Good tile placement is what separates a floor that feels intentional from one that always looks a little off. Most of the result is decided before the first piece goes down: the layout lines, pattern choice, substrate flatness, grout joints, and movement gaps all shape how the surface will look and how long it will last. In this guide I focus on the decisions that matter most for a typical U.S. home floor, with practical notes on what actually prevents lippage, cracks, and awkward cuts.
The strongest tile floor starts with layout, flatness, and patient setting
- Dry-fit first. A few minutes of layout work prevents sliver cuts and awkward thresholds.
- Match the pattern to the room. Straight, diagonal, running bond, and herringbone all solve different layout problems.
- Prep beats patching. Large-format tile needs a flat, stiff base more than almost anything else.
- Joints are structural. Grout joints, perimeter gaps, and movement joints keep the floor from fighting the building.
- Coverage matters. Full mortar support is what keeps tiles from sounding hollow or cracking later.
Build the layout before you mix mortar
I never start by asking where the first tile should go. I start by asking where the room wants the joints to land. A good dry layout keeps cut pieces balanced at doorways, cabinets, and walls, and it saves you from a floor that looks skewed even when every piece is technically square.
My first pass is simple: I find the true center, snap chalk lines, and dry-lay enough tile to see where the cuts land. If one side ends up with a narrow sliver, I move the layout before a single bag of thin-set is opened. I would rather shift the pattern by a couple of inches than hide a one-inch strip at the baseboard or threshold.
- Start from the longest sightline, not just the center of the room.
- Check the doorway and the most visible wall before you commit.
- Leave room for baseboard, trim, or a transition strip at the perimeter.
- Order extra tile for cuts and breakage; about 10% is a sensible baseline for a simple room, and more makes sense as cuts multiply.
That early layout work is what keeps the rest of the project calm, because once the room is mapped the pattern choice becomes a decision instead of a guess.
Choose the pattern that fits the room
The pattern does more than decorate the floor. It changes how wide, long, or calm the room feels, and it changes how hard the installation will be. I choose the pattern to solve the room first, and only then do I think about style.
| Pattern | Best for | What it does well | Tradeoff |
|---|---|---|---|
| Straight lay | Square rooms, minimalist spaces, and tile with strong graphics | Looks clean, uses the least material, and is usually the easiest to keep true | Can feel plain if the room lacks a focal point |
| Running bond | Wood-look planks and rooms that need more movement | Breaks up a long field and hides small size variations well | On long planks, a full half-offset can exaggerate warpage and lippage |
| Diagonal | Small or square rooms that need visual energy | Makes a room feel wider and less boxy | Creates more waste and more perimeter cuts |
| Herringbone | Entries, feature areas, and spaces where the floor should be the visual anchor | Feels custom and adds direction to the room | Demands more planning, more cuts, and a tighter eye during layout |
With plank tile, I stay especially alert to offset. If the tile is long enough to show bow or warpage, I keep the stagger conservative unless the manufacturer says otherwise. The goal is not to chase a trend; it is to make the floor read as intentional when you walk into the room. Once the pattern is settled, the base underneath it becomes the real make-or-break issue.
Make the substrate flat enough for the tile you picked
Tile is unforgiving. The bigger and flatter the piece, the more the substrate matters, because the tile can only follow the surface that supports it. I think about stiffness first, then flatness, then compatibility with the tile and setting materials.
For residential ceramic and porcelain, I still think in terms of a structure that meets tile-specific deflection requirements, commonly expressed as L/360 or better. For large-format tile, especially anything with one edge over 15 inches, I want the floor very close to 1/8 inch in 10 feet and about 1/16 inch in 2 feet or a similar short-span tolerance. That is not perfectionism; it is what keeps edges from rocking and joints from drifting.
When the slab or subfloor is off, I fix the problem instead of hoping mortar will hide it. Grinding high spots, filling lows, and using a self-leveling underlayment or an uncoupling membrane are all valid tools, but each one has a job. An uncoupling membrane is a layer that helps isolate the tile from small movements below; it is helpful, but it is not a license to ignore a floor that is visibly out of plane.
