Demos Are the 80/20 of Finishing Songs

I have a folder on my phone called Ideas. For years it was a graveyard. Hundreds of voice memos, each one a genuinely decent eight seconds, none of them ever becoming a song. The ideas were golden. But nothing ever forced them to grow.

I like to jog, not fast, not far, and never often enough...  but I have taken one thing from it that applies directly to songwriting. The hardest run is never the long one. It is the one where you decide, halfway through lacing your shoes, that today it just isn't happening. A demo is the songwriting version of getting out the door. It is not the marathon. It is just the decision to go round your local park once, at whatever pace, today.

Schrodinger's Demo

Here is the reframe that changed how much I actually finish. A demo is not a small, rough version of the final track. A demo is a device for forcing decisions. An idea in your head can be in three keys at once. It can have two possible choruses. It can be both a ballad and not a ballad. The moment you record it.. actually commit it to one take: it has to pick!.. It has to have a key, a tempo, an order. And most of what we call "finishing a song" is just making those decisions. A demo makes you make them in twenty minutes instead of twenty weeks.

The cheapest demo that still works

You do not need an interface, a session in a digital audio workstation, or a quiet house. You need your phone and one instrument. Here is the whole method: play the song through, start to finish, in one take, singing every part — even the parts where you don't have words yet. Especially those parts. Where the lyric is missing, sing the melody on "la," or on a placeholder line that is obviously wrong. Do not stop to fix anything. The take will be rough. That is not a side effect of the method; it is the method. Just like drawing by hand, you capture the overall shape first, imperfect ovals of the outlines of the shadows of a song. 

The 80 of the 20: The Rough Pass

This is where the 80/20 lives. A one-take phone demo, for maybe twenty per cent of the effort of a "proper" recording, captures around eighty per cent of what the song actually is: its form, how many verses, where the chorus lands.. its key, its tempo, its hook, and, most importantly, where the energy rises and falls. That last one is almost impossible to judge from a lyric sheet and a chord chart. You have to hear the song move. A rough demo is the cheapest way to find out whether the song goes anywhere at all.

You cannot mix a song you haven't finished. You cannot finish a song you cannot hear.

What the demo skips is the other twenty: the production, the clean takes, the bridge lyric you still haven't cracked. And that is completely fine, because you cannot fix the twenty until the eighty exists. The half-songs in my Ideas folder weren't waiting on better production. They were waiting to be heard once, all the way through, so somebody (ME) could decide whether they were worth the rest of the work.

Demo it, then leave it

One more step, and it is the one I would fight for. Once the demo exists, leave it alone for a day. Then listen to it somewhere else. I do my listening in the van. Engine on, moving, the song competing with road noise the way it will one day compete with a kitchen radio and a pair of cheap earbuds. A song that survives the van has real bones. A song that only worked in the silent room where you wrote it was never finished. It was just flattered by the quiet.

When you come back to it, you will know exactly what the song needs, and usually it is something small. If the progression feels soft, the free chord builder at Undercover Zest will let you test a firmer one in seconds. For the lines you sang on "la," RhymeForge will help you find the real words now that you know the shape they have to fit into. The demo told you what to fix. The tools just make the fixing fast.