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 itself barely absorbs anything, but the grout joints between the tiles are porous, and the corners flex every time the house moves, so what you have on your wall is basically a very stylish colander.
That is fine. It is supposed to happen. The point of shower waterproofing is not to stop water reaching the wall, it is to catch the water behind the tile and guide it back to the drain before it gets friendly with your framing. Water always finds a way through. You do not beat it, you just give it a polite exit.
The layer that catches it is a waterproofing membrane, and 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 have probably seen. A liquid membrane gets rolled on in two coats to a set thickness, and yes, that thickness is supposed to be measured with a gauge, not judged by vibe.
Both work. Both fail the same way when they are rushed: thin spots, seams that do not overlap, corners nobody bothered to reinforce. A proper shower waterproofing membrane meets a standard called ANSI A118.10, which is 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, and it is 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 is 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 will not 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 is a screen door on a submarine. I have 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 is not a regrouting job, it is 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 is waterproof by itself. Both are correct. Neither is optional. If a shower quote does not name a waterproofing product anywhere on it, ask which one they are 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. A traditional mortar bed gets floated and sloped by hand: a pre-slope goes down first, then the liner, then the mortar bed on top. Or you drop in a pre-formed foam pan that shows up already sloped, bond it down with thinset, and tie it into the same manufacturer's membrane. Foam is faster and much harder to botch. Mortar is more flexible for odd shapes and big tile. Both are fine, and how to install a shower pan properly comes down to doing whichever one you picked completely, not halfway.
Then there is 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 is lying dead flat, so water pools on it and just stays. That is why so many 1980s showers start to smell long before they visibly leak. The shower is not dirty. The shower has a water feature you did not 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 have 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, holds its shape, needs no sealing, and comes in every look going. Ceramic is cheaper and perfectly happy on a wall. Natural stone, like marble or travertine, looks incredible and is high maintenance: it is porous, it needs sealing, and it reacts badly to acidic cleaners, which is the polite way of saying it can be quietly ruined by one enthusiastic person with a spray bottle. Stone is a real choice, not a mistake, but 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 demand a genuinely flat wall, because a 24 by 48 inch tile shows off every hollow behind it and turns any unevenness into visible lippage, which is the little step you feel where one tile edge sits proud of its neighbour. Big tile is really a flat-wall project wearing a nicer outfit. If the walls are out of true they get made flat first, that is real work, and it belongs in the estimate instead of turning up as a surprise on day three.
On the shower floor, small tile wins, and not for looks. A 12 inch tile cannot follow a floor that slopes a 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 is the actual reason shower floors are mosaic. It is 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% mortar coverage behind tile in a wet area, fully supported at the corners and edges. Gaps behind shower tile trap water and are exactly where tiles crack when somebody 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 is completely invisible once the job is done, which is precisely why it is 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 am 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, which 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 is 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 nothing, because it arrives as one sealed box that bonds straight into the wall membrane, instead of five separate pieces that each want babysitting. Building one in place is completely doable. It just multiplies the number of joints somebody has to get perfect, in a corner, upside down, at four in the afternoon when the coffee has 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 is 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 entire job. It is a grown man watching water refuse to move. Nothing happens, which is the entire point, because it is the one chance to catch a mistake while fixing it is free, rather than after it is sealed under 60 square feet of tile you have already paid for. It is also the day I am most likely to be caught standing in a bathroom, staring meaningfully 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 pitching the top of the curb slightly back into the shower so water runs the right way, because a curb that drains onto your bathroom floor is not a curb, it is a ramp. Tiling around a shower valve means cutting the tile to what the trim plate will actually cover, not to the rough opening, 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 is required and the ceiling needs its own gentle slope so condensation runs off. Everywhere else it is purely a design call. If you do it, carry the waterproofing across the ceiling as well, 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 is odd for a tile guy to put this on his own website. If your grout is just discoloured but still solid, and the tile does not budge when you press on it, you do not 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 is a Saturday, not an invoice.
If one tile has popped loose in an otherwise healthy shower, that is a small repair. I will happily do it, but plenty of people can, and it does not need a specialist showing up with a truck.
And if you are 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 are not flat, because those are the jobs where five things have to go right at the same time.
If you are collecting bids, ask every installer two questions: which waterproofing system will you use, and will you flood test it. You will learn more from those two answers than from staring at the price at the bottom. That is not a sales pitch. That is just how you find the person who is going to do it right, whether or not that person is me.