product are one decision, and picking the color has already picked the line.

There is a second trap inside that number. Of James Hardie's 721 colors, 686 belong to a single made-to-order collection. Their in-stock range is 19 colors. Both are real, both are in the catalog, and the difference between them is weeks of lead time and a different price per square. A visualizer that shows you all 721 as one palette is not lying to you, but it is not telling you which ones a supplier can hand you next week.

And one thing nobody publishes: 10.7% of the measured US siding palette is genuinely dark (L* under 35). If you arrived wanting a near-black house, you are choosing from roughly 276 colors, not 2,848, before availability narrows it further.

What to do with this

The practical order is the opposite of the one the swatch wall suggests.

Start with the line, not the color. Decide the material, the profile and the budget first. That single decision removes most of the 2,848 before color is even discussed, and it is the decision you cannot reverse later without re-siding the house.

Then shortlist colors, and shortlist small. Within a line you are usually choosing between a few dozen, and after the ΔE math, meaningfully between fewer. Five candidates is a real shortlist. Twenty is a stalled decision.

Then look at them at full scale, on your house. This is where a visualizer earns its place, and where the manufacturers' own tools stop being useful: they show you their colors on their demo houses. A color that reads warm on a small chip can read washed out across a two-story elevation facing south, and next to your roof, which you are not replacing.

Then order samples for the last two. A screen narrows the field. A physical sample settles the color. Neither replaces the other, and the mistake is using either one for the other's job.

Try the shortlist on your own house

Upload photos of your house and see siding colors rendered on your actual facade, at full scale, with your roof and windows in the frame, and compare the versions side by side. You see what each direction costs to build before you commit to it. 5 designs to start, topped up daily.

In our catalog the US market has 2,848 active siding colors across 95 product lines and 10 manufacturers. Measured against each other, about 710 of them are distinguishable on a wall, and 49% have a near-identical twin at a different brand. The shelf is much larger than the actual choice.

Often, yes. Just under half of measured US siding colors have a near-identical match at another manufacturer, sometimes down to the identical published value under a different name. What usually does not transfer is the product line: material, profile, warranty and texture are what you are actually paying the difference for.

It depends on whether the color is a coating or the material itself. Some manufacturers repeat every color across their whole range, so color and profile are independent choices. Others, including James Hardie and LP, use each color in exactly one line, so choosing the color has already chosen the product.

No, and it is not trying to. A visualizer shows the color at full scale on your own elevation, next to your real roof and windows, which is what eliminates most candidates. A physical sample shown at your house in daylight is what settles the final one or two.