M7 vs Fitbit Flex - Getting an Accurate Step Count

by Hamilton Chapman

I'm a sucker for iPhones. Every year I tell myself that the new one won't be enough of an improvement to convince me to upgrade. Every year I end up buying the new one.

This year was no different and so I am now the owner of a Space Grey iPhone 5S.

Touch ID is impressive. Having 4G on the go is amazing. However, perhaps the feature of the new iPhone that I was most anticipating was the M7.

I'm a fan of the idea of quantified self and paersonal analytics and have played around with it in the last 6 months or so. I bought a Fitbit Flex when it was released in the UK and had dabbled with using Brett Terpstra's Slogger to track some of my digital activities.

When the iPhone 5S was announced it was the M7 that grabbed my attention. It brought with it the potential to track how I was moving around in the same way that the range of wearable tracking devices could already but without having to wear anything dedicated to tracking me. It would work with just my phone in my pocket.

David Smith's Pedometer++ was one of the first apps I installed when I eventually received my iPhone. I decided to keep on wearing my Flex and to use Pedometer++ to track the number of steps I took during a day.

Below is a graph detailing the numbers that I got from each of the devices.

The difference between the two is pretty consistent. The average number of steps tracked by the Flex was 8479 compared with the M7's 6995. That works out as 22% more steps being tracked by the Flex. 

I thought this seemed pretty high. There are of course some possible explanations. There are a few times during the day where I will not have my phone on me and so having the Flex on me all the time, other than when it's charging, would obviously add on a few steps. These moments aren't that frequent though. For example, I don't take my phone in the shower, but do keep my Flex on. Sometimes I will have my phone on my desk and quickly go off to the toilet or go and grab something from the fridge and so there are again a few steps not being counted by the M7. 

Then there are the discrepancies that will arise due to the Flex being worn on the wrist. Some things that aren't steps may be giving false readings to the Flex. This could range from brushing my teeth to shaking hands with someone. The few days where the M7's tally eclipses that of the Flex's are when the Flex will have been off my wrist, plugged in charging. 

The M7 is likely slightly underestimating my step count while the Flex is probably overestimating by a good chunk. My feeling is that the M7's numbers are more accurate (unfortunately for my health) but still fall slightly short of the real number.

I'll continue  to use both and will revisit the data in the future to see if the differences are the same.