How to train for a Giro d’Italia podium – one rider’s approach, analysed

Rarely do we get such detailed insight into a top rider's Grand Tour preparation.

The subject of our story is somewhere in this photo.

Matt de Neef
by Matt de Neef 31.05.2024 Photography by
Kristof Ramon and Cor Vos
More from Matt +

Grand Tours are called that for a reason. Three weeks is a long time to race your bike; to be so focused on the repetitive daily cycle of racing and recovery. And of course, a successful Grand Tour is built on more than just three weeks – actually preparing for the race takes many months of work, especially if you’re vying for the upper reaches of the general classification.

A new research paper out of Italy shows exactly how much work is involved in such a build-up, following the “day-by-day training and racing characteristics” of one rider in the five months leading up to a podium finish at the Giro d’Italia. And as we’ll see, there are some interesting insights to be gleaned.

How they went about it

There wasn’t anything particularly complicated about the method behind this study. To start with, University of Genoa researcher Gabriele Gallo and his colleagues collected one rider’s training and racing data from the 175 days between the start of their season and finishing the Giro d’Italia. Then, the amount of time the rider spent training or racing each day was catalogued, as was their total load for that day’s efforts, based on their average power. Specifically, their day’s total work (in kJ) was defined as: (average power output (W) x duration of the ride (s)) / 1,000.

Each day’s riding was broken down into three intensity zones, to see how hard the rider was going on any given day, and how that intensity changed throughout the 152 days leading up to the Giro:

The zones used were:

From your own riding you might be more familiar with intensity being broken up into more than three zones – five or seven zones most commonly – but Gallo tells Escape that three makes more sense physiologically.

“In exercise physiology we only have three physiological ‘zones’ separated by two thresholds: the first and second ventilatory/lactate thresholds which separate Z1 / Z2 and Z2 / Z3 in a three-zones model and that can be approximate to some FTP percentage, like in the article,” he says. “For example there is no physiological threshold to divide Z1 and Z2 in a 5/7 zones model – it is just speculation.”

In order for a day’s riding to be included in the study, power data from at least 80% of the day had to be present. On days where 80-100% of the day’s power data were included, “time spent in each zone was increased proportionally.”

Who’s the rider?

So far we’ve referred to the subject of the study only as “the rider”, which mirrors the way the authors refer to him in their paper. The rider isn’t named at any point, nor is the team he rode for, which season the data is from, or what position he finished in the Giro. But there are a bunch of clues which can help us work out who the rider is.

For starters, the rider was 26 years old at the time of the Giro in question, they weighed 64 kg, they were 173 cm tall, they had a VO2max of 81 ml/min, and their 20-minute record power output was 6.6 W/kg. Some of those data points are more useful than others, but when we combine them with the rider’s racing calendar in those 175 days, a clearer picture emerges. 

Here are all the races the rider did, based on what we’re told in the paper:

Gallo didn’t want to reveal or confirm who the rider in question is, but by my reckoning, there’s only one rider and season it could be. Want to try and work it out yourself? I’ll give you the answer after this image.

A graphic from the paper showing which races the rider did (see the brackets at the top) and their total work per day in the latter part of the study period (y-axis).

From my observations, the subject of this research is none other than Mikel Landa who rode to third overall at the 2015 Giro d’Italia (behind Alberto Contador and teammate Fabio Aru) in what was a breakout ride for the then-Astana rider. Here are the races mentioned above:

A few other things to note from Landa’s start to the 2015 season:

Observations from the data

Alright, so what does Landa’s data tell us about his lead-up to and participation in the 2015 Giro d’Italia? Here are a series of observations:

This graphic shows Landa’s efforts from the start of his season, through to the week after Catalunya. Note the increasing percentage of medium-intensity (yellow) and high-intensity (red) work as the months wore on.
Note how much more high-intensity work there is at the Giro (days 153 to 175), compared with earlier in the season (see image above). His Giro stage wins came on days 168 and 170 (separated by a rest day).
A graphic from the research paper which shows the amount of time Landa spent doing medium-intensity (top) and high-intensity (bottom) efforts during his training.
Landa winning stage 15 of the Giro, his first of two stage wins.
The most consistent block in Landa’s whole season. This period ended around two and a half weeks before his first race of the year: the Volta a Catalunya.

Concerns and conclusions

As with any study, there are a few questions that should be asked when considering this project’s findings. For one, and as mentioned already, there are quite a few days in the dataset – 10 to be precise – where Landa’s power data wasn’t available. And with nine of those days being race days, it’s hard to say we have the complete picture of Landa’s pre-Giro preparation. It doesn’t matter terribly much from a big picture perspective, but it would certainly be interesting to dive into the details of his stage win at País Vasco, for example, which isn’t available here (and as far as I can tell, Landa isn’t on Strava).

When asked why power data wasn’t available for every day of the study, Gallo gave a few reasons: “TT with no power meter or power meter didn’t work appropriately: calibration, battery …”

The black bars are day where power data wasn’t fully available (if at all). Note that all of these days, bar one, happened in races. Stage 1 of the Giro (day 153) was a teams time trial and it seems likely he didn’t have a power meter on his bike for that day (or it wasn’t working properly). The day before was the only non-race day in the study where power data wasn’t available. I’d guess that he was doing a final preparation ride on his TT bike that day.

The other question that springs to mind is whether the age of the dataset  – almost a decade old by this point – makes it less relevant today than it might be. While Gallo wouldn’t confirm the age of the dataset (or the rider involved), he did say “I think that there would not be too many differences in the training principles highlighted in the article”.

That said, one training evolution that could warrant a follow-up with more recent data is the increasing reliance on altitude training. While it’s not clear from the data itself how much time Landa spent at altitude in the lead-up to the 2015 Giro, he’s likely spending more in his present build-up to the 2024 Tour de France, say. It would be interesting to see whether that change would lead to different training strategies as well, beyond the location where that training is done. 

Landa (right) on the final podium of the 2015 Giro d’Italia, with overall winner Alberto Contador and Fabio Aru.

Gallo also made the point that, regardless of the age of the data, “this case study shows a possible way rather than the ‘right’ way” of preparing for a Giro podium. That’s an explanation that rings true. Different strategies work for different riders, and no two riders are likely to have the same exact plan in building up for a Grand Tour. 

Even for an individual rider, it’s unlikely that there’s any one optimal strategy, and even if there is, there’s nothing to say that optimal preparation translates to a good result. Landa likely followed a similar approach to preparing for other Grand Tours throughout his career but he certainly hasn’t finished third or better in every Grand Tour he’s done (his third overall at the 2022 Giro is his “only” other Grand Tour podium).

By the same token, Tadej Pogačar’s lead-up to the 2024 Giro d’Italia looks very different to that of his GC ‘rivals’ but had any of those riders followed Pogačar’s program to a tee, it’s hard to imagine the result being any different. That is to say: there’s much more to a rider’s Grand Tour performance than the specifics of their preparation.

And yet, it’s fascinating to dive into the details and learn how the best riders approach the longest and hardest races in the world. And in the case of Landa’s successful tilt at the 2015 Giro, it looked something like this: a long, steady build-up with a focus on base miles; a gradual increase in intensity as the season began; and a bunch of hard racing to tune the engine, before finally tapering towards the Giro. Easy, right?

What did you think of this story?