Your tile is not waterproof. Neither is your grout. Sorry.
Let me get the bad news over with. Water goes straight through a tiled wall. The porcelain barely absorbs a thing, but the grout joints are porous, and the corners flex every time the house moves. So what you've got on your wall is basically a very stylish colander.
That's fine. It's supposed to happen. The point of shower waterproofing isn't to stop water reaching the wall. It's to catch the water behind the tile and guide it back to the drain before it gets friendly with your framing. If Shawshank taught us anything, it's that water finds a way out over time. You don't beat it. You just give it a polite exit.
The layer that catches it is a waterproofing membrane. There are two kinds worth using. A sheet membrane bonds to the wall with thinset (thinset is the mortar that glues tile down) and overlaps at the seams. Schluter's Kerdi is the orange one you've probably seen. A liquid membrane gets rolled on in two coats to a set thickness. And yes, that thickness is meant to be measured with a gauge, not judged by vibe.
Both work. Both fail the same way when they're rushed: thin spots, seams that don't overlap, corners nobody bothered to reinforce. A proper shower waterproofing membrane meets a standard called ANSI A118.10, the test it has to pass before anyone trusts it under tile in a wet room. The Tile Council of North America writes the handbook these systems come from, and any installer worth hiring is following it whether they name it or not.
One last thing, the part skipped most often. Pick one manufacturer's system and use it the whole way through. A sheet membrane on the walls, a rival brand's liquid on the pan, and a third company's tape in the corner is not one waterproofing system. It's three of them forced into a group project. And like every group project, nobody takes responsibility, and neither does the warranty.
Cement board is not waterproofing (the mistake that costs the most)
If you remember one line from this whole post, make it this one. Cement board for shower walls, whether the box says Durock or HardieBacker or WonderBoard, is water-durable. It is not waterproof. One syllable of difference, and a genuinely enormous gap in what happens next.
Water-durable means it won't rot, swell or turn to mush when it gets wet, which is exactly why it beats drywall in a bathroom. It does not mean water stops. Water strolls straight through cement board, and through every screw hole you put in it, whistling the whole way.
It's a screen door on a submarine. I've said that to so many homeowners standing in so many gutted bathrooms that I should honestly get it printed on a shirt.
A shower with cement board and no membrane over it is soaking its framing every single time somebody uses it. That can go on for years with no sign at all. When the sign finally shows up, it's not a regrouting job, it's a demolition back to the studs, and the lovely tile you chose rides to the curb in a wheelbarrow.
So think of cement board as the backing, not the barrier. Either the waterproofing goes over it, or you swap it for a foam board that's waterproof by itself. Both are correct. Neither is optional. If a shower quote doesn't name a waterproofing product anywhere on it, ask which one they're using. The answer, and how fast it comes, tells you almost everything you need to know about that installer.
The shower pan: where all the water goes to make decisions
Everything drains downhill, so the pan collects the most water and hands you the most ways to get it wrong. Two things have to be true. The floor slopes to the drain at a steady 1/4 inch per foot. And the waterproofing runs unbroken from the drain up the walls, with no flat seam sitting in the wettest square foot of the room.
There are two honest ways to build a tile shower pan. The old way is a mortar bed sloped by hand: a pre-slope layer goes down first, then the rubber liner, then the mortar bed on top. The newer way is a pre-formed foam pan that shows up already sloped. You bond it down with thinset and tie it into the same brand's membrane. Foam is faster and much harder to get wrong. Mortar bends around odd shapes and big tile better. Both are fine. How to install a shower pan properly just comes down to doing whichever one you picked all the way, not halfway.
Then there's the dishonest way, which I find constantly in older DMV homes: a single mortar bed with no pre-slope under the liner. The liner is the real waterproofing, and it's lying dead flat, so water pools on it and just stays. That's why so many 1980s showers start to smell long before they visibly leak. The shower isn't dirty. The shower has a water feature you didn't ask for.
Match the drain to the system too. A traditional clamping drain has little weep holes that let water in the mortar bed escape. A bonded-membrane system needs a bonding-flange drain instead. Fit the wrong drain for your setup and you've basically built an expensive bathtub with a decorative hole in the bottom.
The best tile for shower walls, and the one that will fight you
Porcelain is the default, and it earns it. Porcelain tile for shower walls soaks up 0.5% of its weight in water or less. It holds its shape, needs no sealing, and comes in every look going. Ceramic is cheaper and perfectly happy on a wall too. Natural stone, like marble or travertine, looks incredible but is high maintenance. It's porous, it needs sealing, and it hates acidic cleaners, which is the polite way of saying one keen person with a spray bottle can quietly ruin it. Stone is a real choice, not a mistake. Just agree to the upkeep before it goes up, not eighteen months later when it looks tired.
Large tiles for shower walls mean fewer grout lines, which is usually the whole reason people want them. They also need a genuinely flat wall. A 24 by 48 inch tile shows off every hollow behind it and turns any bump into lippage, which is the little step you feel where one tile edge sits higher than the next. Big tile is really a flat-wall project wearing a nicer outfit. If the walls aren't flat, they get made flat first. That's real work, and it belongs in the estimate, not as a surprise on day three.
