Article

Music while you work: when it helps focus and when lyrics cost you

Music helps focus at work less than most people assume, and the damage depends on the words. A 2022 systematic review of 154 experiments found background music generally hurt memory and language tasks, with lyrics worse than instrumental; instrumental music mostly measured neutral. The question is not whether to play something, but what you are about to do. This article covers what that research actually measured: which work lyrics interfere with, why instrumental music comes out neutral rather than useful, and why a playlist that works in week one has stopped working by week five.

Does background music actually help concentration?

The largest review to date says mostly no, and sometimes worse than no. Cheah and colleagues published a systematic review in Music and Science in 2022 covering 95 articles and 154 experiments across six cognitive domains: memory, language, reasoning, inhibition, attention and processing speed. Background music showed a general detrimental effect on memory and language tasks. Everywhere else the most common result was no effect at all, and exactly one positive effect turned up in the entire set, for instrumental music.

Two of the review's other findings matter more for a workday than that headline does. Music hurt performance on difficult tasks and left easy ones alone, which puts the cost precisely where you would least want it: on the demanding hour rather than on the hour of filing. It also hurt introverts while leaving extraverts unaffected, which is the clearest sign in this literature that the right answer varies by person.

So music is not neutral background. It is an input the task either has room for or does not.

Read the evidence for what it is, though. These were laboratory tasks, run mostly on students, measured in minutes, against a silent comparison condition, and the review is explicit about how much of the area remains understudied. None of it watched somebody write a report in an office. What transfers is the shape of the finding rather than its size: the harder the work, the less room it has for anything else, which is the argument Deep Work already makes about every other interruption.

Why do lyrics cost more than the music does?

Words compete with words. In a 2023 experiment published in the Journal of Cognition, Souza and Barbosa tested students on four tasks under three conditions: silence, instrumental lo-fi, and the same style of music with lyrics. Lyrics pulled down verbal memory, visual memory and reading comprehension by about 0.3 of a standard deviation, a small effect but a consistent one. On arithmetic the cost was smaller and did not hold up statistically.

Instrumental music did nothing in either direction, and that is the part worth a second read, because the same participants believed it did. They correctly noticed the lyrics hurting them. Asked afterwards, they sometimes rated the instrumental track as helpful when the scores showed no benefit at all. Your sense that a playlist is working is not evidence that it is.

The 2022 review reached the same ranking from a far wider base, with lyrics more detrimental than instrumental across its experiments. Two different methods agreeing gives you one firm practical rule: keep words out of work that is made of words. Writing, editing, reading a contract closely, drafting a reply that has to land correctly, all of that is language work competing for the same channel.

The rule has an edge, and it is the one people over-apply.

Arithmetic and mechanical work looked more tolerant in both studies, which is why the spreadsheet does feel different from the memo. What tolerance buys you is permission: play what you like while you reconcile numbers, then drop back to instrumental or to nothing when the work turns verbal. If choosing the next track is becoming the ritual that opens the task, the difficulty is starting, and the Pomodoro Technique answers that more cheaply.

Why does your focus playlist stop working?

Because the novelty was doing part of the work. A 2025 randomized study in Behavioral Sciences ran 428 undergraduates through a two-phase experiment with four conditions: no music, high-arousal music, low-arousal music, and one fixed classical track. What it measured was flow, the absorbed state the study treated as the route to performance. The track that helped at the start had lost 53 percent of that effect after 30 days of repeated listening, with some residual benefit left over.

The same study measured high-arousal music as a drag on performance. Energetic music is the obvious pick when you are trying to start something you would rather not start, and it is the version that measured worst here. Easy to adopt, quietly negative.

That changes what a focus playlist is. A checking schedule or a Task Batching block gets stronger with practice; a playlist runs the other way. It wears out with use, and wearing out is its expected behaviour, so a fading playlist says nothing about the album you picked. Most people answer that by hunting for a new playlist, which restores the novelty for a few weeks and costs a browsing session every time.

The honest limit again: 428 students on a memory task is not a year of office work, and the study's explanation for the decline is a proposal rather than a demonstrated mechanism. Take the direction as usable and leave the number with the experiment. If your playlist has stopped working, nothing has gone wrong with you, and swapping it is a smaller answer than asking what it was covering.

Usually it was covering noise. That is a different problem with better solutions, and the focus pillar treats it as one: an agreed quiet hour, a booked room, a changed default. Each removes the sound instead of layering another one over it.

How does your working style change the answer?

The review's introvert finding is the closest this research gets to a personal answer, and it does not map onto the four working styles this site uses. Architect, Sprinter, Visionary and Improviser describe how somebody organises work, not where their attention sits on a personality scale, so nothing in that study says which style should keep the music on. What follows is this site's reading of how the evidence meets each style's ordinary week, not a research finding.

If you run long protected blocks, closest to the Architect, the studies point one way. Your block holds the difficult work the review found music hurting, and it is also the block where arranging silence costs you least, because you already defend it against everything else. Keep Deep Work quiet and leave the music for the shallow end of the day.

If music is how you get started, nearer the Sprinter, you are using it as a starting signal. A timer does that job with no measured cost attached: the Pomodoro Technique begins on a bell and ends on one, and 25 minutes is short enough that going without a playlist for a round is a small ask. Run one round each way this week and compare.

If you work in a room you do not control, which is often the Improviser's situation, the comparison these studies ran does not describe your desk. They tested music against quiet. Your actual choice is between music and a colleague's phone call, and instrumental sound is a defensible answer to that, precisely because it measured neutral rather than harmful. Neutral wins when the alternative is noise.

One change this week, then: keep lyrics out of anything you write or read closely, and leave the rest alone until you have a fortnight of evidence. Write one line at the end of each day about what you played and what actually got finished, the same two-week test this site puts on any new method, and read the lines rather than your impression of them. If you are not sure which of the styles above is yours, the working style assessment takes a few minutes.

Frequently asked questions

Does music help you focus at work?

Usually not: the largest review of the area found background music generally hurt memory and language tasks, and found a benefit exactly once. The cost landed on difficult tasks and left easy ones alone. So the demanding hour is the wrong one to score. That evidence comes from short laboratory tasks rather than from office work, which makes the direction more usable than the size.

Should you listen to music with lyrics while working?

Not while you are writing, reading or editing: lyrics cost verbal memory, visual memory and reading comprehension about 0.3 of a standard deviation in a 2023 experiment. The wider 2022 review ranks lyrics worse than instrumental too. Words compete with words, which is why language work takes the hit and arithmetic mostly does not. If you want music during a stretch of numbers, that is where it costs least.

Why does my focus playlist stop working after a while?

Because part of what was working was novelty: in a 2025 randomized study, a track's measured effect fell by 53 percent after 30 days of repeated listening. What that study measured was flow, not the finished work. A playlist wears out with use, which makes the fade normal. Before hunting for a replacement, check what the music was covering, because office noise has better answers than another album.

Is silence better than music for deep work?

On the evidence, yes for a protected block: music showed its clearest cost on difficult tasks, and a deep-work block is where your difficult tasks live. The exception is a workspace you do not control. The studies compared music against quiet, not against an open-plan office. When the real alternative is somebody else's call, instrumental sound measured neutral, and neutral is the better end of that trade.