How to get the most out of Training Peaks

Multisport and performance lifestyle coach Joel Enoch looks at how you can use online platform Training Peaks more effectively to inform your training

Young runner looking on training results in phone. Synchronisation music with headphones air pods.

The boom in sports technology means that amateur athletes and sportspeople now have access to training data the likes of which the Olympians of Sydney and Athens could only have dreamed.


Culturally, triathlon is a sport that embraces technology quickly, and so from GPS watches to power pedals, carbon bikes to carbon shoes and head-up display goggles to real-time blood glucose monitoring, triathletes have a wealth of options that may help optimise their training and performance.

The use of technology in sport is perhaps optimised by the widespread use of Training Peaks for athletes and coaches.

What is Training Peaks?

Launched around 20 years ago, Training Peaks was developed by lauded triathlon coach Joe Friel, his son Dirk and their friend Gear Fisher as a way to programme, log and analyse training data.

The system attributes a Training Stress Score (TSS) to sessions based on power, heart rate or your own subjective measurement, on the principle that going as hard as you can for 60mins equals a score of 100 TSS.

This data is then used to predict your fitness (CTL), fatigue (ATL) and form (TSB) suggesting how effective your training may have been and how ready you might be to perform.

Who uses Training Peaks?

Over 20 years after its inception the platform is the most widely used coaching and training log platform in the world, utilised by pros, Olympians, national governing bodies such as British Triathlon and Tour De France cycling teams, as well as hundreds of thousands of amateurs and coaches globally.

I am one of that number, using Training Peaks as an athlete and coach, in addition to using their analysis software (WKO5) to dive deeper into the data produced. However, I have a few issues…

Are there any potential issues with Training Peaks?

Firstly, there’s an obvious need for accurate data. Without it, the information given is little better than computer- generated guesswork.

With watches and power meters ever more advanced, you’d think this wouldn’t be an issue, yet data spikes and dropouts occur regularly and if you don’t look over the files you might not spot these.

Also, if basic metrics like body weight, HR zones or threshold aren’t accurate, TSS estimates will also be poor. This takes effort, as while WKO predicts your bike threshold (FTP), Training Peaks doesn’t, and you must manually update it when it changes, requiring accurate data from testing (and if you’re testing your FTP, then do you need an algorithm to tell you how much your FTP might have shifted?).

In addition, run and bike HR zones can be quite different, so you need to input these accurately too, but the biggest issue in terms of data collection for the triathlete is the inaccuracy of swim and S&C stress scores.

For example, let’s say you swam 3,000m with fins at 1:40mins/100m and the next week swum 3,000m at 1:40mins/100m with paddles on – would those two sessions feel the same? The answer is likely to be a resounding no.

However, based on the pace, TP would give you the same TSS for both. The issue with gym work is even greater.

Officially, TSS for S&C work should be dictated by HR, but during heavy lifting, HR can remain very low in endurance athletes while the load might be extremely high and cause a significant amount of stress to the muscles.

You may only get a TSS of 25 but feel like you’ve done 100TSS race the next day. Similar issues exist with over-gear work on the bike and, to a lesser extent, hilly or windy run sessions.

In TP, sessions change colour according to how accurately they were completed, but as a result, training can become a box-ticking exercise where turning all sessions green is more important than quality, recovery or listening to what your body is saying.

Data versus athlete feedback

While useful, Goodheart Law states that ‘when a measure becomes a target it ceases to be a good measure’. This can impact coaches too.

If data suggests an athlete is overtraining and but subjectively feedback that they feel fine – which does the coach believe? The computer, or the person in front of them?

Care must be taken to balance these two valid points of view. To do this, maybe there should be more reliance on the subjective measures of training available on Training Peaks.

Further, if my CTL (fitness) drops and my TBS is the same, the prediction is that I will be slower, but is this true? Also, athletes have limits of time and energy and what if you reach that capacity? Are you to then believe that you cannot become faster?

Surely these assumptions can be questioned in any number of ways, yet many athletes are slaves to such data. Stress scores seem more weighted towards volume than power, which doesn’t represent how stressful a session of 6 x 50m sprints can be and the data can’t ‘see’ technique improvement through a session or understand the stress of technique work.

How to make the data work better for you

Psychologically we can be derailed if our numbers look ‘worse’ and so we need to couple data with better questions such as: Am I improving? What have I learned? What will I do differently? When will I do it? How was the session? And not, ‘Did I hit the numbers?’.

It’s too easy to stop asking these questions when presented with data – potentially making us less self-assured, self-reliant, and resilient.

Yes, TP provides a lot of information, but information doesn’t equal knowledge and knowledge alone does not lead to wisdom – it’s how we action the information we have that demonstrates wisdom (or otherwise).

That requires a more holistic view, and it might be that rather than a ‘coach sets – athlete does – coach analyses’ system across a four-five week block, we instead migrate towards readiness measurement tools such as Whoop, FitrWoman or Rewire to better inform what training is undertaken based on the athlete, not the planned session.

Understanding the limitations and the benefits

In conclusion, Training Peaks, I believe, is the work of a great team of people doing a good job. It’s a good tool for logging training and has some useful analysis features. But, in my opinion, it lacks the capacity to understand the full context of your training or your readiness to perform as well as you can.

A useful development would be the capacity for TP to utilise recovery data such as HRV or its own fatigue index TBS to suggest changes to the load of sessions if someone appears to be over-reaching.

While it is one of several platforms that you can use to help you train more effectively, how you use such tools is important and there is a need to understand the limitations of such tech as well as the potential benefits.

This article is written with the simple aim of highlighting some of those areas for your consideration. Ultimately, your body and mind should be the central factor in deciding when to push and when to back off and your experience of the sport is more important than a fitness number generated by tech – no matter how clever that tech is.

Training Peaks were contacted for a response to this article. 

Joel Enoch is an award-winning triathlon coach to multiple world and European champions, performance lifestyle advisor to TriWorks Edinburgh and is soon to launch a new project to take the idea of why we do what we do into the wider triathlon community – watch this space!


Top image: Getty Images