Skip to content
Practice Timer

What to Practice on Piano Every Day: Four Skills, One Session

The four skills a daily piano session buys — technique, harmony, melody, reading. What each means at the keyboard, and how to fit all four into one session.

19 min read

The short answer: four skills, in rotation. Technique sets the ceiling on what you can play. Harmony explains why the music does what it does. Melody and improvisation are the part that’s yours. Reading gives you access to everything anyone else has written. A daily session is a decision about which of those four your thirty minutes is buying.

Ask what to practice on piano every day and nearly every guide answers the same way: warm-up, scales, sight-reading, repertoire, something fun at the end. That’s a list of activities, and activities are not skills. You can play scales for five years and still be unable to move from one chord to the next without a page in front of you. You can read fluently and freeze the moment someone asks you to make something up.

So this article is organized the other way round. First, what the four skills actually are and the precise exercises that build each one. Then how to assemble them into a daily piano practice routine that fits one session — without it taking two hours.


Skill 1 — Technique: your ceiling

Technique is the ability to get the sound you want at the moment you want it. Not speed, not strength. Your technical limit is the real limit on your repertoire: it decides which pieces are available to you at all, and how much of your attention is left over for music once the notes are handled.

Two facts change how you should practice it.

A single note has essentially one variable: how fast the key goes down. The reason is mechanical. The hammer is not pushed into the string — it’s thrown. The action releases it a few millimeters short and it crosses the last gap in free flight, which means nothing your finger does after that instant can reach the string at all. The only thing the string ever knows about your playing is how fast the hammer was travelling when it arrived. Press a key slowly and you cannot produce a loud note, no matter how much force you put into it.

That is almost the whole story, and the exception is worth having. Goebl, Bresin and Galembo miked and instrumented grand actions under two touches — pressing the key from resting contact, versus striking it from above — and found that notes matched for identical hammer velocity stayed audibly distinguishable, because striking adds a finger-on-key noise that pressing doesn’t (Goebl et al., 2005). So touch does leave a mark of its own. It’s a thin layer on top of the variable that decides nearly everything, and hard is still the wrong word for it: what you are varying is speed, and the noise is a by-product of how you arrived at that speed.

Everything else people call “tone” is that speed, plus the timing between notes, plus the balance between notes sounding at once, plus the pedal. As pianist Albert Frantz puts it in his overview of piano technique, the ear directs the fingers and the sound feeds back into the next one. Every technical problem is a control problem.

Which means the strength model is wrong — and that’s a claim about a rationale, not about a book. Hanon’s stated promises are finger strength, wrist and forearm strength, and endurance, and none of the three describes what a control problem needs. Anything that moves your fingers on keys builds independence; power comes from the whole body rather than the forearm, and chasing forearm strength produces tension and an aggressive tone; and playing on without releasing is how injuries start.

The patterns themselves are neutral, and there’s a real argument for them. They are trivial to memorize, and a pattern you don’t have to read or think about frees your entire attention for the thing that does need it. As a vehicle for evenness, for dynamic control, for rhythm variants, for transposing into the key of the day, Hanon is a perfectly good tool — provided you can say what this particular repetition is for before you play it. What doesn’t survive is the version on the cover: reps for strength, at speed, unlistened to. The difference isn’t the notes on the page. It’s whether there’s a named target attached to them.

The categories, and the drill for each

Category What lives here The drill
Single notes Scales, broken chords, repeated notes, trills, tremolos Scale of the day, hands together, two octaves, even tone. Repeated notes with finger substitution 3-2-1, not one finger hammering
Double notes Thirds, sixths and octaves in one hand Thirds up a scale, hands separately, listening for both notes landing together. Octaves from a loose wrist, staccato before legato
Position changes Adjacent (thumb passing) and remote (leaps) Scales for the thumb. For leaps: play the target chord, look away, jump to it silently ten times, then with sound
Touch and articulation How the key goes down, and what happens between notes One phrase, five ways: legato, staccato, portato, melody louder than accompaniment, accompaniment louder. Hear the difference
Pedal Syncopated (legato) pedal, half pedal, release timing Play a chord progression and change pedal after each new chord sounds, never with it. Record it and listen for blur

Your daily technical minimum

One key per day, moving through the circle of fifths: scale, arpeggio, and the cadence (I–IV–I–V–I) in that key, hands together. Twelve days covers every key, and you’re never in an unfamiliar one for long. Add whichever category above your current piece is actually demanding — that’s the warm-up that earns its place.

