Methods of spectrophotometric Analysis one of the easiest and least expensive methods of chemical analysis. How to increase its efficiency and reduce its cost
Also use a more 'sensitive' wavelength. An example are 'molybdenum blue' methods which are normally quantitated at 680 nm but can also be quantitated at 1000 nm (?) which is outside the wavelength range of most spectrophotometers.
To reduce cost, prepare the reactants yourself. For example, If you prepare the solutions you need for COD (the colorimetric closed reflux method) the cost of reactants per sample is about 8X less than if you buy the prepared KIT. Also, try to use preservation techniques in the samples for the parameters you want, so that you run the maximum possible number of samples at once.
Scale your reagent usage. Most of the colorimetric methods are written in terms of 100 or 200 mL of sample with the appropriate amount of reagent. Scale that back to 10 or 20 mL; a standard spectrophotometer cuvette only holds 3-4 mL. Also, get a good UV/visible system; good systems will have 10 times the sensitivity of cheap spectrophotometers and you'll have more ability to change the size of the cuvette.
You can also maximize the concentration of your reagents. Many commonly prescribed reagents are for pure standards and not samples. Thus, always perform a 'spike and recovery' experiment.
As Mark suggested, use a good UV-VIS spectrophotometer. The sensitivity also depends on the type of detector that is used in the instrument. For example, a PMT detector is much more sensitive than a Silicon Photo diode detector. You can also use long path length cuvetts to increase the absorbance in case the target concentration is very low.