To be regarded as an invention, an idea needs to include an inventive step. An inventive step must be non-obvious - that is, it would not readily occur to an expert in the relevant technology.
The word ‘obvious' comes from the Latin term for ‘upon the road' (ob via), and in the sense of inventions it means something that would be the next logical step along your path from the problem to the solution.
Judging what might be obvious can be very difficult. Many inventions involve combining equipment (for example, fitting a miniature torch to a key-ring). The result of such combinations might be a new product, but its properties or functionality might be entirely predictable as soon as one knew its components. As such, it could be considered obvious.
A product in which one component has been replaced for a different one with equivalent properties could be considered to be obvious (for example, a small metal spring is replaced with a rubber cone).
In another situation there might be a new problem which can be solved with a well known piece of equipment: the ‘novel' process for solving this problem might be considered obvious if there was only one solution to the problem, and it would be known to the typical technician facing the problem (the so-called ‘person skilled in the art').
On the other hand, when components are combined to make a product or process with properties which are greater than the sum of its parts, or better than expected, then that could be a non-obvious invention. Or an invention could come from where there are many possible solutions to a problem, but the inventor has had to research and select the best one. Or an inventor might defy some technical prejudice and solve a problem by doing something every other expert had previously believed would not work.
When it comes to prior art searching, many inventors simply scratch the surface. By far the most common mistake they make is to assume that their idea is novel, when a simple patent search would tell them that it was not. They then make things worse by spending often large sums of money on an idea that is extremely unlikely to be commercially successful.
For example, the inventor of an odourless toilet bowl visited many companies with a complete working toilet, performing 1600 demonstration flushes. But he had never done a patent search. Eventually he found a company interested enough to do its own patent search. They soon found so much prior art that it became obvious that the idea was unprotectable. The inventor threw away his toilet.