The diagnostic: can you hear exactly how a passage should sound, and fail to produce it? That’s a technique problem. If you can’t hear it in the first place, that’s the next skill.


Skill 2 — Harmony: the logic underneath

Harmony is why one chord follows another, where the tension builds and where it lets go. It’s the difference between memorizing a sequence of hand positions and understanding a piece — and it’s what lets you play without the page, accompany a singer, or work out a song by ear.

The realization that unlocks most of it: chords don’t move as blocks. Individual voices move, and they move as little as possible. The inevitability you hear in a good progression lives in those lines, not in the chord names.

Take the most common progression in Western music, ii–V–I. In C major that’s Dm7 → G7 → Cmaj7. The two notes doing all the work are the 3rd and 7th of each chord — the guide tones. In Dm7 they’re F and C. Move to G7: the F simply stays put and becomes G7’s 7th, while the C falls a half step to B, G7’s 3rd. Move to Cmaj7 and the same thing happens in mirror: the B stays put and becomes Cmaj7’s 7th, while the F falls a half step to E, Cmaj7’s 3rd.

ii–V–I in C, as two voices
Guide tones · 3rds and 7ths
The guide tones of a ii–V–I progression in C major. At each chord change one voice holds its note while the other falls a half step.
VoiceDm7iiG7VCmaj7I
Upper voiceC7th−½ stepB3rdholdsB7th
Lower voiceF3rdholdsF7th−½ stepE3rd
Read the markers diagonally: at every change one voice holds and the other falls a half step, and the two voices swap which of them is doing it. That alternation is the progression — the chord names are just what those two lines happen to spell. Play it as a shell voicing (root in the left hand, these two notes in the right) and it transposes to all twelve keys without relearning anything.

Two voices, and at each change one holds while the other moves a half step. That’s the entire progression, and once you hear it that way you can play it in any key.

What to practice, precisely

  1. Roman numerals instead of chord names. I ii iii IV V vi vii°. The point isn’t theory for its own sake — it’s that a progression written in numerals transposes to any key with no relearning. Every working musician thinks this way.
  2. The three functions — and the seven names underneath them. Tonic (I, vi, iii) is home. Predominant (ii, IV) is leaving home. Dominant (V, vii°) is tension that wants to resolve. Almost every progression you’ll meet is a route through tonic → predominant → dominant → tonic. Label eight bars of your current piece this way and the shape appears. But the grouping is a working shortcut over seven degrees that each have their own name and their own character, and it’s worth knowing which is which — the table below is the version to keep.
  3. Shell voicings around the circle of fifths. Root in the left hand, 3rd and 7th in the right. Play ii–V–I in all twelve keys, moving the guide tones by step, five minutes a day. This is the single highest-yield harmony exercise there is.
  4. The four cadences. Authentic (V–I), plagal (IV–I), half (ending on V), deceptive (V–vi). Play all four in the key of the day. You are training the sound of an ending, not the vocabulary word.
  5. Analyze what you already play. Take eight bars of your current piece, write the numerals above the staves, mark the cadence. It costs three minutes and almost nobody does it.
  6. Follow the leading tone. The 7th degree of the scale — B in C major. Find it in your piece and watch where it goes. It rises to the tonic nearly every time, and the times it doesn’t are the composer being deliberate. Tension and resolution are, mechanically, this.
  7. Spot secondary dominants. When a chord that “should” be minor turns up major, it’s usually V of the chord that follows it. Learning to recognize V/V and V/vi explains most of the accidentals in the music you’re playing.

