How Much Tile and Grout You Actually Need

Area divided by tile area is always short

The naive calculation treats tiles like water: pour enough square feet of tile into a floor of the same area and it fills. Tiles are not water. They are rigid modules laid on a grid that starts at one point and marches outward, and it meets each wall wherever it happens to meet it. If the wall arrives three-quarters of the way through a module, you cut a tile and throw away a quarter of it. Four walls, four rows of cuts, and the offcuts from one wall are almost never the right size for the opposite one.

There is a second, quieter error hiding in the same calculation. The number printed on the box is the tile size, and the tile is not what occupies the floor — the tile plus its share of joint is. A 12 in tile with 3/16 in joints occupies 12.1875 in of floor in each direction. That is only 1.6% per axis, about 3.1% by area, and it is the difference between a full row and one more row you have to buy. On small formats it is worse: a 4 in tile with 1/8 in joints occupies 4.125 in, a 3.1% axis penalty and 6.3% by area.

Count the grid: a worked room

Take a bathroom floor 8 ft 6 in by 11 ft 4 in — 96.33 sq ft. Tiles are 12 in square, joints 3/16 in, so the module is 12.1875 in. Across the short axis, 102 in ÷ 12.1875 = 8.37, which rounds up to 9 columns. Along the long axis, 136 in ÷ 12.1875 = 11.16, which rounds up to 12 rows. Nine times twelve is 108 tiles: 108 sq ft of tile for 96.33 sq ft of floor, 12.1% over, and that is before a single tile has been dropped.

Add 5% for breakage and you are at 114 tiles. Tiles come in cases; at 10 sq ft per case you need 12 cases, 120 sq ft. Add one full case of attic stock and you are ordering 13 cases, 130 sq ft, for a 96 sq ft floor. That is 35% more than the naive answer, and none of it is padding — every step of it is either geometry, handling or insurance.

Whether the grid step costs you anything at all is largely luck. Run the same 12 in tiles through a 10 ft by 13 ft room and the arithmetic gives exactly 10 columns and 13 rows — 130 tiles for 130 sq ft, no rounding loss at all, because the room happens to be a whole number of modules on both axes. Two rooms, the same tile, the same method, 12.1% waste in one and 0% in the other. This is why a fixed percentage rule cannot replace the count: the answer depends on your room, not on a convention.

What a diagonal actually costs

Turn the grid 45 degrees and every perimeter tile becomes a triangle. That is the whole story, and it is worse than it sounds, because a triangle cut from a square leaves a second triangle whose hypotenuse runs the wrong way — it fits the opposite corner of the room, not the next position along the same wall. In a straight lay a cut piece and its offcut often tile two opposite edges; in a diagonal they usually do not.

The trade rule is 10% extra for a straight lay and 15% to 20% for a diagonal or a herringbone. It is worth being clear about where that comes from: it is supplier and installer practice, not a published standard. The TCNA Handbook and BS 5385 cover installation methods, substrates, movement joints and workmanship — they do not tell you how many tiles to buy. So the percentages are conventions, and conventions can be checked.

So I checked them geometrically: place the tile grid at 45 degrees over the room outline, try every starting offset, and count the grid cells the room touches. On an 8 ft 6 in by 11 ft 4 in bathroom with 12 in tiles, the best diagonal layout needs 110 tiles against 96.33 sq ft of floor — 14.2% over, where the straight lay needed 12.1%. On a 10 ft by 13 ft room the gap is far wider: 150 tiles diagonally, 15.4% over, against a straight lay that came out exactly even. The diagonal penalty is not a fixed number; it is the loss of the lucky cases.

Herringbone is harder still and deserves its own warning. Rectangular tiles set at 45 degrees produce a repeat that almost never matches the room dimensions, every course terminates against a wall on both axes, and each terminating tile is cut on an angle that leaves an unusable remainder. Twenty per cent is a floor, not a ceiling, and it is worth dry-laying the first two square metres before you commit to the order.

Grout from first principles

Most calculators go vague at exactly this point, which is a shame because grout is the easiest part of the job to compute exactly. The joints form a continuous grid, and the share of the surface they occupy is 1 minus the tile area divided by the module area: f = 1 − LW / ((L+w)(W+w)). Multiply that fraction by the joint depth and you have a volume per unit of floor. Multiply again by the density of the mixed grout and you have a mass. There is nothing else in it.

