Due to the nature of-SSB, in order to properly recover the fidelity of the original audio, a pilot carrier was distributed to all locations (from a single very stable frequency source), such that, the phase relationship of the demodulated (product detection) audio to the original modulated audio was maintained. 
 

Also, SSB was used by the U.S. Air force's Strategic Air Command (SAC) to insure reliable communications between their nuclear bombers and NORAD. In fact, before satellite communications SSB-was the only reliable form of communications with the bombers. 

The main reason-SSB-is superior to-AM,-and most other forms of modulation, is due to the following:

  SSB-ver-AM
(1) Since the carrier is not transmitted, there is a reduction by 50% 
      of the transmitted power (-3dBm).   --In AM @100% modulation: 1/2 of the power is comprised of the carrier; with the remaining (1/2) power in both sidebands. 

(2) Because in SSB, only one sideband is transmitted, there is a further reduction 
      by 50% in transmitted power (-3dBm (+) -3dBm = -6dBm). 

(3) Finally, because only one sideband is received, the receiver's needed 
      bandwidth is reduced by one half--thus effectively reducing the 
      required power by the transmitter another 50% (-3dBm (+) -3dBm (+) -3dBm = -9dBm).  --Remember, if a receiver's bandwidth can be reduced by 50%: the needed transmitter power is also reduced by 50%, i.e., the receiver's Signal to Noise Ratio (SNR) is improved as the receiver bandwidth is reduced. This of course implies that the signal containing the information is not lost--which is the case in this instance.    --Huh? Its true: if I'm Lying, I'm Dying!

Example:
A HAM running 2000 Watts AM, would sound no better than another 
HAM running 250 Watts PEP (Peak Envelop Power) SSB. 

Sometimes, on 20 meters, I would go down to the AM band (14,200 - 14,250 MHz), more often than not, I would hear only the carriers, no modulation--I kid you not! 

I also ran a Collins KWM-2 (180 W pep) in my room with little trouble making contacts around the world. 

I had often felt that there was even more advantages to SSB (than the 9dBm) due to the received signal's intelligibility (filtering, ability to tune off frequency changing pitch, etc.). 

In Fact to take advantage of this, I built an acoustic delay-line[2] which delayed the audio to one ear ~16 ms more than the other, giving me a pseudo stereo like effect. 

I could de-tune the receiver just a little and the effect was like placing different people in the QSO at different spatial positions within the room, with the receiver noise having its own spatial position--it was a Hoot!

It worked best with head phones, but it worked well with speakers also. The effect of all this, on top of improving the intelligibility, was to relieve a lot of the fatigue inherent in attempting to understand conversations in noisy environments over a prolonged time--which can be pretty taxing during poor reception. 
 

[2] The acoustic delay-line used about sixteen feet of 3/16" neoprene tubing, driven at one end by a 16 ohm, electromagnetic ear bud. At the other end was a (piezo) crystal ear bud with a gain stage to make up for the -50 dB inherent loss. The frequency response of the overall system (300 Hz to 2400 Hz) was dictated by the receiver's SSB mechanical filter.