Maxwell's equations imply a wave equation for light that has a speed in it
$$ c = \frac{1}{\sqrt{\epsilon_0\mu_0}} \approx 3.0 \times 10^8 \,\mathrm{m/s} \,.$$
That speed must be measured relative something. The initial thought (consistent with the treatment of waves on a string, waves on the surface of a liquid and sound) it that it is measured relative a medium (dubbed the luminiferous ether).
The interferometry experiment, seems to imply the non-existence of such a medium (which has been confirmed to much high certainty since then) or some other moderately exotic options such as having Lorentz contraction with the ether.
By itself the Michelson-Morley result doesn't prove much.
Einstein sought and found a theory that encompasses the usual Galilean relativity at low relative velocities and the relativity implicit in Maxwell's equations (which had been found by Lorentz) in one neat bundle. That kind of neatness is appealing, and when the ability to test it came around it passed with flying colors.
In principle people could have tried to figure out how to get along with two relativities, but then you have endless corner cases. Frankly that is one reason neatness is desired when possible.
The reason the desire survived is, of course, because the Special Relativity agrees with the world we live in (up to the need for General Relativity to explain, for instance, the gravitational red-shift).