Why are the bald spots in those particular places?

Because the pattern is a record of how the floor was worked, and it usually points straight at the cause. Thin areas at the far edge of each throwing position mean the flake was not reaching, which is a throwing distance and technique problem. Thin areas near walls, columns and doorways mean the awkward spots got less attention than the open floor, which is the most common version of all. A floor that goes progressively sparser as you follow the sequence of work means the base coat was setting under the crew, so later areas took less flake because there was less wet resin to hold it. And a floor that is sparse only at the very end of the sequence usually means somebody was rationing what was left in the box. So before deciding anything, work out the order the floor was laid in and see whether the thin areas line up with it.

How do I tell a broadcast problem from flake that came off later?

Look at the edge of the thin area rather than the middle of it. A broadcast problem fades. Density drops gradually from properly covered to sparse, because a thrown handful lands heaviest in the middle and thins towards the edges of its reach. There is no line to point at. Flake that was there and then left has a defined edge, and there is usually loose flake or a rough patch around it. That is a bond or a wear problem rather than a coverage one, and the question becomes why it released rather than why it was never there. The two also differ in what is underneath. A bald spot from a bad broadcast has clean, intact base coat in it, because nothing ever covered that resin. A spot that lost its flake usually has damaged or pitted base coat where the flake pulled out of it.

Is this a material problem or a technique problem?

Technique, in the large majority of cases, and it is worth saying plainly because the material is usually blamed first. Flake is inert. It does not settle, separate or lose potency, and one box of a given blend behaves like the next. What varies enormously is how it is thrown: the height it is released from, the angle, whether it is flicked upward to fall as rain or thrown outward to land in a clump, how much is taken per handful, and whether the person covering a bay overlaps their positions or works from a standing spot. The one genuine material contribution is the flake size and the blend. Larger flake is heavier, so it drops faster and travels less far for the same throw, and a floor broadcast with a large blend using technique tuned for a small one will read as patchy. That is still a technique correction, just one driven by the material.

Can I go back over the thin areas and add more flake?

Only while the base coat is still wet enough to hold it, and that window is much shorter than people expect. If the resin is still genuinely wet, more flake will seat and the repair is invisible. That is why the last job before leaving a bay should be to walk it in raking light and fill what you find, not to move on and inspect at the end. Once the base coat has begun to set, added flake sits on the surface rather than in it, and the topcoat will not hold it there permanently. You end up with loose flake under the clear coat, which looks worse than the bald spot did and cannot be corrected without removing the topcoat. After the base coat has cured, the honest fix is not more flake. It is to abrade, apply a fresh resin coat over the affected area and broadcast into that, which means blending a repair into a decorative floor and is difficult to make invisible on anything but a busy blend.

Does it matter, or is it only cosmetic?

Cosmetic in almost every case, but on a flake floor cosmetic is the point. The structural work is done by the base coat and the topcoat, and both are continuous across a bald spot. The floor is as strong and as chemically resistant there as anywhere else, and it will wear at the same rate. What you have lost is the thing the customer paid for. Flake is specified for appearance and for the slip resistance its texture provides, and a bald patch has neither. On a decorative garage or a showroom that is a legitimate complaint about the finished article, not a trivial one. The exception worth checking is a floor where the flake was also carrying the slip resistance specification. There a bald area is genuinely a smoother area, and on a wet-exposed floor that is a safety question rather than an appearance one.

How much flake should I be ordering?

More than the arithmetic says, and the reason is exactly this defect. A full broadcast to refusal means throwing flake until the surface will not accept any more, and the excess is swept up and discarded rather than being used. Ordering the theoretical amount guarantees running short somewhere, because the parts of the floor done first will always take more than their share while people find the rhythm. Running short is one of the causes of this defect rather than an inconvenience separate from it. A crew who can see the box emptying starts rationing, and rationing produces exactly the progressive thinning towards the end of the sequence described above. Work the quantity out properly before you start rather than estimating from the last job, size the order with a margin on top, and have the surplus on site rather than at the supplier.

How do I stop it on the next floor?

Light it properly, plan the coverage, and inspect before you leave rather than after. Lighting first, because it is the cheapest and the most often skipped. Flake density is genuinely hard to judge under overhead light looking down. Put a light low and rake it across the floor and thin areas become obvious immediately. Most bald spots were visible at the time and nobody was in a position to see them. Then plan it: broadcast to refusal rather than to a target, throw upward so the flake falls as rain rather than outward so it lands as a clump, overlap the throwing positions rather than treating each as a separate patch, and do the walls, columns and corners first while there is time to be careful with them. And size the working area to the base coat, the same discipline that prevents lap marks. If the resin is setting before you have finished broadcasting into it, the sections are too big.

Flake Bald Spots and Uneven Broadcast

A flake floor is the one system where the defect is usually something absent rather than something present, and that changes how you diagnose it. Thin patches are almost always a record of how the flake was thrown and how much time the base coat gave you, not of the material itself. Here is how to read the pattern and work out which.

Quick answer

This is the one defect on the site where the problem is something absent rather than something present, and that changes the diagnosis. The film is fine. What is missing is flake, and the pattern of what is missing is a record of how the floor was worked: how far somebody could throw, which corners got the least attention, and how much time the base coat gave them before it stopped accepting anything. Work out the order the floor was laid in, and the thin areas usually line up with it.