The seven degrees, their proper names, and what each one is actually for:

Degree Name Function What it does
I Tonic Tonic Home. The chord everything else is measured against
ii Supertonic Predominant The smoothest way out of home — and the first half of every ii–V–I
iii Mediant Tonic The rarest and most ambiguous of the seven; shares two notes with both I and V
IV Subdominant Predominant Leaving home the plain way. The plagal “amen” chord
V Dominant Dominant Maximum tension. Contains the leading tone, which is why it pulls
vi Submediant Tonic Still home, but shaded minor — and the chord a deceptive cadence lands on
vii° Leading-tone Dominant A V7 with its root removed. Same pull, less weight

Two things fall out of that table once you have it. Mediant and submediant are named for sitting a third above and a third below the tonic — not for anything they do, which is why their names tell you less than the others’. And the reason vi can substitute for I is arithmetic: they share two of their three notes. Function is mostly a question of which notes two chords have in common.

Level Fifteen minutes of harmony looks like
First year The I, IV and V chords in three keys, played until they’re under the hand without thinking. Harmonize a nursery tune using only those three
Intermediate ii–V–I with shell voicings through four keys a week. Roman-numeral analysis of the piece you’re learning. All four cadences in the key of the day
Advanced Rootless voicings and inversions; reharmonize a melody you know two different ways; transpose a full progression by ear on sight

Skill 3 — Melody and improvisation: the part that’s yours

Technique and harmony are inputs. This is output — and it’s the skill self-taught players postpone forever, because it feels like an advanced topic. It isn’t. You can improvise on your first day with three notes. What makes improvisation sound like wandering isn’t lack of ability; it’s lack of constraint.

What to practice, precisely

  • Constraint drills. Loop a two-chord progression in the left hand. Improvise over it using only chord tones. Then only the 3rds and 7ths. Then a single octave. Then three notes total. Every constraint you add makes the result more musical, not less — the limit is what forces you to shape a line instead of running a scale.
  • Target notes. Before each bar, decide which note you’ll land on at beat 1, and land on it. The 3rd and 7th are the best targets: they’re the notes that tell the listener what the chord is.
  • Phrase, then stop. Three to five notes, then silence long enough to hear the next idea arrive. Beginners play continuously because the silence feels like failure. Silence is most of what makes a solo sound like speech rather than typing.
  • One motif, developed. Take four notes. Repeat them. Move them up a step. Turn them upside down. Stretch them to double length. Break off the last two and repeat those. Getting eight bars out of one four-note idea teaches more than eight bars of fresh material.
  • Play a tune you know in three keys. By ear, no page. This is where technique, harmony and your ear meet, and it’s the fastest way to find out which of the three is actually weak.

If you play strictly classical, none of this is off-limits — the improvising tradition it belongs to is the one Bach, Mozart and Beethoven worked in. Fifteen minutes a week is enough to stop the freeze.


Skill 4 — Reading: access to everyone else’s music

Reading is what makes the rest of the repertoire available to you, and it’s the skill most self-taught players quietly avoid because progress is invisible.

Two research findings are worth knowing before you practice it.

Sight-reading is partly an ear skill. When Kopiez and Lee tested 23 candidate predictors of sight-reading ability, the best combination was trilling speed, sight-reading experience acquired before age 15, speed of information processing, and inner hearing — explaining just under 60% of the variance between players (Kopiez & Lee, 2008). Inner hearing is the one on that list you can straightforwardly train as an adult. Reading well is not only moving your eyes faster; it’s hearing the page before your hands get there.

And it isn’t purely a matter of hours. Working memory capacity predicts sight-reading skill on top of accumulated deliberate practice, and more practice doesn’t erase the difference (Meinz & Hambrick, 2010). Two useful conclusions follow: if you read badly it isn’t laziness, and the techniques that help most are the ones that reduce how much you’re holding in your head at once.

