The single biggest thing you can change about a home pizza is what it sits on. Here is what each material actually does, with numbers.
Everything else about a home pizza - the flour, the hydration, the sauce, how long you proofed it - shows up in the crust only if the surface underneath can deliver heat into the base fast enough. A domestic oven tops out somewhere around 500 to 550 F, which is roughly a third of what a wood-fired oven floor runs at. The surface is how you claw back some of that gap.
This is also the category with the most confident wrong information in it. The claim you will meet everywhere is that steel conducts heat "18 times" better than stone. That number is doing at least three different jobs on different pages, and it cannot be right for all of them. So we did the arithmetic ourselves from published material constants, and published the working.
8.5x
steel's thermal effusivity against cordierite
Our pick in this category
Baking Steel Original, 16 x 14 x 1/4
The default surface for a domestic oven, and the one whose maker publishes the most
#ad We hold no verified live price for this product, so we show none. How this site is funded.
Heads up: some links on this page are affiliate links. If you buy through one we may earn a commission at no extra cost to you. It never changes a ranking, and we accept no payment for placement. The full version.
Five stones ranked on published specs and computed preheat behavior, with the two that publish nothing called out.
Best Pizza Stone
FibraMent-D Home Oven Baking Stone, 20 x 15
#ad We hold no verified live price for this product, so we show none. How this site is funded.
Five steels ranked on published mass and thickness, with the preheat arithmetic that shows a steel is slightly slower to heat than a stone.
Best Pizza Steel
Baking Steel Original, 16 x 14 x 1/4
#ad We hold no verified live price for this product, so we show none. How this site is funded.
Four surfaces ranked for grill use, where thermal shock and one-sided heating - not stored energy - decide what survives.
Best Grill Stone
Unicook Heavy Duty Cordierite Pizza Stone, 15 x 12
#ad We hold no verified live price for this product, so we show none. How this site is funded.
The comparison table first: conductivity, heat capacity, effusivity and preheat, all computed from published constants.
Cast iron matches carbon steel to within 5 percent on effusivity at a fraction of the price. Why the rim, not the physics, is the argument.
Screens and perforated pans buy you an easy launch and cost you contact area. What each is for, and when the trade is worth it.
Mass, preheat time and stored energy at each steel thickness, with the case for a quarter inch made in numbers.
The first-use routine in six steps, plus the two handling mistakes that account for most broken stones.
A five-step cleaning routine, the burned-cheese method, and why every stone goes black and should be left that way.
Six causes ranked by how often they are the culprit, the physics behind each, and whether a cracked stone can still be used.
When people say a steel is "better" than a stone they are usually collapsing three separate physical properties into one figure. They are not the same, and they do not have the same ratio.
Thermal conductivity is how fast heat travels through a material. Published figures: carbon steel 50 W/m.K, cordierite 1.3 to 1.7 W/m.K. That is a ratio of roughly 30 to 38 times. Volumetric heat capacity is how much energy a given volume holds per degree. Steel comes out at 3.71 MJ/m3.K against cordierite's 1.73 - a ratio of 2.1 times, not thirty. And thermal effusivity, the square root of conductivity times volumetric heat capacity, is the one that actually governs how hot the contact patch stays when a cool wet base lands on a hot surface. There the ratio is 8.5 times.
Effusivity is the number that matters for crust, and it is the one nobody quotes. It is also why a steel and a stone can hold almost the same total energy - a 16 by 14 inch quarter-inch steel and a 16-inch cordierite round are within about 15 percent of each other on stored kilowatt-hours - and still produce visibly different bases.
The other thing the marketing gets backwards. Run the Biot number for any of these plates and it lands at or below 0.1, which means heat crosses the surface far more slowly than it travels inside. Internal conduction is not the bottleneck. Mass is.
Work the lumped-capacitance model through and a 7.3 lb cordierite round reaches 95 percent of oven temperature in about 25 minutes; a 15 lb quarter-inch steel takes about 27; a 27 lb three-eighths steel takes about 42. The steel is not faster to preheat. It is heavier, so it is slightly slower - and Baking Steel's own published guidance of 45 to 60 minutes agrees with our model for their plate, which is the closest thing to validation we can offer for a site that owns none of this equipment.
Start with your oven, not the material. Measure the rack, subtract an inch of clearance on each side, and buy the largest surface that fits. Usable area is the upgrade almost nobody makes and everybody notices - a base that overhangs the surface bakes pale at the edge no matter what the surface is made of.
Then decide what you bake. If it is mostly pizza and you want the crispest possible base, a steel. If it is bread as much as pizza, a stone - the gentler effusivity is more forgiving with a long bake, and a stone does not need seasoning. If you want most of the steel effect for a fraction of the money, cast iron sits within about 5 percent of carbon steel on every thermal number that matters; the catch is the rim, which changes how you launch.
Then be honest about lifting. A quarter-inch steel is 15 pounds. A half-inch is 32. If a plate is heavy enough that you will not take it out to clean it, buy the lighter one - the surface you actually use beats the surface that wins on paper.
Who should not buy a steel: anyone whose oven runs below about 450 F, anyone who bakes mostly enriched or sweet doughs that scorch on a fierce base, and anyone who cannot comfortably move 15 pounds at chest height. None of those are edge cases. All three are common, and none of them are mentioned in the roundups this page is competing with.
Every figure above is worked through in full on our methodology page, with the material constants, their sources and a calculator-checkable example.