I’m always anxious about my singing voice, which is one reason why I always am ready to call anything I record a “demo” and cover myself from criticism (that I can’t sing as well as I should be able to, given how many years I’ve been at this).

My next step will be to record some lead vocals, which I will do as another track in Soundtrap. I may add some more live keyboards (myself, playing, as opposed to loops) at the end, to give the second half of the song more texture. In regards to vocals, I aim to do my best, but in my mind, I keep wondering: Who else could I ask to sing this? Or help me sing this? A backing track might make all the difference in the world.

L’unico grande cantautore italiano che se la tiri più di De Gregori

One of the advantages of being a songwriter instead of a performer is that, while artists have a hard time recovering their reputation when a record bombs, the only time the public pays attention to who wrote a song is when it's a hit. Most people don't realise that, even for successful writers, the good-v-great ratio is low: Chambers has written more than 1,000 songs in the last 15 years, of which 21 ranked in the top 10 – that's one hit for every 47 songs. That may sound like a frustrating process, but most writers would agree it's necessary to write non-hits to get to the nuggets. As with athletes, it's important to exercise the writing muscle.

If anyone was hoping to stumble on a secret formula, that hope was quickly shattered. As Motown legend Lamont Dozier once said: "I've written about 78 top 10 songs, and I still don't know what a hit is. I can only go by what I feel.