How to make a tile finish at the intended size

A tile can finish exactly 100 mm wide and still be too big for a layout planned on a 100 mm repeat. The missing detail is the grout joint. Before reaching for a shrinkage ruler, I would put two outlines on paper: the tile itself, and the space it occupies beside its neighbours.

Decide what the finished size includes

Layout pitch is the repeat distance from one tile to the same point on the next tile. For a regular grid, that is one tile width plus one joint width. It is not the size to cut in clay.

Suppose your drawing calls for a 100 mm by 150 mm repeat, with a 3 mm joint in both directions. The intended finished tile is then:

• Width: 100 − 3 = 97 mm.
• Length: 150 − 3 = 147 mm.

These are example design choices, not recommended joint widths for every installation. Agree the actual joints and perimeter allowances with whoever is installing the tiles. If “100 by 150” already describes the tile itself, do not subtract the joint again.

For a finite row, count the gaps rather than assuming one per tile. Ten 97 mm tiles have nine internal 3 mm joints: 10 × 97 + 9 × 3 = 997 mm, before any perimeter gaps. A framed panel or a replacement tile needs its own measured opening and clearance, not just a nominal tile size.

Also decide which edges you are measuring. The example below assumes the fired clay body's edges define the footprint, with no glaze build-up extending past them. If glaze wraps around the edges, check the final glazed width in your trial rather than treating glaze thickness as clay shrinkage.

Work backwards along both edges

Use a measured linear shrinkage rate: the loss of length between two named stages, not a change in area or volume. For a wet cutting template and a glaze-fired target, you need wet-to-glaze-fired data from the same clay, forming method and intended firing conditions. A supplier's dry-to-fired figure leaves out drying shrinkage.

If you do not have that measurement yet, start with a clay shrinkage test. A bar gives a useful starting rate; a full-size tile trial checks whether that rate works for your slab process.

For each edge, the calculation is:

Cutting dimension = intended finished dimension ÷ (1 − shrinkage percentage ÷ 100)

For this example only, suppose repeated tests support 12% wet-to-glaze-fired shrinkage in both directions. The clay retains 88% of each starting dimension, so divide the fired target by 0.88.

The 97 mm fired width needs 97 ÷ 0.88 = 110.23 mm wet. The 147 mm fired length needs 147 ÷ 0.88 = 167.05 mm wet.

A practical first template might therefore be about 110.2 × 167.0 mm, measured with an ordinary ruler. The decimal places show the calculation, not a promise that soft clay can be cut or fired to that precision. Do not simply add 12% to the target: that uses the smaller, fired dimension as its starting point.

Measure both directions on your trial tiles, and note their orientation in the slab. Do not assume that a good width result proves the length is right. If repeated trials show a different rate along each axis, calculate each dimension with its own tested rate. If results vary unpredictably, stabilise the process before making a more elaborate template.

There is another easy mismatch here: cutting at leather-hard rather than wet. Some shrinkage has already happened. Use a rate measured from that cutting state to the final firing, not the full wet-to-fired allowance. The stage-specific sizing guide explains that distinction.

Make one master template, then check the clay

I would label the template with the intended fired size, actual cutting size, clay body, starting stage and firing conditions. That is more useful than a loose rectangle marked only “tile”. Keep it flat and use material that will not readily swell or distort against damp clay.

Check all four edges and the two diagonals with an ordinary ruler. Opposite edges should match; matching diagonals are a useful squareness check alongside straight edges and correct side lengths. For a printed template, measure the print itself; a printer's “fit to page” setting can undo the calculation.

A shrinkage ruler can help transfer the intended finished dimensions into enlarged cutting distances. Do not apply the allowance twice: either use the matching compensated scale for the 97 × 147 mm fired target, or an ordinary ruler for the calculated 110.2 × 167.0 mm wet dimensions. If the scale's direction is unfamiliar, check its compensated distances first.

Before cutting the batch, make a few pilot tiles: full-size samples to try the whole process. Mark which edge follows the slab's rolling direction so you can compare like with like later. Check their actual dimensions after any edge tidying, at the same moisture state used for the calculation. A reliable template cannot correct a tilted knife, stretched slab or extra trimming.

Drying is a separate part of getting the size right

A tile of the right length can still rock on the table. Shrinkage compensation plans size; it does not guarantee flatness.

Keep slab thickness consistent and avoid bending the tiles when moving them. Aim for even moisture loss rather than a fast-drying top face and a damp underside. Digitalfire's tile-making guidance recommends minimal disturbance after cutting, board-supported turning rather than bending, and even drying. It also warns that tiles drying between boards must remain free to shrink without excessive drag.

Lakeside Pottery's drying guide singles out large flat forms as unsuitable for rushing through drying. There is no set number of drying days: thickness, clay and studio conditions all matter. Follow your studio's drying and firing practice, and keep the pilot and main batch processes alike.

Lay out the pilot batch before committing

Take the pilots through the intended drying and firing sequence, including the planned decoration and edge treatment. If you normally bisque then glaze-fire, do both; if you single-fire, test that route. Once cool, record each tile's final width and length, and check straightness, diagonals and rocking on a flat surface. Keep the range of results, not only the average.

Then lay them out without adhesive, using the intended spacers. This catches a problem that one good tile cannot: small errors accumulate across a row. Ten tiles each 0.5 mm too wide add 5 mm to the row when the joints stay unchanged.

If flat tiles finish consistently too small or too large, use their recorded cutting and fired dimensions to recheck the rate along each axis, then recalculate the template. If the sizes scatter or corners lift, investigate forming, drying and firing rather than merely resizing the template. Test a revised template before committing the rest of the clay.

I would keep an approved fired sample beside the labelled template. One records the plan; the other shows what that clay and process actually delivered.