Compaction is taking the air out, nothing more mysterious than that
Pour a bucket of stone out on the ground and you have a pile that is part stone and part empty space. Push on it and it gets shorter. That shortening is the settlement your slab is going to experience, and the only real question is whether it happens under a machine this week or under your concrete over the next ten years.
Running a compactor over the fill does two things at once. It drives the air out, and it works the angular pieces of crushed stone into each other so they lock together and stop sliding. That interlock is why crusher run makes a good base and why rounded material and dirt do not. Once it is tight, it behaves less like a pile of loose pieces and more like one firm layer.
Moisture is part of it too. Fill that is bone dry will not tighten properly and fill that is soaked cannot be tightened at all, because water does not compress and it has nowhere to go. That is why we will stop and wait on a base after a hard rain rather than run a machine over mud and call it done.
Layers versus one deep dump
This is the fork that separates a base that holds from one that only looks like it does. A compactor works down into the fill from the top, and its reach is limited. Stone brought up in shallow lifts, with each lift compacted before the next goes on, ends up tight all the way through. The same quantity of stone dumped in one deep layer and rolled once ends up tight on top and loose underneath.
From the surface, those two bases are indistinguishable. Both are flat. Both will hold a boot. Both will hold forms. One of them has several inches of air still in it waiting to be squeezed out by a parked truck, and you find out which one you got about two winters later.
That is the honest reason to care about a step you will never see. It is not that anyone doubts the concrete. It is that this specific shortcut is invisible on the day, cheap to take, and expensive for you alone.
Compacted stone versus putting the dirt back in the hole
The other version of the same shortcut is backfilling with whatever came out of the excavation. Around here that usually means red clay, and clay is the worst thing to build a slab base out of. It holds water instead of draining it, it swells when it takes water on and shrinks when it dries out, and it does both of those things through every freeze and thaw the winter brings at this elevation.
So we strip the topsoil and the soft spots out, get down to firm ground, and bring the grade back up with stone rather than with the excavated material. The stone does two jobs. It compacts to a firm, even bearing layer, and it lets water pass through and move away instead of standing under your concrete.
There are ground conditions where the answer is different, and we say so on site. But raw clay graded flat and poured on is a slab that is going to move, and no thickness of concrete and no amount of steel makes that a good plan.
What skipped compaction looks like two years on
You do not see it right away. New concrete looks the same either way for the first season, which is what makes this worth explaining before the pour rather than after. What eventually shows up is a particular kind of failure that is easy to recognize once you know it.
A crack that runs across the middle of a slab with no joint anywhere near it. A corner that has dropped so a shovel catches on the lip. A walk panel that rocks under your weight and squirts water out at the joint when it rains. A low spot in a driveway that holds a puddle after every storm and holds ice all of January. None of those are concrete failing. They are the slab reporting on the ground under it.
The repair is also worse than the original prep would have been. Getting a base right costs machine time and stone. Getting it right after the fact costs the same machine time, the same stone, plus breaking out and hauling off perfectly good concrete first.
Proving the base before the forms go down
The check is simple and physical. We proof roll the pad with weight on it and watch for anywhere the ground flexes, ruts or pumps water up. A spot that moves under a loaded wheel will move under a slab, and it gets dug out and rebuilt before anything else happens. Grade and fall get set at the same time, so water leaves the finished surface in the direction we chose.
If a pad will not take compaction because there is water in it or the ground below it is too soft, we tell you that when we find it, not after the concrete is on the ground. It is a conversation nobody enjoys and it is the cheapest moment it will ever be had. The base and the prep are written on your free written estimate the same as the concrete is, so you know what you are buying under the part you can see.