preconcentration
Preconcentration is the trick of packing the thing you want to measure into a smaller space so there is more of it per drop, making a faint signal strong enough to see. It is like boiling a big pot of stock down to a thick, intense reduction: the amount of flavour hasn't grown, but it is now crammed into far less liquid, so every taste hits harder.
In analysis, preconcentration raises the concentration of a trace analyte before measurement — by evaporating off solvent, trapping the analyte on a solid and releasing it into a tiny volume, or extracting it into a small amount of another liquid. A litre of river water might be reduced to a millilitre of extract, multiplying the analyte's concentration a thousandfold.
It matters most in trace analysis, where the analyte would otherwise sit below the instrument's limit of detection — too faint to register against the noise. The honest caveats: concentrating the analyte usually concentrates the interferences and contaminants too, and any losses or contamination introduced during the step are magnified by the same factor as the signal.
To detect a pesticide present at a few parts per trillion, an analyst passes 1 L of water through a cartridge and elutes the trapped pesticide into 1 mL of solvent — a 1000-fold concentration that lifts it above the detection limit.
Crowd a trace analyte into less volume to lift it above the noise.
Preconcentration is the mirror image of dilution: dilution spreads the analyte thinner to bring a too-strong sample down into range, while preconcentration crowds it tighter to bring a too-weak one up. Many methods use both in turn.