When we talk electric power, we think in terms of load. We exercise energy on a load, which might be: pure resistive (like in a boiler resistance) inductive (as in an electric motor) and capacitive (some ultrasound generators).

Applying the driving voltage on the load, there will be a certain current, determined by the voltage and the impedance of the load (resistive, inductive or capacitive).

Considering an ac voltage, say the good old 120V, a sinusoidal wave. It swings from zero to a max positive value, back to zero, to a max negative value and back to zero in one period, or 360°). There are, in the USA, 60 such periods for the residential power. In a resistive load, as the voltage increases (the first quarter of one period, or 90°), the current increases the same way, in the same time. The voltage peaks at the 90° and so the current. We know that the power is the product between voltage and current (P=UxI) so, at 90° the power is maximum.

An inductive load will oppose to the increase of the current, so when voltage reaches its maximum value at 90°, the current has yet to come to its maximum value (and this depends on the magnitude of the inductance, the residual resistance of the inductance, etc), So, if we take again, the product UxI, it is obvious that the power is not the maximum one, voltage and current do not peak in the same moment. The difference, in degrees, of the voltage and current peaking, called φ will be the subject of the power factor definition, Cos φ. If the difference is φ=0 (resistive load), cosφ=1, power is maximum