- Remove paint, drywall mud, dust, or curing compounds that can break the bond.
- Patch low spots before setting tile, not after the first row is already stuck down.
- Use an uncoupling system when the assembly calls for it, especially over wood or cracked concrete.
- Back-butter larger tiles when needed; that means skimming a thin coat of mortar on the back so the whole piece is supported.
- Check mortar contact as you work. I want at least 80% in dry interior spaces and about 95% in wet or exterior work.
When the base is honest, setting becomes a control job instead of a rescue operation, and that is a much better place to be.
Set each tile with control, not speed
Once I start setting, I keep the mortar ridges straight and the pace steady so the thin-set does not skin over before the tile goes in. I press the tile into place, then give it a small back-and-forth movement to collapse the ridges and lock the piece into full contact. The only tile that matters is the one under your eye right now, so I check coverage often by lifting a fresh piece and looking at the back.
Leveling clips can help neighboring edges land at the same height, but they are not a cure for a crooked substrate. I use them as a support tool, not a rescue tool. If the floor is flat, they make the finish cleaner; if the floor is wavy, they only make the problem look more organized for a while.
| Tile type | Practical starting joint | Why it matters |
|---|---|---|
| Rectified tile | 1/8 inch | Factory-cut edges are more consistent, so a tighter joint can work when the tile itself is uniform. |
| Calibrated tile | 3/16 inch | A slightly wider joint gives more room for small size differences. |
| Gauged thin tile | 1/16 inch minimum | Only use this when the product and system are designed for it, because the margin for error is tiny. |
I also stay conservative with running bond on long planks. For pieces with one edge longer than 18 inches, a 33% offset is the safer ceiling if the manufacturer allows that pattern at all. A full half-offset can make even small bowing show up as lippage, and lippage is one of those flaws people feel before they can explain it.
That brings the floor from “laid” to “built,” but it still needs the joints that let it move without tearing itself apart.
Grout joints and movement joints keep the floor alive
Grout joints and movement joints are not interchangeable. Grout fills the visible gap and finishes the field, while movement joints absorb expansion, contraction, and small shifts in the building. I never treat one as a substitute for the other.
In controlled interior spaces, I usually expect movement joints every 20 to 25 feet, and I tighten that spacing to about 8 to 12 feet when the floor gets direct sunlight or lives outdoors. I also place them at doorsills, transitions, changes of plane, and anywhere the tile field meets a restraining surface. Those joints get flexible sealant or a purpose-made profile, not cement grout.
At the perimeter, I leave a gap and cover it later with trim or baseboard. If you grout hard into the edge, the floor loses its room to move, and that is how stress gets transferred into the tile field. The same thinking applies to lippage: a flat substrate, a sensible offset, and full mortar support are the real defenses, because no spacer can fix a surface that was already out of plane.
Once those joints are in the plan, the final check is mostly visual and tactile.
The checks I make before I call the floor finished
Before I consider the job done, I walk the room under raking light and look for anything that breaks the rhythm of the surface. Doorways, thresholds, and the first few rows near the entrance get special attention, because that is where a layout mistake usually reveals itself fastest. I also look at the grout lines from standing height, not just kneeling at tile level, because the room has to read well to the person who will live with it.
- Confirm that cut pieces are balanced at walls and transitions.
- Check that grout joints stay consistent through the full run of the room.
- Clean squeeze-up from joints before it hardens and becomes a chisel job later.
- Tap for hollow spots and investigate any large area that sounds loose.
- Verify that perimeter gaps and movement joints were left open where they belong.
If the room is large, the tile is long, or the slab is uneven, I slow down rather than trying to win on speed. Good tile work is mostly disciplined preparation: the right layout, the right substrate, and the patience to reset what does not look right. When those pieces line up, the floor reads as one continuous surface instead of a collection of individual tiles.
