If you think of a 1 dimensional wave (just a sine or a cosine for instance) moving with time, you can think of it really as series of coupled armonic oscillators: you can think of any point of a particular straight line to be "perturbed", thus oscillating vertically, moving about the equilibrium position, as if it was attached to a spring; each of these points, of coordinates $x_i$, would be coupled with its neighbours of coordinates $x_i \pm dx$ in the sense that there is an infinitesimal phase difference between their oscillations.
Now, if you try to describe the wave from a global perspective, you don't care about the motion of each point, but you rather care of the motion of the set of points as a whole. Therefore, at a certain instant in time you can see a series of peaks, that is a series of points that have reached the maximum distance from the equilibrium point.
If we focus on the case of periodic waves (you can think of a non periodic wave as one with a period $T \to +\infty$), then we can define the distance between each of the peak that marks the start of a new period and the one that marks the start of the following period, to be the wavelength $\lambda$. Now the wave pattern moves in time. In particular, the period $T$ of a wave is defined to be the time interval after which a peak at $x_i$ reaches $x_i \pm \lambda$. Hence it is possible to define the phase velocity of a wave (the direction in 1 dimension is obliged and the verse is given by the sign of $v$) to be the ratio between its wavelength and its period:
$$ v = \frac{\lambda}{T} $$
It is important to notice that this velocity isn't referred to any of the points that make up a wave in particular, as none of these moves horizontally at all, for a transverse wave; it is rather referred to the global motion of this set of points, viewed as a whole.
Actually we can say that waves do not carry matter, but rather energy. It is also possible to say that a wave carries information that moves through space-time, therefore you can think of the phase velocity of a wave as the velocity at which this information propagates through time (the information being how much each point at $x_i$ is out of phase with respect to the points at $x_i \pm dx$).
I hope this was clarifying! ;)