Quick Answer

Gloss variation on a polished floor almost always comes from uneven slab density, not from the polishing pad itself. Test gloss at a 60 degree angle in a grid and you will find the dull zones track exactly with patch repairs, older pours, low-density concrete, or spots where densifier pooled or ran dry. Fix it in order: correct the densifier rate first, then re-grind with finer resin pads, then burnish. Do not chase a low reading with more guard coat, because that only hides the problem for a few weeks.

Why One Floor Reads Two Different Gloss Levels

Gloss is a mirror: the smoother and denser the surface, the more light bounces straight back to the meter. Concrete is not a uniform material. A slab poured on Monday may have a different cement content than one poured on Friday, a patched repair uses a different mortar that absorbs water faster, and trowel finishes vary by bay. Densifier then amplifies the difference: porous patches soak up more chemical and come out harder and glossier, while dense patches repel it and stay dull. The diamonds cut both, but they cannot make a soft patch reflect like a hard one.

Machine technique adds a second layer. If the operator leans on one side of the grinder, lets the machine dwell in a corner, or changes traverse speed between bays, the scratch pattern depth changes and gloss drops wherever the cut is shallower. Heavy machines press harder on the outside edge of the pad than the center, which is why the middle of a large burnish often looks flatter than the perimeter.

Reading Gloss Correctly Before You Panic

Use the same angle and the same meter

Measure gloss at 60 degrees, which is the standard angle for polished concrete. Readings at 20 degrees are far more sensitive to micro-scratches and make normal variation look dramatic, while 85 degrees is for very matte floors. Calibrate the meter on its black glass standard before each walk, take three readings per spot, and average them. Gloss units on polished concrete typically range from 20 on a matte finish to 80 on a high polish; a normal, well-executed floor varies by about 5 to 10 gloss units between bays.

Map the dull zones on a grid

Lay out a 5 m grid across the floor and mark every reading on a sketch. Dull patches that follow a straight line usually sit over a control joint or a cold joint. Round dull spots with a halo are patch repairs. Large, broad dull areas track with a different pour day or a different trowel crew. Once you see the pattern, the cause is usually obvious, and you stop guessing.

The Most Common Causes and Their Signatures

• Uneven densifier spread: dull streaks where the product dried too fast or was walked through; glossy streaks where it pooled.
• Patch repairs and cold joints: darker or duller islands because repair mortar absorbs densifier differently.
• Machine weight and head pressure: dull bands along walls where the grinder tilted, or glossy ovals where the operator dwelled.
• Traverse speed differences: dull lanes where the machine moved faster and skipped a diamond grit.
• Prior coatings or curing compounds: glossy ghost patches where old sealer was not fully removed before grinding.

How to Correct the Variation

Recheck densifier on the dull zones

Start by rechecking densifier coverage on the dull zones. If the water bead test shows the dull patches still absorb water, a light second coat, scrubbed in and picked up after 15 minutes, raises the density enough to even the reflection. Wait the usual 12 to 24 hours, then resume polishing. If the dull zones are caused by soft, porous concrete rather than missed chemical, no amount of extra densifier will lift them; you need finer resin pads to close the scratch pattern and a longer burnish at moderate speed.

Re-grind the whole bay, not the spot

Re-grind the entire floor at the next finer grit, not just the dull patches. Spot re-grinding creates a brighter island inside a dull field, which reads worse than the original variation. Run the same number of passes, same pad pressure, same speed across the whole bay, then burnish in two directions with a clean white pad. Re-measure the grid after the burnish; a normal correction narrows the spread from 15 gloss units down to about 5.

When Guard Coat Makes It Worse

A guard or sacrificial coat masks gloss variation for a month or two, but as traffic wears it through, the underlying difference comes back stronger because the coat wore thin on the high spots and stayed thick in the valleys. Treat guard as the final uniform layer after the base gloss is even, not as the fix. If the client demands a mirror finish over a naturally uneven slab, agree on the tolerance in writing on the sample panel, because no system can make a soft patch reflect like a dense one.

Common Glossing Mistakes

The number one mistake is measuring gloss only in the brightest showroom corner. That corner has the right light, the right slab, and the right crew, and it tells you nothing about the loading-dock bay poured three weeks later. Walk the whole floor with the meter. The second mistake is chasing one low reading with an extra burnishing pass in a single spot, which creates a hot, over-polished oval that everyone notices from 20 meters away.

FAQ

Is a 10 gloss unit difference across the floor acceptable? On a commercial warehouse, typical tolerance is around 5 to 10 gloss units at 60 degrees and is usually accepted. On a showroom or retail floor where the client expects a mirror, agree on the exact number on the sample panel before starting.

Why does the floor look glossy in the evening but flat at noon? Angle and ambient light change perceived gloss far more than the meter does. Judge the floor under the lighting it will actually live under, and take gloss readings in the same direction the lights run, because light traveling along the scratch pattern reads brighter than light crossing it.