What to practice, precisely

  • Read below your level, every day. Two grades below what you can play. Five minutes. Never stop, never go back — a wrong note played in time is a success here, and a correct note played late is not.
  • Read intervals, not note names. “Up a third, down a fourth.” Naming every note individually is the single biggest thing making reading slow.
  • Clap the rhythm first. Most sight-reading collapses are rhythmic, not pitch-related. Take the rhythm alone before you touch the keys.
  • Stay a bar ahead. Cover what you’ve already played with an index card. Your eyes belong where your hands are going.
  • Chunk with your harmony. Once you know your chords, you stop reading four separate notes and start reading “F major, second inversion”. That’s a direct cut to working-memory load — and it’s the reason skill 2 makes you a better reader.
  • Hear it before you play it. Look at a line, sing it or hear it internally, then play. You’re training the strongest trainable predictor on the list.

The one part of that list a tool genuinely does better than you can is supplying the material — an endless stream of notes you have not seen before, graded, and answered on the keys rather than by naming them out loud. That’s the whole job of a reading trainer.


Fitting four skills into one session

You cannot do all four properly every day, and you shouldn’t try. The session has a fixed shape — named blocks, timed, decided before you sit down — and the four skills rotate through it.

Block Share of playing time What goes in it
Warm 15% Technique — the key of the day, plus whatever shape today’s hard passage is made of
Repair 45% The specific bars that broke yesterday
Build 25% One rotating slot: harmony, improvisation or reading
Close 15% Play through, then write down what broke

Two answers give you today’s version:

Build today’s session
Two answers · no sign-up
1 · How long do you actually have today?
2 · Where are you right now?
Pick a length and a level and today's session appears here — four blocks, with the minutes worked out and what goes in each one at your level.
Warm3 min5 min8 min
Open the hands on the shapes you are about to need.A five-finger pattern and one scale, hands separately — in the key of the piece you are about to work on. Not C major out of habit.The technical shape of today’s hard passage — its arpeggio, its scale, its leap — in its actual key. Two minutes of that beats ten of Hanon.Passage-derived technique, plus the one system you are keeping alive this month: a scale cycle, an étude, an octave routine.
Repair8 min16 min24 min
The bars that broke yesterday. The largest block, always.Two to four bars. Hands separately at a tempo you cannot miss at, then together. Stop the instant it is wrong — do not play on and hope.Four to eight bars. Hands separately, then together, metronome starting where failure is impossible. Then add a bar on each side and drill the joins.Two separate problems, alternated — not one ground down. Name a restart cue for each: a bar you could begin from cold, on stage, from silence.
Build4 min9 min14 min
Reading, harmony or improvisation — rotate across the week.Reading: one short piece you have never seen, straight through, no stopping. Or harmony: I, IV and V until they are under the hand.Reading two grades below level, ii–V–I with shell voicings through three keys, or one constraint drill over a two-chord loop.Rootless voicings and reharmonization, transposing by ear, or the un-memorized section. Mental practice away from the keys counts.
Close3 min6 min8 min
Play through, then write tomorrow’s repair list.Play something you already own, for pleasure. Then write down the bar numbers that broke today.One unbroken run of the whole piece. Mark what fails; fix nothing. That mark list is tomorrow’s repair block.Perform it — record, or play as if someone is listening. Log where the recording differs from what you thought you played.
18 minutes of playing, 20 on the clock. The difference is 3 gaps of ~40s of real rest between blocks — hands off the keys, phone face down. Those gaps are not lost time; they are where the session consolidates.36 minutes of playing, 40 on the clock. The difference is 3 gaps of ~80s of real rest between blocks — hands off the keys, phone face down. Those gaps are not lost time; they are where the session consolidates.54 minutes of playing, 60 on the clock. The difference is 3 gaps of ~2 min of real rest between blocks — hands off the keys, phone face down. Those gaps are not lost time; they are where the session consolidates.
Twenty minutes is not a shortened hour — it is the same session at a smaller scale. Repair stays at roughly 44% of playing time whichever row you pick, because the passage that broke yesterday is the only block that reliably changes what you can do tomorrow.

