Blog (Page 2)

Venture Beat’s view

Mark Sullivan writes: “There’s no shortage of apps and devices that generate numeric data but stop there. We need technology that interprets it all, but, so far, the attempts we’ve seen to do that have been clunky. New Zealand-based Performance Lab [has] unveiled its answer to the problem”.

Here’s the full article.

Why are we doing this?

For more than 20 years Performance Lab worked face-to-face and side-by-side with clients who wanted to change their lives through exercise, or achieve their fitness and competitive goals. There was never any doubt about why my team came in to work in the morning: it was to make a difference in those people’s lives.

We know how to do that, and we’ve learned how to use data to inform that process. A few years ago we saw that a moment in time had arrived where it was possible to scale this process. The specialist sensors we had been using were becoming commoditized and ubiquitous. Remote exercise coaching on a mass scale had become possible.

We want nothing less than to change the trajectory of people’s physical lives. To do this, data and numbers are not enough. Motivation and social pressure is not enough. Meaningful and trustworthy advice delivered at exactly the right time is what is needed, and is what we are making possible. That’s because timely and relevant advice improves performance, which leads to success. And it turns out that success is the best motivation of all.

What to do with a broken heart rate

As a coach who’s had to get runners and cyclists to strap on chest-based heart rate monitors for a few decades, I’m excited to see the emergence of more user-friendly options in fitness devices. But I’m confused about the trend to see heart rate as the basis of so-called coaching apps. Let me explain.

Heart rate is a reaction to something. In order to use it as an input to coaching advice you need to know what is causing that reaction – and heart rate is subject to a lot of variables. Is it heat? Is it altitude? Or is it effort? And if it’s effort, have you factored in how many days off the athlete has had, and how this will translate into delays in the heart rate response?

If an app or device hasn’t factored in these things then it risks saying something meaningless or even outright wrong in response to the heart rate reading. That’s why most users quietly put their fitness wearables in a drawer after a few months. The answer of course is to base coaching advice on a full spectrum of parameters — which is the approach we’ve taken at Performance Lab. Not reporting back on each parameter, but running algorithms that integrate all this information to build up a picture of what’s going on and how the athlete is performing, and then provide advice based on that knowledge. I’m excited to see more and more devices going beyond just heart rate monitoring and adding barometers, GPS, accelerometers and so on. Single device coaching is now a real possibility, if you know how to interpret the data.

Closing the coaching loop with context

Coaching is actually pretty simple. A coach tells you:

  1. What you should do
  2. What you in fact did
  3. How you performed
  4. What you should do next.

You’ll see that it’s a loop: step 4 takes you back to step 1.

Context is the key to every step. Most fitness devices are like a coach who’s sitting in a dark room, yet on the phone telling you what to do, based on a few numbers like heart rate.

What you should do needs to be informed by recent performance and what your goals are. It needs to be informed by your fatigue levels. And it needs to be informed by the terrain!

What you did also needs context. That you ran at a good stride rate up a hill can look very much the same as slowing down and getting tired on the flat (and that’s what most apps seem to think is going on).

How you performed is, obviously, also very much dependent on knowledge of what you did.

So although coaching is simple, packing it all into a tiny device is hard.

More about trust

The two things I always hear from the users of incumbent fitness trackers is:

  1. What they tell me isn’t interesting; or
  2. I don’t trust what they are telling me.

In one camp we have devices or apps that lack ambition and merely report back what they are measuring (steps, miles, etc). This doesn’t do much to change behaviour or performance, and is the main reason why 50% of users put their new devices in a drawer after a few months.

In the other camp we have devices or apps that attempt to coach. But because they don’t have a full picture of the user’s context, the advice doesn’t ring true. Most commonly they tell a user that they are out of their ‘training zone’ — because they detect elevated heart rate, for instance. But if this advice is ignorant of the fact that the user just started a steep hill, or a scheduled sprint session, or any of the other 57 legitimate reasons for having an elevated pulse, it is worse than silence.

As Kerri has mentioned, trust is the key to coaching. That’s why our ARDA Coaching Engine is based on combining multiple data streams from a range of parameters to understand the user’s geography (terrain, weather, incline); physiology (fatigue, effort, endurance); and history (goals, recent performance, outcomes). You need to know what’s actually going on in order to provide meaningful advice. And for users to trust you, it’s important that they realise that you know what’s going on.

Our main value proposition to partners who are seeking to develop world class coaching platforms is the trust that underpins that coaching.

Focus on trust, not motivation

People often ask me how we got a 90% half-marathon completion rate for the hundreds of novice runners we worked with over the years. Actually, they are usually more specific than that. They ask: “How did you motivate them to do that?”

I try to explain that motivation is a bit of a red herring. What we do with any athlete, novice or elite, is understand their physiological starting point (These days, using wearable technology, we only need someone to go for a couple of quick jogs to get a read on this). We then put in place a training programme, and tell the athlete what she’s going to experience on that first run. Then we tell her why she’s doing a different kind of run next time, how it’s going to feel, and what the performance improvement is going to be for her strength and endurance. And so on, until the goal is reached.

The effect of our predictions coming true (this is how you’ll feel, and this is the performance improvement you’ll experience) is magical. The athlete trusts the coach. Because she trusts the coach, she sticks to the plan. Because she sticks to the plan the performance improvements are achieved. Sticking to a plan is motivating. Nailing performance improvements is motivating. And then going the distance is almost trivial.

For a wearable fitness device to succeed, the user must trust it, just like a coach. That means getting things right in the moment: accurate information and advice that makes sense right then and there.

Why coaches care about weight

Remember high school physics, where an object’s energy was proportional to its mass multiplied by the square of its speed? When you start a new fitness program with a coach or if you track fitness through an app, they will always ask your weight. Sure, it’s a good way to track how much weight you might lose on a certain program; however, how much you weigh is also an important stat to base how you exercise.

There are clever ways to work out your speed, which together with your weight (or technically your mass) means we can now work out your energy – which lets us talk to you about calories burned and the effort you’re putting into your exercise program or training. And because we know all sorts of things about your technique, the terrain you’re on, and your recent performance, now we can coach you.

The ARDA Coaching Engine essentially bottles all this into a software program and can act as a personal coach. But first, you must divulge the numbers on the scale (don’t worry, we won’t tell anyone).