I have been sent a great many demos. Cassettes through the letterbox at the radio station, CD-Rs handed over at gigs, links in emails, links in messages, links attached to messages that just said "no pressure."
What I remember most is not the quality. It is the sheer number of artists who sent me two songs and then, eighteen months later, sent me the same two songs slightly remixed.
Meanwhile the ones who went on to have careers sent me two songs, then four more, then an EP, then something that sounded like they had finally worked out what they were doing. Not because they were more gifted. Because they finished things.
Talent is common. Completion is not.
This is the least romantic thing I will ever tell an artist, and it is the truest.
Raw musical ability is genuinely widespread. Every town has people who can write a striking melody or turn a phrase in a way that stops you. What is rare — vanishingly rare — is the person who takes that idea all the way through arrangement, recording, mixing, mastering, artwork, metadata and release, and then does it again, and then does it again after that.
A brilliant unfinished song is worth exactly nothing. A merely good finished one is worth something, and more importantly it teaches you things an unfinished one never will.
You do not learn how to make records by starting them. You learn by finishing them, because all the hard lessons live in the last twenty per cent — the part where the interesting decisions are gone and what remains is judgement, tedium and compromise.
Why artists stall at eighty per cent
It is almost never laziness. In my experience it is one of three things, and they need different remedies.
The song stopped matching the idea in your head
Every recording disappoints its writer at some point, usually around the second week. The version in your head was perfect because it was not real. The version in front of you has actual decisions in it.
Artists who stall here believe the disappointment means the song is bad. It does not. It means the song is now real, and this feeling arrives on every record ever made, including the ones you consider flawless.
Finishing means being judged
An unfinished song cannot fail. It exists in a permanent state of promise, and promise is comfortable. The moment you release it, it can be ignored — which is far more likely and far more painful than being disliked.
A great many perfectionists are not perfectionists at all. They are people protecting themselves from indifference, and endless revision is the mechanism.
Nobody ever taught you what finished looks like
This one is practical and it is the easiest to solve. Plenty of artists genuinely do not know the difference between a good demo and a finished master, so they keep polishing without a target.
Pick three commercially released tracks in your genre that you admire. Listen on the same speakers, at the same volume, immediately after your own. The gap will tell you specifically what is missing — and it is usually one of arrangement, performance, or mix, rather than the vague general inadequacy that fear tells you it is.
The ten-song rule
Here is the practice I give artists who are stuck, and it is deliberately unsubtle.
Finish ten songs. Not ten good songs — ten finished ones. Take each all the way to a master you would be prepared to put out, whether or not you actually do.
Numbers one through three will be hard and probably not very good. Four through six will be noticeably faster. By seven you will start making decisions in an afternoon that used to take a fortnight. By ten you will be a different maker of records, and it will have taken less time than most artists spend agonising over a single EP.
Set a deadline per song and treat it like a gig. You would not cancel a show because you were not feeling inspired. Same principle.
Decide in advance what finished means. Written down, before you start. "Mixed, mastered, artwork done, metadata complete." Then finished is a checklist rather than a feeling, and feelings are the problem.
Get one outside opinion, not five. One person whose taste you trust and who will be honest. Five opinions is a way of never having to decide.
Do not start anything new until the current one is done. New ideas arrive precisely when the current song gets difficult. That is not inspiration. That is avoidance wearing a convincing costume.
Register it with your PRO the day it is finished. Build the admin into the definition of done and it never becomes a backlog.
Catalogue is the asset
There is a business argument here too, and it is stronger than most artists realise.
A career is not built on a song. It is built on a body of work. Every finished recording is a permanent asset that can be licensed, playlisted, discovered, re-released, sold and collected on, for as long as you live and then some.
An artist with forty finished songs has forty chances of something landing. An artist with four has four. Over a decade that difference compounds into two completely different careers, and it has nothing to do with which of them writes better melodies.
You cannot get lucky with a song that does not exist. Volume of finished work is the mechanism by which luck becomes available to you.
Start today, with the one you abandoned
Not a new song. You have a new song; new songs are easy and that is the trap.
Go back to the one sitting at eighty per cent that you stopped believing in. Give it two weeks and a definition of finished. Get it to a master. Put it out, or do not put it out — but finish it.
Then do that nine more times. It will teach you more about making records than any course, and at the end of it you will have ten assets you did not have before.