Repair gets the largest share because it’s the block that quietly gets eaten when a session is arranged badly, and because the evidence on how experts actually practice points straight at it. When researchers filmed pianists learning the same unfamiliar passage, what separated the strongest learners wasn’t total time or number of repetitions — it was that they stopped the instant something went wrong, isolated the exact spot, slowed it down, and didn’t move on until it was correct (Duke et al., 2009).

The repair loop: find the smallest broken unit, usually two to eight bars. Hands separately at a tempo where failure is impossible. Hands together at that same tempo. Then add a bar before and a bar after — the hard spot is often only hard because of how you arrive at it. Then come back to it a second time later in the block with something else in between, because alternating two problems retains better than grinding one — the contextual interference effect (Shea & Morgan, 1979), covered in interleaved practice for musicians.

One myth to drop while you’re there: the claim that every wrong repetition needs five, or seven, or ten correct ones to undo. No study supports that number. What the evidence supports is narrower — mistakes left uncorrected get encoded, and skilled learners fix them on the spot instead of playing on and hoping.

The weekly rotation

Day Build slot
Mon Reading — five minutes below level, then a new line
Tue Harmony — ii–V–I through three keys, or analyze eight bars
Wed Reading
Thu Improvisation — one constraint drill over a two-chord loop
Fri Harmony
Sat Free — learn something by ear, transpose a tune, play for pleasure
Sun Off, or ten light minutes

Consolidation happens between sessions rather than during them, so the day off is part of the method rather than a break from it (why rest improves practice). On total time, the research is covered separately in how long you should practice — the short version is that twenty structured minutes beats an unstructured hour.

Close the loop before you stand up

Play one unbroken run of the piece. Mark what fails; fix nothing. Then write the bar numbers down — bars 22–25, left hand rushing; bar 40, fingering unresolved. That list is tomorrow’s repair block, which is why you never sit down wondering what to practice. It’s also the single field most practice plans leave out, and the cheapest one to add.

For anyone memorizing, there’s a stronger version. When Chaffin and Imreh recorded a concert pianist learning the Presto from Bach’s Italian Concerto from scratch, what she built alongside the notes was a set of performance cues — landmarks in the score she deliberately attended to and could restart from. Those cues predicted where she hesitated when first playing from memory, and still predicted her recall two years later (Chaffin & Imreh, 2002). Name one bar per section you could begin from cold, from silence, in front of people.


Which skill is your bottleneck?

Most practice problems announce themselves. The trick is knowing which of the four they point at.

What you notice The bottleneck What to add tomorrow
You hear exactly how it should sound and can’t produce it Technique The specific category — double notes, leaps, pedal — that the passage demands
Your hands tire or ache, and tone gets harsh when you push Technique (efficiency) Stop. Slower keyspeed, released wrist, real pauses. Pain is information
You can’t play anything without the sheet music in front of you Harmony Roman numerals and shell voicings. Learn one song by its progression, not its notes
Someone asks you to make something up and you freeze Melody Constraint drills over a two-chord loop, three notes only
A new piece takes months of note-hunting Reading Five minutes below level, daily, never stopping
You’ve played for years and every new key feels foreign Harmony + technique The key of the day through the circle of fifths
Clean alone, falls apart inside the piece Repair method Start the repair from the bar before, end one bar after
Fine at 60 bpm, collapses at 100 Fingering, not speed Hands separately at 60, audit every finger number
Fine at home, gone in front of anyone Memory structure Name three restart landmarks per section, practice beginning from each
You finish with no idea what improved No close block Write the bar numbers before you leave the bench

Frequently asked questions

What should I practice if I only have 20 minutes?

Three minutes of technique in the key of the day, eight repairing yesterday’s broken bars, four on whichever skill the rotation says, three closing with a run-through and a written list. Twenty minutes is not a shortened hour — it’s the same session at a smaller scale.

Should I practice scales every day?

Yes, but one key at a time rather than all twelve. Scale, arpeggio and cadence in the key of the day, moving through the circle of fifths, takes about three minutes and covers every key in under two weeks.

Are Hanon and Czerny worth it?

Both are good vehicles and poor goals. The patterns are neutral material, and being memorable is exactly their advantage: with nothing to read, your whole attention is free for evenness, dynamic control, rhythm variants or transposition. Use them that way, with a named target for each repetition, and they earn their place. What doesn’t hold up is the rationale on the cover — finger strength isn’t what the piano rewards, and playing on without releasing is how tension builds. If you enjoy Czerny as music, play it as music.

Do I need to learn theory to play the piano?

You need harmony, which is a smaller and much more practical subset. Roman numerals, the three functions, the four cadences and guide-tone motion will change how you learn pieces within weeks. You can skip almost everything else until you want it.

Can I improvise if I only play classical music?

Yes, and the tradition is on your side — Bach, Mozart and Beethoven all improvised publicly. Start with chord tones over a two-chord loop from a piece you already know.

How do I actually get better at sight-reading?

Read below your level every day without stopping, read intervals rather than note names, and practice hearing a line before you play it. That last one matters more than most players expect (Kopiez & Lee, 2008).

How do I know whether a session worked?

Minutes won’t tell you. Did the repair block get its share, or did comfortable run-throughs eat it? Were the gaps real rest, or scrolling? Can you name, in bar numbers, what’s better than yesterday? That reasoning is the method we built the app around — the shape of the session, not its length.


The bottom line

Technique buys you control, and it’s a problem of efficiency rather than strength. Harmony turns a sequence of hand positions into music you understand. Improvisation is available to you today, at three notes. Reading is partly an ear skill, and chunking is what makes it fast.

Practice technique daily, repair yesterday’s failures with nearly half your time, rotate the other three skills across the week, and finish by writing down what broke. The list of activities you were handed isn’t wrong — it just never said which of these four it was buying.


References

  • Goebl, W., Bresin, R., & Galembo, A. (2005). Touch and temporal behavior of grand piano actions. Journal of the Acoustical Society of America, 118(2), 1154–1165. doi:10.1121/1.1944648
  • Duke, R. A., Simmons, A. L., & Cash, C. D. (2009). It’s not how much; it’s how: Characteristics of practice behavior and retention of performance skills. Journal of Research in Music Education, 56(4), 310–321. doi:10.1177/0022429408328851
  • Chaffin, R., & Imreh, G. (2002). Practicing perfection: Piano performance as expert memory. Psychological Science, 13(4), 342–349. doi:10.1111/j.0956-7976.2002.00462.x
  • Kopiez, R., & Lee, J. I. (2008). Towards a general model of skills involved in sight reading music. Music Education Research, 10(1), 41–62. doi:10.1080/14613800701871363
  • Meinz, E. J., & Hambrick, D. Z. (2010). Deliberate practice is necessary but not sufficient to explain individual differences in piano sight-reading skill: The role of working memory capacity. Psychological Science, 21(7), 914–919. doi:10.1177/0956797610373933
  • Shea, J. B., & Morgan, R. L. (1979). Contextual interference effects on the acquisition, retention, and transfer of a motor skill. Journal of Experimental Psychology: Human Learning and Memory, 5(2), 179–187. doi:10.1037/0278-7393.5.2.179

Further reading

23 min read

Best Apps to Learn Piano: 6 Tools, One Per Problem

The best apps to learn piano are not courses: a metronome, a timer, a chord reference, a score scanner, a reading trainer and a backing band. All six tested.

  • practice-tips
  • product
  • routines