I received a crystal radio kit for my 9th birthday and upon assembling it, it worked. I could hear the two local radio stations, a ham on 75 meters and the nearby Coast Guard station and her boats. This led to Broadcast Band and Shortwave listening with subsequent receivers and set the stage for interest in amateur radio. In the summer prior to my 14th birthday, the local radio club began conducting licensing classes, theory, and Morse from a mil surplus machine that read inked tape. Learning Morse was easy enough and soon I had a Novice license in hand. I didn't have any interest in cw, so I didn't get on as a Novice. A few months later I had the Conditional license in hand and I headed for the 20 meter phone band, with a dipole up about 20-30 ft high, 30-35 watts on AM and was having a ball, working all over North and Central America, including the Caribbean. Many of those qso's were successful but quite a few ended prematurely when qsb/qrm decided to pull the switch on me. One day, I was lamenting that fact to a group of older locals on 6 meters, and one suggested I try cw. I scoffed but he insisted, even offering to loan me a J-38 key. I agreed and accepted his offer, looked up the protocol for cw contacts and began pursuing it. In no time at all, it became second nature, fun, qso's had a much higher success rate, seldom ending prematurely. I realized we were transferring the very same info, albeit, at a slower rate, and that cw had a degree of precision, lacking in phone, namely the confusion of similar sounding words and letters. No phonetics needed on cw. But the best of all was international DX became available. Europeans, Africans, South Americans, Pacific Islands, Asians, all began to populate my log. I was thrilled! I never looked back, nor ever felt that early interest in phone operating again.
An article appearing in QST a few years ago, suggests that cw is 24 or 25 db more effective than AM and 17 db more effective than unprocessed audio SSB, 11 db over processed audio ssb. Consider that 11 db figure. With 6 db average power increase with processed audio ssb, parity is 1250 watts ssb to 100 watts cw. And in the unprocessed audio case, the parity is 5000 watts ssb to 100 watts cw. That's a powerful difference either way, and serves to illustrate why I gravitated to cw and continue using it to this day.
Maybe you have a story to tell?