A skateboarder waits at the top of a 2 m ramp. She has gravitational potential energy and no kinetic energy. As she rolls down, her height falls and her speed rises; at the bottom the potential energy is gone and kinetic energy has taken its place. With no friction at all, she would roll up the other side to exactly 2 m, stop, roll back, and keep going forever. The total, potential plus kinetic, never changes; it only shifts between the two.
In real life she reaches a little less than 2 m each time and eventually stops. The missing energy has not vanished: friction in the wheels and air resistance have turned it into thermal energy, slightly warming the wheels, the ramp and the air, plus a little sound. Count those forms too and the total is exactly what she started with. That is the law: energy changes form and moves between objects, but the total in an isolated system stays constant.
Conserving the total is not the same as conserving one kind of energy. When a lump of clay hits the floor and sticks, its kinetic energy is gone, but the total energy is not: the clay and the floor are slightly warmer and the clay is squashed. A common error is to set the kinetic energy before equal to the kinetic energy after in every problem; that works only when nothing turns motion into heat or deformation.
Think of it like this. Pouring water between glasses is a close match: you can pour from the 'height' glass into the 'motion' glass and back, and some splashes onto the table as heat, but count the puddle too and the total amount of water never changes. Unlike water, energy is not a substance; it is a quantity you calculate.
Core idea. In an isolated system the total energy never changes: energy only moves between objects or changes form, so the energy counted at the start equals the energy counted at the end, heat included.