Start with y=x2, whose lowest point is (0,0). Now look at y=(x−3)2+1. The squared part (x−3)2 is zero when x=3 and positive everywhere else, so the smallest possible output is 0+1=1, reached at x=3. The vertex is (3,1), found just by asking what makes the bracket zero.
That is why h comes with a minus sign. In (x−h)2 the bracket is zero when x=h, so (x−3)2 puts the vertex at x=3, while (x+3)2=(x−(−3))2 puts it at x=−3. Subtracting inside moves the graph right: the input has to be 3 larger to produce the output x2 gave at 0.
The number a sets the shape. With a=2, outputs grow twice as fast as for x2, so the parabola is narrower; with a=−21 it opens downward and is wider, and the vertex becomes the highest point. In f(x)=−2(x−4)2+9, the squared part can only take away from 9, so 9 is the maximum, at x=4.
Factored form, a(x−r1)(x−r2), also has brackets of x minus a number, but those numbers are roots. In 2(x−1)(x−5) the graph crosses the axis at 1 and 5, and the vertex is halfway between, at x=3. Vertex form has a single bracket, squared, with a constant added outside it.
Think of it like this. Vertex form is a set of directions for placing a stamp: pick up the basic x2 shape, slide it h to the side and k up, then stretch it by a, flipping it over if a is negative. Where the stamp's tip lands is the vertex. The one twist is that the sideways direction is the opposite of the sign printed in the bracket.
Core idea. Vertex form a(x−h)2+k puts the parabola's vertex (h,k) right in the formula.