On the shower floor, small tile wins, and not for looks. A 12 inch tile can't follow a floor that slopes 1/4 inch per foot toward a centre drain without being cut into little wedges. A 2 inch hex bends around that slope with no drama. That's the actual reason shower floors are mosaic. It's geometry, not fashion, and the extra grout lines give your feet some grip as a bonus.
Whatever goes on the wall, coverage matters more than the tile does. The standard, ANSI A108.5, wants 95% of the tile back stuck to the wall in a wet area, with the corners and edges fully supported. Gaps behind shower tile trap water, and they're exactly where tiles crack when someone leans on the wall. You hit that number by back-buttering, which means spreading mortar on the back of each tile as well as on the wall, and by using the right size trowel. It's completely invisible once the job's done, which is exactly why it's the first corner a rushed installer cuts.
A shower niche is a hole in your waterproofing, on purpose
This is where things get niche. I'm sorry. It had to happen somewhere, and better here than in the pan section.
A niche is a hole you deliberately cut into a wall you just spent a day waterproofing. That makes it the single most likely leak point in the whole shower. It has five inside surfaces and eight inside corners, and every one of those corners has to be sealed and reinforced before a single tile goes near it.
The bottom shelf should tip forward about 1/8 inch so water runs out instead of sitting at the back. Tiny detail, huge difference. It's what separates a niche you wipe clean in two seconds from one that grows a permanent grimy tideline no matter how hard you scrub.
A pre-formed foam niche makes a waterproof shower niche far more reliable than building one from scratch. It arrives as one sealed box that bonds straight into the wall membrane, instead of five separate pieces that each need babysitting. Building one in place is completely doable. It just adds more joints somebody has to get perfect, in a corner, upside down, at four in the afternoon when the coffee's worn off.
Sort the layout before the walls close up. A niche centred between the studs, lined up with your grout lines, and sized so no skinny sliver tiles land inside it, looks deliberate and expensive. A niche shoved in wherever it happened to fit looks like an afterthought forever, and no amount of pretty tile rescues it.
The order it actually happens in, start to finish
Framing and blocking first. Blocking is just extra wood added inside the wall anywhere a bench, a niche or a grab bar is going, so there's something solid to fix them to later. Then the backing board goes on, fastened at the spacing the manufacturer specifies, not at the spacing that gets everyone to lunch faster.
Then the pan, then the waterproofing, joined into one continuous skin that runs from the drain, up the walls, and over the curb. Then the flood test: plug the drain, fill the pan, leave it 24 hours, and check the ceiling below for surprises.
The flood test is the most boring day of the whole job. It's a grown man watching water refuse to move. Nothing happens, and that's the entire point. It's the one chance to catch a mistake while fixing it is still free, instead of after it's sealed under 60 square feet of tile you've already paid for. It's also the day I'm most likely to be caught standing in a bathroom, staring at the floor, apparently doing nothing.
Then tile. Dry-fit the layout first so the cut tiles land where you want them, not wherever gravity and optimism drop them. Walls before floor, as a rule. Tiling a shower curb means tipping the top of the curb slightly back into the shower so water runs the right way. A curb that drains onto your bathroom floor isn't a curb, it's a ramp. Tiling around a shower valve means cutting the tile to what the trim plate will actually cover, not to the rough hole behind it, so the round cover plate hides the gap instead of leaving a little cave around the handle.
Grout once the thinset has cured, not the same afternoon. Then silicone, not grout, in every spot where two planes meet: the inside corners, the joint where the wall meets the floor, and around the niche. Those joints move a hair. Grout is rigid. Rigid stuff in a moving joint cracks, which is less a tile problem than a physics problem, and physics has never once agreed to meet me halfway.
As for tiling the ceiling of a shower, you only have to in a steam shower, where it's required and the ceiling needs its own gentle slope so the condensation runs off. Everywhere else it's purely a design call. If you do it, carry the waterproofing across the ceiling too, and use small tile, because holding a big heavy tile over your head is a job for a younger man with better shoulders than mine.
When you should not hire me
Genuinely, and I know it's odd for a tile guy to put this on his own website. If your grout is just discoloured but still solid, and the tile doesn't budge when you press on it, you don't need me. You need a stiff brush, an oxygen bleach cleaner, and maybe to have the silicone in the corners cut out and redone. That's a Saturday, not an invoice.
If one tile has popped loose in an otherwise healthy shower, that's a small repair. I'll happily do it, but plenty of people can, and it doesn't need a specialist showing up with a truck.
And if you're patient, reasonably handy, and tiling a simple alcove with a foam pan and a sheet membrane, you can absolutely do this yourself. These systems are designed to be installed by careful beginners, and the manufacturers publish genuinely good step-by-step guides. The Ceramic Tile Education Foundation exists for exactly this. Where DIY tends to come apart is curbless entries, steam showers, benches, and large format tile on walls that aren't flat, because those are the jobs where five things have to go right at the same time.
If you're collecting bids, ask every installer two questions: which waterproofing system will you use, and will you flood test it. You'll learn more from those two answers than from staring at the price at the bottom. That's not a sales pitch. That's just how you find the person who's going to do it right, whether or not that person is me.