A training plan is not rocket science — but it is more than “go running three times a week”. Work through the seven steps in this guide in order, and you will build a plan that fits your goal, your current fitness and your daily life. One that does not leave you guessing, in the week before your race, whether any of it was enough.

Step 1: Define the goal precisely

“Getting fitter” is not a goal — it is a wish. A usable training goal has three properties: it is measurable (a distance, a target time), it has a date (race day), and it is realistic for your current level. “Sub-4-hour marathon on 18 October” is a goal. “Run faster someday” is not.

Also decide on the goal's priority: an A-race gets the full build-up including a taper, a B-race is trained through, and a C-race serves as a test under race conditions. This distinction later determines what the final weeks before the date look like — details in our article on season planning with A-, B- and C-races.

Step 2: Assess your current fitness honestly

The most common mistake in self-made plans: they are built on the fitness you would like to have — not the fitness you actually have. So establish three numbers before you plan a single week:

  • Current weekly volume: how many hours or kilometres do you really train in a normal month — not in your best week?
  • Current performance level: a 5K time trial, a 20-minute FTP test on the bike or your last race provides a reliable pace or power baseline. From a 5K test you can derive every training zone.
  • Available time: how many sessions per week are sustainably possible alongside work, family and sleep? Plan with the minimum, not the ideal.

If you cheat here, you build a plan that collapses in week three. Honesty about your starting point is the cheapest injury prevention there is.

Step 3: Set the timeframe and periodisation

Count the weeks backwards from race day to today. That is your planning horizon — and it should match the goal: 8–12 weeks is enough for a 10K, 10–14 for a half marathon, 16–20 for a marathon. Plan too short and you have to ramp up too fast; plan too long and you lose focus.

Then divide the timeframe into blocks. The classic three-part split has stood the test of time: in the base phase (the largest block), volume in easy zones takes priority. In the build phase, threshold and race-specific work is added. In the peak phase with the taper, volume drops and freshness rises. The principle behind it — the structured alternation of load and recovery — is called periodisation and is the backbone of every plan that works.

Just as important is the wave-like build: three to four weeks of gradually rising load, then a recovery week with 30–40 % less volume. Without these waves there is no supercompensation — only creeping fatigue.

Step 4: Build the weekly structure

Now it gets concrete: distribute your sessions across the week. Three rules prevent most problems:

  • Avoid hard on hard: between two demanding sessions (intervals, threshold work, a long run with pace blocks) there are at least 48 hours or an easy session.
  • The long run is fixed: it sits on the weekend when time allows, and grows by roughly 5–10 % per week during the build — never in jumps.
  • Do not skip strength training: once or twice a week, 20–30 minutes of core and leg work. It costs little and measurably lowers injury risk.

An example with four run sessions per week: Tuesday easy run, Wednesday strength, Thursday quality session (intervals or threshold), Saturday relaxed run or bike, Sunday long run, Monday off. The order is interchangeable — the pattern of load and recovery is not.

Step 5: Distribute the intensity correctly

The best-evidenced distribution key is the 80/20 principle: about 80 % of training time in easy zones (conversational pace), 20 % moderate to hard. In practice it fails in two places: the easy sessions are run too fast (the notorious “grey zone” with medium fatigue and little effect), and the hard sessions are not hard enough.

Control your zones via heart rate, pace or perceived exertion — whatever you use, use it consistently. For easy work the rule is: you must be able to hold a conversation. For the hard parts, orient yourself around the anaerobic threshold, which you can estimate from a test or from your 5K/10K time.

Step 6: Build in adjustment and flexibility

A plan that cannot forgive missed sessions is not a plan — it is a wish list. Colds, business trips, family commitments: they will happen. Set the rules for them before the first missed session occurs: a skipped workout is not made up — it is deleted. If you miss several days, re-enter with an easy week instead of a backlog catalogue. And review the plan briefly once a week — but do not rebuild it. Volume adjustments of 10–15 % are fine; anyone who tears up half the concept every week is training without a plan that happens to be written down.

Step 7: Measure whether the plan works

Without monitoring, you cannot tell whether the plan is working. Track at least three signals: morning resting heart rate (the trend, not the daily value), pace or power at the same heart rate (if it gets faster, fitness is growing), and your subjective sense of sleep and legs. If you want it more systematic, work with load metrics like CTL, ATL and TSB — they turn feeling into numbers and show whether you are actually getting fresh during the taper.

The five most common mistakes in self-made plans

  • Ramping up too fast. More than roughly 10 % added volume per week is the classic route into overuse injuries.
  • No recovery weeks. Four consecutive build weeks feel productive — and end in a performance plateau.
  • Everything at medium pace. The grey zone feels like work but delivers neither base nor speed.
  • Copying a plan from the internet. Someone else's plan does not know your starting point. Fine as inspiration, risky as a blueprint.
  • No taper. Anyone who “makes full use” of the final two weeks gives away on race day the freshness they trained months to earn.

When an AI plan is worth it

The seven steps work — but they take time and require training knowledge you first have to acquire. That is exactly where an AI training plan comes in: it handles periodisation, zone logic, weekly waves and tapering automatically, and adjusts the plan every week to the sessions you actually completed and your recovery. For an honest comparison of both approaches, read AI training plan vs. traditional plan. And if you prefer building your own plan anyway, you can use Peakora as a control instance: the recovery monitor shows, independently of any plan, whether load and recovery are in balance.

Frequently asked questions

How many weeks should a training plan cover?

It depends on the goal: 8–12 weeks is enough for a 10K, 10–14 for a half marathon, and 16–20 weeks for a marathon. What matters is that the plan builds on your current fitness — not on the fitness you wish you had.

How many training sessions per week make sense?

For most recreational athletes, 3–5 sessions per week is ideal. Consistency matters more than volume: three fixed sessions sustained over months beat six sessions that fall apart after two weeks.

How should I distribute intensity in my training plan?

Follow the 80/20 principle: about 80 % of training time in easy zones (conversational pace) and 20 % moderate to hard. This distribution is best supported by studies of both elite and recreational athletes.

How often should I adjust my plan?

Review it briefly once a week, but change it only gently: adjust volume by no more than 10–15 % and avoid complete rebuilds. Every 3–4 weeks, schedule an easier recovery week.

Should I build my own plan or use an AI training plan?

If you understand training science and have the time, you can build a good plan yourself. An AI plan like Peakora handles periodisation, adjustments for missed sessions and load management — and adapts weekly to your data.

Let Peakora build your plan

Periodisation, 80/20 logic and automatic adjustment for missed sessions — tailored to your goal and your recovery. Start for free, no credit card required.

Start for free