Introduction: Too Much Data… Not Enough Clarity?
Today, every running or cycling workout generates an impressive amount of data: distance, duration, speed, heart rate, watts, cadence, estimated VO2max, recovery score…
With Garmin, Strava, Kinomap, and other connected apps, everything is measured. But one key question remains: How do you analyze your training data to actually improve?
Collecting numbers doesn’t guarantee performance. In fact, misinterpreting metrics can create confusion, stress… and sometimes even slow down your progress.
The goal isn’t to have more data. The goal is to understand which training metrics are truly reliable and useful for improving performance in running and cycling.
In this article, we’ll structure training data analysis from the fundamentals to the more technical aspects, so you can progress intelligently—without drowning in numbers.
1) The Foundation of Progress: Consistency and Weekly Coherence
Before analyzing heart rate or watts, you need to return to the foundation of long-term performance: consistency.
The first metric to observe isn’t technical. It’s your regularity.
The real question isn’t: “How fast am I running?”
But rather: “Am I training every week?”
An athlete who trains three times per week for three months will almost always progress more than someone who has one exceptional week followed by two weeks without training. Weekly consistency is decisive.
If your volume looks like this: 5h → 0h → 3h30 → 0h
The body doesn’t adapt. It absorbs stress.
By contrast, a stable progression like: 3h → 3h10 → 3h15 → 3h30
Allows gradual and sustainable adaptation. It’s not massive volume that drives improvement. It’s continuity over time.
When analyzing your training data, first check:
- The regularity of your sessions
- The stability of your weekly volume
- Continuity over 4 to 8 weeks
If this foundation is solid, you’ve already built the core of your progress.
2) Intensity: Measuring Your Real Efficiency
Once consistency is established, your analysis can become more precise. The key question becomes: Am I more efficient at the same perceived effort? This is where metrics start to matter.
In Running: Heart Rate and Pace
In running, the two most reliable metrics to analyze progress are heart rate and pace per kilometer (or mile).
Unlike cycling, speed in running varies relatively little on similar terrain, making it a relevant performance indicator.
The method is simple: compare your pace at a similar heart rate.
If two months ago you ran at 6:00/km at 150 bpm, and today you run at 5:40/km at 150 bpm, you’re more efficient. It’s not absolute speed that matters—it’s your ability to go faster for the same energy cost.
In Cycling: Power Comes First
In cycling, the logic changes. Speed depends heavily on wind, elevation, aerodynamics, and environment. It’s therefore unreliable for measuring true progress.
The most reliable cycling metric is power output. Watts directly measure mechanical effort. If you can sustain more watts—or maintain the same power for longer—your progress is objective and measurable.
In general, the hierarchy of reliable cycling metrics is: Watts → Heart Rate → Cadence → Speed
This hierarchy prevents you from overvaluing less relevant data like speed.
3) Cadence: The Often Overlooked Technical Indicator
After raw intensity comes movement quality. Cadence, in both cycling and running, is often underestimated in training data analysis. Yet it reflects technical efficiency and muscular economy.
In Cycling
Cadence is measured in RPM (revolutions per minute). A smooth, stable cadence—typically between 85 and 95 RPM in endurance riding—helps distribute effort and limit excessive muscular fatigue.
A cadence that’s too low increases muscular strain and accelerates fatigue. Over time, maintaining a more stable cadence at similar intensity improves efficiency.
In Running
In running, cadence is measured in steps per minute (SPM). An efficient cadence generally falls between 170 and 190 steps per minute, depending on your morphology.
A cadence that’s too low often results in overstriding, increased ground impact, and higher energy cost. Gradually improving cadence (without forcing it artificially) can enhance running economy and reduce joint stress.
This type of improvement rarely produces spectacular numbers… but it’s fundamental in the long term.
4) Intensity Distribution: Avoiding the “Silent Plateau”
After analyzing consistency, intensity, and cadence, one key factor remains: intensity distribution.
Many athletes train constantly in a moderate zone—neither easy enough for recovery nor hard enough for meaningful adaptation. This is often called the “gray zone.”
When analyzing your training data, check:
- Whether you truly include easy sessions
- Whether you include high-quality sessions
- Whether recovery has a proper place
A balanced week typically includes:
- One or two quality sessions
- One or two easier endurance sessions
- Adequate recovery time
A balanced week generally includes one or two quality sessions, one or two easier endurance sessions, and sufficient recovery time.
Analyzing metrics should help you balance your training load, not stack it up.
5) Secondary Metrics: Useful, but Relative
Estimated VO2max, recovery scores, training status… These indicators can provide interesting trends. However, they’re algorithm-based and heavily influenced by sleep, stress, and accumulated fatigue. They should never become the center of your training analysis.
If your estimated VO2max drops slightly but your efficiency at comparable effort improves, your progress is real. Concrete performance always outweighs estimations.
6) Perceived Effort: The Invisible but Essential Metric
No training data analysis is complete without considering how you feel. Sensors measure external performance. Your sensations measure internal state.
If a previously easy pace feels difficult for several weeks, it may indicate:
- Accumulated fatigue
- Insufficient recovery
- Professional stress
- Disrupted sleep
Progress isn’t just numbers. It’s the balance between objective data and subjective perception.
A good analysis always combines: Numbers + Feelings + Life Context
Conclusion: Data Doesn’t Create Performance — Interpretation Does
Training data analysis isn’t about collecting metrics. It’s about prioritizing them.
Your priorities should be:
- Consistency
- Weekly coherence
- Improvement at comparable effort
Then come:
- Power (cycling)
- Pace (running)
- Cadence
- Intensity distribution
And always, your sensations. A sensor gives you numbers. You give them meaning.
And with tools like Kinomap, you already have everything you need to analyze your training data intelligently—without drowning in unnecessary metrics.
Par Fanny Marre
Entraîneur en cyclisme, running et triathlon
fannymarre99@gmail.com
See previous coaching articles:
LACTATE THRESHOLD IN RUNNING: HOW TO TRAIN IT TO RUN FASTER FOR LONGER
OVERTRAINING : SIGNS, CAUSES, AND AN EFFECTIVE RECOVERY PLAN
CHAMPION’S MINDSET : EXERCICES TO IMPROVE YOUR PERFORMANCE IN CYCLING AND RUNNING