When we start MDing, we usually peek a detector and then swinging it more or less by chance. There is nothing wrong with that, but in finding things it is helpfull to be more methodical and systematical. I will not discuss the research beforehand, since this has been done many thimes elswhere, but rather more tecnical aproach to the problem. Take this as my personal advice, if you like.
First, one should focus on target and write down two separate lists; one list is what is known about it, such as its size, depth range, material, magnetic and electric properties, surrounding material (soil), origin, time period of production etc. This will ofently be an empty list. The second one is what it is expected about the target. Though both lists contains same elements, one should not confuse the two of them.
Second, one should focus on search time frame and expected disturbances; can that search be done anythime or in specific time frame(s) only, will it be an open search or a more hidden one etc.
Third, the terrain itselfe; how big will be the area to be searched, the vegetation and relieph of it, the visible obsticles, naihgbourhood and the EMI, type of soil, humidity of soil and mineralization, resistivity of soil, realtive ease or difficulty to dig it, homogenity and uniformity of soil deposits where the target is expected, homogenity of soil electric resistance in the area of interest, presence of other metal targets in the area, their expected size and comosition, .... to name a few.
When all that is considered, then it is the time to peek the right equipement. No metter what you do, keep some form of written documentation of what, where and when you did it, for your own refference.