That formula has a consequence worth internalising: halving the tile size roughly doubles the grout. At 1/8 in joints and 5/16 in depth, 24 in tiles need about 3.0 lb per 100 sq ft, 12 in tiles 5.9, 6 in tiles 11.6 and 3 in tiles 22.5 — each halving multiplies the requirement by about 1.95, approaching exactly two as the joint becomes small relative to the tile. It is the same reason a mosaic sheet eats grout: you are not buying tile any more, you are buying joint.

Joint width pulls in the same direction and almost as hard. Keep 12 in tiles and widen the joint from 1/8 in to 3/16 in and the requirement goes from 5.9 to 8.7 lb per 100 sq ft — a 48% increase for a change most people make on aesthetic grounds without thinking of it as a quantity decision. Wider joints also need a sanded grout for crack resistance, which is a different product with a different density, so the bag you were pricing may not be the bag you buy.

Why the coverage on the bag may not be your coverage

Every bag prints a coverage figure, and every one of them is conditional on a joint profile the printer had to assume. The most important assumption is depth, and depth is not tile thickness. The joint runs from the tile face down to the top of the adhesive bed, and adhesive squeezed up the tile edge during bedding takes some of that depth away. On a 10 mm tile you may have 8 mm of joint, or 6, depending on how the tile was bedded. Hold everything else at 12 in tiles with 3/16 in joints and vary depth alone: 1/4 in needs 7.0 lb per 100 sq ft, 5/16 in needs 8.7 and 3/8 in needs 10.5. That is a 50% spread from one number nobody measures.

The tile edge matters too. A rectified tile has a square, mechanically ground edge and gives you the joint you specified. A pressed tile usually has a slightly rounded or bevelled edge, which widens the joint at the surface even though your spacers are the same size — more grout, and a visibly wider joint line than the spacer suggested. Density is the third variable: cementitious grouts run roughly 1.5 to 1.9 kg per litre depending on whether they are sanded, and epoxy grouts behave differently again, so a coverage figure is specific to one product and not transferable.

So use the bag's figure as a sanity check, not as the calculation. Derive your own from the geometry, add 10% for what stays in the bucket and on the float, and round up — grout is cheap, a second trip is not, and a bag opened three days later will not colour-match the first because cementitious grout darkens as it cures and each mix carries slightly different water. Mixing a whole floor from one batch on one day is worth more than the few units the extra bag costs.

Frequently asked questions

Is 10% extra enough?

Sometimes, and you cannot tell without counting. Ten per cent is a convention from suppliers and installers, not a published standard, and the grid arithmetic shows why it is unreliable: two rooms tiled with the same 12 in tiles came out at 12.1% and 0% waste purely because of how the room dimensions fell against the module. Count the grid first; use the percentage only to cover breakage and cuts on top of it.

How much extra for a diagonal or herringbone layout?

Fifteen to twenty per cent on top of the grid count is the working rule, and the geometry supports the upper end. Rotating the grid 45 degrees over two real rooms pushed the tile count 14.2% and 15.4% above the bare floor area, against 12.1% and 0% for the same rooms laid straight. Herringbone is worse than a plain diagonal because the pattern repeat rarely matches the room, so treat 20% as a minimum and dry-lay a corner first.

Do I subtract the area under the bath and the toilet?

Subtract a built-in bath or a fitted cabinet plinth, because you genuinely tile up to them and not under them. Do not subtract a toilet, a pedestal basin or a freestanding appliance: you tile under and around those, the cuts are fiddly, and the offcuts are wasted. In practice a toilet increases your tile requirement rather than reducing it, because the cuts around the pan and the pipe leave unusable pieces.

How deep is the grout joint really?

Not the tile thickness. The joint runs from the tile face down to the top of the adhesive bed, and adhesive squeezed up the edge during bedding shortens it — on a 10 mm tile you often have 8 mm of joint or less. Depth matters more than people expect: holding tile and joint width constant, going from 1/4 in to 3/8 in of depth raises the requirement from 7.0 to 10.5 lb per 100 sq ft, a 50% swing.

Why does a mosaic use so much grout?

Because grout scales with total joint length per unit of area, and that length grows as tiles shrink. Halve the tile size and you roughly double the joint length, so you roughly double the grout: 24 in tiles at 1/8 in joints need about 3.0 lb per 100 sq ft, 12 in need 5.9, 6 in need 11.6 and 3 in need 22.5. A fine mosaic can be more joint than tile by surface area, and it should be budgeted that way.

How much attic stock should I keep?

One full unopened box, kept flat and dry, with the batch code visible. That is enough for a plumbing repair or a cracked tile, and it is worth far more than its price because tile models are discontinued within a few years and shade batches never match across production runs. Keep half a bag of the same grout with it, sealed, and write the mixing ratio on the box.