I use a four stage basic process.
1. Direction - what is the key question I want answered. for example: "What were the richest gold bearing areas around Smithville"
2. Collection. This requires a collection plan, usually run through excel with a list of sub-questions and where I will collect the information from. Example:
a. What are the historical records for gold recovery, type of gold, associated minerals?
b. What were the mining stages - alluvial to deep lead, alluvial only, deep lead only, rush details (numbers, dates, length of rush) etc
c. Land status. Crown, private, on the exempt list, state park, national park, heritage listings etc
d. Mapping. Google earth, Bonzle, Memory Map, Forestry Maps, hard-copy, historic (published and unpublished)
e. Indicators. What is the geology of the area. What types of soils, minerals, vegetation, *****, faults, water etc was associated with gold.
f. Recce of area by vehicle/foot.
Check local library for hard-copy reference material, Trove (National Library of Aust) for electronic on-line records, State Library (ditto), various Government Depts, local museum and/or historical society, google, private hard copy library and electronic library checked etc etc.
The collection list also involves the dates and places to be or have been visited and/or contacted so there is no double dipping of questions.
3. Processing. Once the information starts to pile up, it is necessary to compile the data into a semblance of order so that the original questions can be answered. For example, a Mining Company Report alone may have answers to numerous questions in their background research because they want to know the same things, who/what/when/where/why/how - the six interrogatives. I also asked the question when reviewing material 'so - what? what does this mean?'
4. Dissemination. A final dot-point of the relevant data I have found which is then checked against the first question to determine if I have answered it or not.
Then the cycle goes around the circle again. Sound complex? It is not - its the same process we would use when buying a car or house, or looking for a fishing spot etc. I need to buy a car - what type? Collect brochures and web data. Collate all the info. make a decision.
Any sort of detecting requires a methodical approach to determine best options. If you fish, you would not use a shark rig to try and get bream from a tidal estuary. Same with detecting - right information for the right spot to get the best result. A desk recce will however only answer some of the questions. Time spent on the ground is never wasted.