Search techniques vary according to who you are looking for. A significant proportion of people 'missing' have some form of mental difficulty (frequently suicidal). In this case statistics are a fairly good guide to likely search areas. Police information also helps hugely, a large number of 'missing' people have gone missing before and you may have an idea where they are likely to be.
Searching in an urban area is very difficult (lots of people, lots of obstacles to avoid, lots of cover). The large number of people in the area means that, generally, people are not 'lost' for too long.
People that are mentally sound fall into two categories, those that are lost and those that are injured/incapacitated. The first category is, generally, easier as they want to be found and quite often make themselves seen fairly well (by waving etc). Hand held GPS and experienced local MRT members also mean that accurate positions are not uncommon. The second category is exceptionally difficult. The vast majority of searches in this category are planned 'on the hoof' with a little guesswork involved (land based teams can use statistical probabilities to improve chance of detection - air units can also do this but, in my experience, it is too time consuming and the time is better spent out looking).
In all this I have used lots of words like 'generally' and 'often' because it is not an exact science and people turn up in the strangest of places.
Add to this, the RAF also has the addition of FLIR and TV which adds to the confusion. Should the search be conducted visually at low level or in the hover at 1000' using the camera? Does this change if the person is suicidal (the noise can sometimes be a deciding factor). What about at night? Is it best to use FLIR or NVG?
There are loads more factors that affect a search including weather/equipment that the casualty has/experience etc.
So many decisions, so little time!!