Read the pattern first, not the patch

Almost every bald spot is explained by where it is rather than by what it looks like. Before deciding anything, establish the sequence the floor was laid and broadcast in, then see which of these it matches.

Thin at the edge of each reach

Sparse bands at a regular distance from wherever somebody was standing
Cause: throwing distance. A handful lands heaviest in the middle and thins towards the limit of its reach
Worse with a large flake blend, which is heavier and travels less far for the same throw
Fix next time: overlap the throwing positions instead of treating each as its own patch

Thin at walls, columns and doorways

The open floor is fine and the awkward parts are sparse. The most common version of all
Cause: the difficult areas got the least attention, usually because they were done last and quickly
Fix next time: do the edges, corners and obstructions first, while there is still time to be careful with them

Progressively sparser along the sequence

The floor gets thinner the further through the work you follow it
Cause: either the base coat was setting under the crew, or somebody was rationing what was left in the box
Those two look identical, and both are real. Check what was ordered against what the floor needed
Fix next time: smaller sections, and order with a margin

Was it never there, or did it leave?

This is the one distinction worth being certain about, because it sends you to a different page and a different repair. Look at the edge of a thin area rather than the middle of it.

A broadcast problem fades. Density falls gradually from properly covered to sparse, with no line you could point at. The base coat inside the thin area is clean and intact, because nothing ever covered that resin
Flake that left has an edge. There is a defined boundary, usually loose flake or a rough patch around it, and the base coat underneath is often pitted or damaged where the flake pulled out of it. That is a bond or wear question, and peeling is the page for it

It is almost always technique, and that is not a criticism

Flake is inert. It does not settle, separate or lose potency, and one box of a blend behaves like the next. What varies enormously is how it is thrown: the height it is released from, whether it is flicked upward to fall as rain or thrown outward to land as a clump, how much is taken per handful, and whether somebody moves through the bay or covers it from a standing position.

The one genuine material contribution is flake size. A larger blend is heavier, drops faster and travels less far for the same throw, so a floor broadcast with a large blend using technique tuned for a small one reads as patchy through no fault of the person throwing it. That is still a technique correction, just one the material asked for.

The base coat is the clock

Flake only seats while the resin underneath is genuinely wet. Once the base coat starts to set it stops accepting flake, and anything thrown after that sits on the surface rather than in it. This is the same discipline that prevents lap marks, and it fails the same way: the working time you actually have is much shorter on a warm slab than the figure on the tin suggests.

The practical rule is that if the resin is setting before you have finished broadcasting into it, the sections are too big. Check the working time you really get at the temperature on site with the cure time calculator and size the bays around it rather than around the room.

Order more than the arithmetic says

A broadcast to refusal means throwing flake until the surface will not take any more, and the excess is swept up and discarded rather than used. Ordering the theoretical figure therefore guarantees running short somewhere, because the areas done first always take more than their share while people find the rhythm.

Running short is a cause of this defect rather than a separate inconvenience. A crew who can see the box emptying starts rationing, and rationing produces exactly the progressive thinning described above. Work the quantity out before you start with the flake broadcast calculator, size the order with a margin on top, and have the surplus on site rather than at the supplier. The quantity guide covers why the margin matters.

What can be fixed

The window is narrower than people expect, and it closes with the base coat.

While the resin is genuinely wet: more flake seats properly and the repair is invisible. This is why the last job in a bay should be to walk it in raking light and fill what you find, rather than moving on and inspecting everything at the end
Once the base coat has begun to set: added flake sits on top rather than in the resin, and the topcoat will not hold it there. You get loose flake under the clear coat, which looks worse than the bald spot did and cannot be corrected without removing the topcoat
After cure: the fix is not more flake. It is to abrade the area, apply fresh resin and broadcast into that, which means blending a repair into a decorative floor. Realistic on a busy blend, difficult to hide on a simple one

Does it actually matter?

Structurally, almost never. The base coat and the topcoat do the work and both are continuous across a bald spot, so the floor is as strong and as chemically resistant there as anywhere else.

But flake is specified for appearance and for the texture it gives underfoot, so a bald patch has lost exactly what it was bought for. On a showroom or a decorative garage that is a legitimate complaint about the finished article rather than a trivial one.

The exception worth checking is a floor where the flake was also carrying a slip resistance requirement. There a bald area is genuinely a smoother area, and on anything that gets wet that is a safety question rather than an appearance one.

Preventing it on the next floor

Light it properly, and do it low. Flake density is very hard to judge under overhead light looking down. A light raked across the floor makes thin areas obvious instantly, and most bald spots were visible at the time with nobody positioned to see them
Broadcast to refusal, not to a target. Throw until the surface stops accepting flake rather than until a measured amount is gone
Throw upward, not outward, so the flake falls as rain and lands evenly instead of arriving as a clump with sparse ground around it
Do the hard parts first. Walls, columns, doorways and anywhere awkward, while there is still time and material to be careful
Overlap the throwing positions rather than treating each standing spot as its own patch of floor
Size the sections to the base coat, and walk each one in raking light before leaving it rather than inspecting the whole floor at the end