Think of an assembly line with three stations. Station h adds 1, station g squares, station f doubles. Feed in 3: h gives 4, g gives 16, f gives 32. In notation this is f(g(h(3))), written (f∘g∘h)(3). The function written last, h, is the first to touch the input, because it sits innermost in the brackets.
Reorder the stations and the product changes. h(g(f(3))) doubles first to get 6, squares to get 36, then adds 1: 37. Composition is not commutative, so f∘g and g∘f are usually different functions. You can also turn the chain into one formula by substituting from the inside out: g(h(x))=(x+1)2, so f(g(h(x)))=2(x+1)2. At x=3 that gives 2⋅16=32, matching the station-by-station answer.
Composing is not multiplying. (f⋅g)(x)=f(x)⋅g(x) evaluates both functions on the same input and multiplies the results: with f(x)=2x and g(x)=x2, (f⋅g)(3)=6⋅9=54. Composition feeds one output into the next function instead: (f∘g)(3)=f(9)=18. The small circle ∘ means feed in; a dot, or no symbol at all, means multiply.
Think of it like this. Nested gift boxes: you put the gift in the smallest box first, then that box into a bigger one, then into the biggest. The innermost box is handled first even though, in f(g(h(x))), it is written last. Unwrapping runs the other way, outer layer first, which is how an inverse undoes a chain.
Core idea. A composition chain feeds one function's output into the next, evaluated from the innermost function outward.