Does focus music actually work?

Strong

Background music does not make you smarter

The largest analysis of this question pooled 97 studies. The overall effect of background music on cognition was r = 0.03, which is effectively nothing. Reading comprehension came out at r = negative 0.11, and all eight reading studies pointed the same way.

One outcome was reliably positive: emotion, r = 0.11. Music makes work feel better. That is the honest ceiling for this entire product category, ours included, and it is a perfectly good reason to press play.

If a focus app tells you its music raises your IQ, sharpens your brain, or unlocks a flow state, it is selling you something the literature does not support.

Strong

Lyrics genuinely hurt, and instrumental music genuinely does not

This is the best established finding in the field. Music with lyrics measurably impairs verbal memory, visual memory, and reading comprehension. Instrumental music behaves very differently. In a direct comparison, instrumental music was four times more likely to have no effect than to disrupt verbal recall.

The honest way to say this: instrumental music will not make you better at writing or code, but it will not get in the way either. Not getting in the way is the real product.

One exception worth knowing. The same study found instrumental music clearly disrupts tonal recall. If your work is pitch based, transcription, ear training, learning an instrument, music is the wrong choice and no app can fix that.

Strong

Slow tempo calms your nervous system

This is the most robust lever any music app has. In a controlled study across five tempos, slow music near 60 BPM raised parasympathetic (vagal) activity and lowered heart rate.

The half everyone gets wrong: in that same study, sympathetic activity did not differ significantly between tempos. Fast music raises how aroused you feel, but the tidy story about it flipping your nervous system into high gear is not what the data showed. Response also varied by person. There is no universal tempo dose.

Moderate

Covering distractions is real, but music is bad at it

Intelligible speech is the worst common office distraction, and masking it works. Here is the part focus apps do not mention. When researchers compared five different masking sounds at matched intelligibility, a water like natural sound completely removed the disruption, matching a silent control. Filtered pink noise came second and people liked it.

Their conclusion was that music cannot be recommended as a general purpose masker. Vocal music was worst. So if your main problem is a noisy room or a talkative office, a good noise bed beats any music, including ours.

Mixed

Binaural beats do something, just probably not what you want

Two meta analyses put the average effect on cognition around g = 0.40 to 0.45, which is real. But the detail matters more than the headline.

Frequency band changes the direction of the effect. In one controlled study, beta beats improved recall while theta beats made it worse than a plain noise control. And the most relevant recent work, an 80 person study with EEG confirmation, found that binaural beats entrained the brain but did not stop attention declining over time, which is exactly what a long work session needs.

Binaural beats also require headphones. On speakers the two tones mix in the air and the effect collapses. That part is physics, not opinion.

Speculative

40 Hz gamma is not a concentration booster

The serious 40 Hz research is about Alzheimer's pathology, amyloid clearance and slowed brain atrophy. It is not about making healthy adults think faster. A 2025 study found a single session of 40 Hz stimulation entrained EEG without improving perception, attention, or memory.

Some recent work does report short term vigilance gains from 40 Hz flickering light, but that is a different sense, and the strongest result is still a preprint. None of it transfers to a 40 Hz tone inside a music mix.

Strong

Music that never repeats is not automatically better

Drift generates music live and never plays the same session twice. We used to present that as a focus benefit. The research does not support that, so we stopped.

Two findings pushed us. First, the disruptive effect of ignored background sound fades after about eight repetitions. Repetition is how your brain learns to stop noticing something, which is precisely what background music should want. Second, sound that keeps changing is more disruptive than steady sound, and that particular kind of distraction cannot be overcome by concentrating harder.

So the goal is not music that never repeats. It is music that never surprises you.

That is a real design constraint and it is what we build toward. Variation belongs in the slow stuff, texture and density and register drifting over minutes. It does not belong in sudden events. No instrument bursting in unannounced, no abrupt change of key or kit. A surprise is the one thing reliably shown to steal your attention back.

We still think generating live is the right approach, because it lets the music move with a long session instead of looping a four minute track forty times. But we are not going to tell you novelty makes you focus better.

Where the evidence changes the product

Drift is fully instrumental, which is the one strong finding we can actually build on. Most moods sit in the slow tempo range where the parasympathetic evidence is best. The moods meant for study and deep work use the frequency bands the memory research supports, and the dreamier moods that use theta are never sold as study tools.

Entrainment fades in over three minutes, stays quiet enough to sit under the music, switches to a speaker safe mode when you unplug your headphones, and can be turned off completely. Individual response varies enough that an off switch matters more than fine tuning.

When we found errors in our own research notes, including a citation that did not say what we claimed it said, we fixed them and wrote down what changed. We would rather be correctable than impressive.