The peaks in a XRD pattern represent the crystallographic planes, like 111, 100, 200, etc. The position of each peak is determined by the Braggs' Equation:
n . Lambda = 2 . d . sin(theta)
Each crystallographic plane has a interplanar distance (d), and with the wavelength used by the equipment, you can solve this equation for theta and find the position of each peak.
If your peaks shifted, it means that the values of interplanar distances (d) changed. So, your unit cell dimensions changed, and if the peaks shifted to the left, it means that your unit cell got bigger.
you performed your relative measurements on two different samples;
so, in addition to the answer of Ricardo Tadeu Maia ,
you should consider, that a decrease of the sample height (sample surface relative to the focal point of the goniometer) will also cause a peak shift to the left.
So you should check your height adjustment in order to judge whether the peak shifts*) are due to a change of sample property or simply due to sample height misalignment...
I would like to encourage you to do a very instructive experiment by performing some mutual height misalignments in order to 'see' the magnitude of the peak shift(s) as a function of sample height changes...
From a simple geometrical sketch you will find such a dependence as: