An intermediate guitar practice routine that changes every week.

An intermediate practice routine works when the session is split into five blocks with different jobs: a warm-up with a stated intent, one repair target, tempo work on that target, repertoire played under pressure, and a cold-start test. A 50-minute session runs 5 / 15 / 10 / 15 / 5 minutes. A 25-minute session runs 3 / 10 / 8 / 4, with the tempo ladder folded into the repair block. Keep the five blocks fixed and change the material inside them as the evidence changes. A routine that is identical in week four to week one has stopped producing change, because the fault it was built to fix is either fixed by now or was never the real one.

Why the intermediate stall is a structure problem

You can almost play a hundred things and fully play none. That is rarely a talent ceiling and it is usually not a discipline problem either. It is what happens when every session is built from the same material: the songs you can already mostly play, at the speed you can already mostly play them, stopping and restarting whenever a bar goes wrong.

That session feels like work and produces very little, because nothing in it is aimed at a specific fault and nothing in it measures whether the fault moved. The five blocks below exist to give each part of the hour one job and one output. If the sessions have stopped producing any change at all, read why guitar practice plateaus first, because a plateau needs a diagnosis before it needs a routine.

The five blocks, and what each one is for

Run them in this order. Each block hands its output to the next, and the last block generates the input for tomorrow.

1. Warm-up with intent, 5 minutes

This block gets your hands and ears working and commits you to the one thing the session will fix.

Use material that matches the motion your repair target needs. If the target is a barre change, warm up moving between barre shapes slowly and cleanly. If it is string crossing at speed, cross strings at a tempo where nothing misfires. A warm-up made of unrelated scale patterns is just the first five minutes of playing.

End the block by naming the target out loud or writing it down. If you cannot finish the sentence “today I am fixing ___”, you do not have a session yet. You have practice time.

2. The one repair target, 15 minutes

This block makes one specific thing measurably better. It is the only block that is allowed to feel bad.

Take the smallest unit that actually breaks. Usually that is two bars, one chord change, one position shift, or one string crossing. Not a section, and never a whole song. Work it well under performance speed with your phone recording the entire block, so you are judging the take instead of the memory of the take.

The test is countable: five correct repetitions in a row, counted, not “it is mostly there now”. If you cannot get five, the unit is too big or the tempo is too high. Halve the unit before you lower the tempo again.

Choose the target from your own evidence, not from a video you saw: the place the pressure block collapsed last session, or the break point the cold-start test exposed. The practice plateau diagnostic is a faster route to it when the recording is ambiguous. For chord changes specifically, Strumly's guide to smoother chord transitions breaks the motion into the parts worth isolating.

3. Tempo work, 10 minutes

This block finds the speed at which the repaired unit falls apart, which is the real edge of the skill.

Build a ladder: a starting tempo where the unit is genuinely clean, the tempo the music actually runs at, and steps between them. Climb one step at a time. When the fault reappears, drop two steps, play it clean three times, and climb again. The tempo ladder builder lays out the steps if you would rather not do the arithmetic mid-session.

Two rules keep this block honest. The top of the ladder is the song's tempo, not a personal record, because playing faster than the music requires is a separate skill and it is not the one that finishes your song. And clean at a slow tempo counts for nothing if you never climb. The point is to move the break point, not to enjoy the bottom rung.

4. Repertoire under pressure, 15 minutes

This block proves the repair survives contact with the actual song.

Pressure is anything that removes your ability to stop and restart. A click. A backing track. One take with no restarts. Standing up. Recording. Another person in the room. Pick one constraint and keep it for the whole week so the sessions stay comparable.

The failure this block catches is specific: the passage is clean in isolation and still collapses inside the song, because the bar before it and the bar after it were never part of the repair. Play into the passage and out of it, not just through it.

5. The cold-start test, 5 minutes

This block measures what survives with no runway.

Play one piece you claim to know, once, cold, with no practice run and no restart. Everyone plays better on the third attempt. The first attempt is what you would have on a stage, at a jam, or when someone hands you a guitar.

Keep a rotating list of four to six cold-start pieces and note two things after each: whether you finished, and where it broke. That break point becomes a candidate target for the next session, which is how the routine feeds itself. The song readiness score checks the same piece against the parts that usually fail under pressure.

This is the cheapest block and the only one that measures anything. When the session gets cut short, cut something else.

The two session lengths, side by side

The 50-minute session:

• 5 minutes. Warm-up with intent.
• 15 minutes. The one repair target.
• 10 minutes. Tempo work on that unit.
• 15 minutes. Repertoire under pressure.
• 5 minutes. The cold-start test.

The 25-minute session:

• 3 minutes. Warm-up with intent.
• 10 minutes. The repair target, with the tempo ladder run inside it.
• 8 minutes. Repertoire under pressure.
• 4 minutes. The cold-start test.

Notice what did not happen: the short session is not the long session halved. Tempo work merges into repair because both are working the same unit. The pressure block absorbs most of the cut because it is the block that tolerates being short. The cold-start test gives up one minute, because removing it saves almost nothing and costs you the only measurement in the session.

How to swap material in without rebuilding the routine

Only the inputs change. The blocks do not.

• Warm-up. The motion your repair target needs, at a tempo where it is clean.
• Repair. The smallest failing unit from last session's pressure block or cold start.
• Tempo. That repaired unit, plus the tempo the song actually runs at.
• Pressure. The song the unit belongs to, played under the one constraint you kept all week.
• Cold start. The next piece on your rotating list.

The repair input is the one that matters. It should come out of your own last session, not out of a feed. If you cannot say where this week's target came from, you picked it because it looked interesting, and the routine has quietly turned back into lesson collecting. Strumly's guide to designing a practice routine covers the same problem from the routine-design side.

Retire a cold-start piece once it survives three cold starts in a row and put a harder one in its place. A list you always finish has stopped telling you anything.

What should change week to week

The blocks hold their shape across a month. The emphasis inside them moves.

1. Week 1. Record a baseline before deciding anything. The repair target is whatever that recording exposes, not what you assumed on the way to the guitar.
2. Week 2. Same target, bigger unit. Two bars become the phrase; the phrase becomes the section it sits in.
3. Week 3. The target moves to the seams, meaning the bar before and the bar after. Pressure gets the longest block this week.
4. Week 4. No new repair. The session is pressure and cold start only. Record the final take beside the baseline and name the change the two recordings actually support.

Change one variable at a time. If you swap the target, the tempo ceiling, and the song in the same week, the routine can no longer tell you which change did anything.

Three signals say a target has finished its work: the error rate stops dropping across two consecutive sessions, the unit stays clean at the song's tempo for three sessions running, or the cold-start test starts breaking somewhere else. Any one of them means pick a new target. Two of them mean you should have picked one already.

The ten-minute day

Do not run all five blocks at two minutes each. A two-minute repair block is not practice, it is a checklist you ticked.

A ten-minute day is a repair day:

• 2 minutes. Warm up on the exact motion the target needs.
• 6 minutes. The repair unit, recorded, and nothing else.
• 2 minutes. One cold start, one piece, no second attempt.

The tempo ladder and the pressure block are gone for the day, and that is correct. Both need enough repetitions to mean something, and neither survives being handed ninety seconds.

Mark the session short in whatever you use to log it. A week with three ten-minute days and two full sessions is a fine week, but if the log does not say so, you will read the flat cold-start results as a plateau instead of as a short week.

Questions about this routine