How Do I Know If A Song Is ‘Done’?

You’ve been working on the same song for three weeks. The chorus has been rewritten four times. The first verse keeps changing depending on what mood you’re in. You don’t know if it’s good, you don’t know if it’s done, and you definitely don’t know whether to send it to a producer or shelve it for another month.

If that sounds familiar, this one’s for you. I want to walk through how I (and most of the artists I see come through CD Unity) actually decide when a song is finished, what the warning signs are that it isn’t, and how to stop yourself from spending hours perfecting something that was already there two drafts ago.

I’m not going to pretend there’s a magic checkbox. There isn’t. But after talking to bands every week and listening to hundreds of songs land on my desk in various states of “almost there”, I can tell you there are real, honest signals that separate a finished song from one that’s just being fiddled with. Let’s get into it.

The first thing worth saying is that “done” and “perfect” are two completely different ideas, and most songwriters confuse them. A song is done when your changes stop making it better. That’s it. That’s the whole definition. If you tweak the second chorus and it’s not clearly stronger, just different, you’ve crossed the line from finishing into fiddling.

I’ve watched bands sit on a great song for six months because they kept making small changes that were lateral, not upward. The song was done after the third draft. Drafts four through fifteen were just nervous energy.

So before you ask “is this song finished?”, ask yourself a harder question: “are my last three changes actually improvements, or am I just moving things around because I’m afraid to call it?” Be honest. You’ll know.

This is the test I’d recommend to every songwriter who comes to us asking whether a track is ready. Sit down with just your voice and one instrument, a guitar or a piano, and play the song from start to finish. No effects. No backing vocals. No production. Just you and the bones of it.

If the melody still feels strong, if the chorus still hits, if the lyric still says what you wanted it to say, the song is done. Or near enough done that any remaining work is arrangement, not songwriting.

If the song doesn’t survive that test, no amount of mastering or fancy production will save it. I see this every week from artists who send us their masters expecting magic, when the truth is the song itself wasn’t finished. The arrangement was hiding it.

What To Listen For

When you do the acoustic test, you’re listening for three things: does the melody have a shape that builds and releases, does the lyric tell a story or land an emotion, and does the chorus feel like a payoff after the verse? If those three are present, the song’s bones are right. Everything else is decoration.

Walk away for 48 hours. I mean it. Don’t listen to the song, don’t open the session, don’t even hum it. Then come back and play it once, all the way through, without stopping or skipping. What you feel in those three to four minutes is the truth.

If it still moves you, if you find yourself singing along to a section of your song that you’d half-forgotten, if the chorus lands the way you hoped it would, you’ve got a great song on your hands. If it sounds a bit flat, or if you suddenly hear a clunky line in the first verse you didn’t notice before, that’s your gut telling you what your tired brain couldn’t.

The reason this works is simple. When you’ve been hearing a track on loop, your ears stop noticing things. You can’t tell the difference between what’s actually working and what you’ve just got used to. Distance gives you back the perspective of a listener who’s hearing it for the first time, which is exactly what your audience would be doing.

Bonus: Get Someone Else’s Ears On It

Better yet, after your 48-hour break, play it for someone who isn’t a musician. A friend, your partner, anyone. Don’t tell them what to listen for. Just press play and watch their face. If they zone out during the second verse, you’ve got a problem there. If they tap their foot during the chorus, that section is doing its job.

I trust a non-musician’s reaction more than a producer’s notes about 80% of the time. Producers can fool themselves into liking clever production. A regular listener can’t. They either feel something or they don’t.

Every finished song has a single big idea at its core. It might be a melodic hook that won’t leave your head, it might be a heartfelt lyrical moment that hits you in the chest, or, if you’re lucky, it might be both. But you should be able to say in one sentence what your song is about and why anyone should care.

If you can’t, the song isn’t finished, no matter how many takes you’ve recorded. You haven’t worked out what it is yet.

The catchy songs that stick are usually built around one undeniable melodic moment, a hook in the chorus that the listener can sing back after one play. The heartfelt ones are built around one lyrical truth that makes someone go “yeah, that’s exactly how it feels”. The best ones, the ones that record labels actually sign, do both. They give you a melody you can’t shake and a line you can’t forget.

The “Sing It Back” Test

Here’s a quick check. Sing your chorus to yourself right now, away from the recording. Can you do it? Is the melody distinct enough to hold its shape without the instrumental? If yes, you’ve got something genuinely catchy. If you can’t sing it without the backing track propping it up, the melody isn’t doing enough work yet.

Same goes for the lyrical side. Read the chorus out loud as if it were a sentence in a conversation. Does it sound like something a real human would say with feeling? Or does it sound a bit like a placeholder you forgot to replace?

I want to talk about the thing that keeps most artists from finishing: the fear that the song isn’t quite ready yet. You’ve spent hours perfecting the snare reverb. You’ve gone back and forth on whether the bridge needs an extra bar. You’ve muted and unmuted the same backing vocal eleven times.

None of that is finishing. That’s avoiding finishing. And I get it, because finishing means committing, and committing means people will hear it and might not love it. So your brain keeps inventing small changes to make so you don’t have to face that.

The cure is a deadline. A real one, not a “soon” one. Pick a release date. Book a mastering slot. Tell your bandmates. Once the calendar is set, the song will tell you it’s done because there’s no more time to pretend it isn’t.

I’ve had a few artists ask me recently if generative artificial intelligence can help them work out whether a song is finished. Honestly? No. Not in any meaningful way. AI can tell you whether your audio levels are within spec, or whether your song fits the average tempo for your genre, but it can’t tell you if your chorus makes a listener feel something. That’s still a human job, and it always will be.

Use AI for the boring stuff if you want. Loudness checks, metadata, transcription. But the question of whether a song is truly done is one only you, your gut, and a small handful of trusted listeners can answer. No model has ever cried at a song. People do.

Once you’ve decided your song is genuinely finished, the next stages are mixing, mastering, and release prep. The mix balances everything, the master gets the song to commercial loudness and prepares it for streaming and physical formats, and then you’re into artwork, release strategy, and physical formats if you’re going that way.

If you’re heading towards a physical release, this is where we usually come in at CD Unity. The bands we work with at our pressing plant have already gone through everything I’ve talked about above by the time they send us their masters, and you can hear the difference. A finished song masters better, sounds better on a pressed disc, and makes a much stronger first impression on whoever picks the CD up.

If you’re not sure what comes next after the song is done, our step-by-step checklist for release day walks you through it from final master to launch day. And if you’re still working on the artwork side, the Ultimate CD Artwork Guide is free and covers a lot of ground.

Josh McKenzie

Hello, I’m Josh, and I’ve been honing my graphic design skills for almost 15 years now, catering to the needs of bands and businesses alike. What really fascinates me is the business aspect of the music industry. In addition to my design work, I also happen to play the Hammond organ, and I strive to share my knowledge through helpful articles that I write exclusively for you all!