How to Pick Which Unfinished Song to Finish

The honest bit first: the folder is not a folder. It is one decision, avoided thirty times. More writing will not fix it, because a new idea feels better than an old chorus every single time, the way new running shoes feel better than a run. So the folder grows and nothing comes out of it. The way out is not another idea. It is a pick.

Think setlist, not library

You cannot play thirty songs in a forty-minute room. Nobody can. On the road you pick the set that earns the slot, and you get good at picking fast, because the room tells you when you are wrong. Treat the folder the same way. You are not choosing the best song. You are choosing tonight's song. The rest are not deleted; they are resting.

And drop the guilt while you are at it. The folder is not homework you failed to hand in. It is a pantry, and songs keep. I finished one last year that started in a festival dead hour three summers earlier, and the wait had quietly done half the editing: by the time I came back I could hear which verse was cardboard in about four seconds. Some songs are not late. They are not ready, and that is different. More ideas will keep arriving whether you want them or not; supply was never the problem.

You are not choosing the best song. You are choosing tonight's song.

The three questions

Here is what I actually ask, in order. The whole thing takes ten minutes.

One. Which one do I still hum without meaning to? Not which is cleverest. Which one follows me to the kettle. The hum is data: some part of you is still writing that song.

Two. Which one is closest to surviving the van? Play the roughs back somewhere unfriendly... over road noise, on the phone speaker in the kitchen with the extractor fan going. The one that still has bones in bad conditions is closer to done than the pretty one that needs silence and good headphones to exist. (Sol once judged a whole running order off a phone speaker balanced on Doreen's dashboard, and he was right about every track but one.)

Three. Which one scares me a little? There is usually one that feels too honest, or too simple, or too much like you. That charge is worth more than polish. If two songs tie on the first two questions, the scary one wins.

Then leave the others alone

Pick one and give it a clean week. No visits to the folder, no new ideas logged unless they belong to the chosen song (voice memo them and walk away). Finishing is a muscle, and it trains on completion, not volume. I ran three times this week, which for a home month is a personal record, and not one of those runs was long or fast or anything you would put on a poster. Steady beats heroic. Same with songs: one, this week, done badly if necessary. Badly finished is still finished, and it teaches you things polishing never will.

Then do the unglamorous end of it. Rough demo. Kitchen playback. Sent to one person who is waiting. RhymeForge for the rhyme that has stalled the second verse since 2023, the chord builder if the bridge needs the two chords it never got. Small exact jobs. That is all the last twenty percent ever is.

One finished song will teach you more than nine you keep polishing in your head. The folder can have you back next month.

Pick one. Go.