Three Activity Trackers, Quantified Self

July 12th, 2014  |  Published in tech  |  2 Comments

Activity trackers really work for me, less for counting steps than all the stuff you do around them, like recording food and just generally thinking about activity levels.

Some of this will sound more complicated written out than it is in practice, but the benefits have been pretty clear because I’ve come to understand through observation and recording just how much activity, diet and sleep are key to my sense of wellbeing, how interdependent they are, and how being mindful of my general state where one or the other is concerned can help me make better decisions.

Three Pillars: Sleep, Food, Exercise

For instance, I know that I need about seven hours of sleep a night to feel well rested. Much more than that and I tend to be over-rested, and that triggers a bout of sleeplessness. So I try to keep my sleep steady: 6.5 — 7.5 hours a night is just right.

I learned how much sleep I needed by using a sleep-tracking alarm clock app that included a simple diary I could fill out at night (what did I do that day that might effect my sleep, e.g. caffeine intake, exercise, stress at work, working late, late use of computer/iPad), and in the morning (to record how rested I felt).

When I record my sleep and make it a habit to think about it, it can help me keep from making bad decisions. For instance, if I get only six hours of sleep, I can offset some of the fallout from that (feeling crabby and irritable) by getting some exercise first thing in the morning. Keeping track of sleep keeps me thinking about that dynamic, making it harder to say “oh man, I feel pretty bad … I’ll just skip the ride to work and take the Max in.”

Same thing with food: I know if I’m tired I’ll be more inclined to go after carbs and other cheap, fast energy. Thinking that through first thing in the morning really helps keep me from eating a bunch of stuff that’ll make me unhappier later in the day, when I crash from it.

That helps with mood and interpersonal stuff: Under six hours of sleep, and other people become pretty hard for me. Having that reminder that I’m probably not going to be receiving other people very well makes it easier to deal with things that might seem like provocations otherwise. And getting exercise generally lightens my mood.

In the big picture, it all comes together to help serve a longer term goal, which is losing some weight. When I’m in the habit of paying attention to all this stuff, I behave better, feel better, and I lose weight. When I stop paying attention, I behave worse, feel worse, and I gain weight.

The total amount of time I spend per day paying attention to this stuff comes out to maybe five minutes, tops: about .3 percent of my day. Seems worth it.

I’ve had three activity trackers in the past several years. Here are some notes in case you’re considering buying one for yourself.

Jawbone Up

I bought one of the earliest Jawbone Up’s a few years back. I don’t know what else was on the market at the time besides FitBit, and I think the Up was the first step tracker you could wear on your wrist (vs. clipping into a pocket) or wearing a giant, bulbous device on your arm.

Jawbone got a lot right about the Up: It looked nice, the iPhone app that came with it looked and worked really well, and the device integrated with calorie-counting services like MyFitnessPal or digital scales like the ones from Withings.

The Up could do a few things:

  • It could track steps, and if you tell it that some activity it tracked wasn’t walking but was actually something like jogging or riding a bike, it could adjust the calories it calculated you burned.
  • It could track sleep quality. Tell it you were going to sleep, and it would record periods of motion and let you know how well you were sleeping at night.
  • It could provide gradual alarms in the morning, gently vibrating you awake as you began to naturally stir at the top of your sleep cycle.

The Up was made of a textured rubber with silver highlights, and it looked pretty nice: Closer to a piece of jewelry than a step tracker. The FitBit Flex and Garmin Vivofit look pretty clunky by comparison.

The Up took input via pressing one of the ends of the wristband and it offered feedback via tiny embedded lights and vibrations.

In terms of ease of use, the first Ups were o.k. You had to snap an end off and plug it into your phone’s headphone jack to sync with the app. The caps become loose over time (especially if you’re an anxious twiddler who’s prone to fiddling with things like bracelet bits). Later models involved Bluetooth for sync but still require removal of that cap to charge the device (which lasts about 7 days on a single charge).

With all that stuff going for it, I wish I could still bear the thought of using the Up, but it has (or had, maybe it’s gotten better) one pretty big problem: It breaks a lot.

The issue appears to be endemic to the design. Unlike the Garmin Vivofit or FitBit Flex, which keep their guts in a discrete plastic lozenge with no moving parts, the Up hardware is built into the rubber bracelet. That means if you flex it too hard, you can break the innards.

“But Mike,” you’re thinking, “don’t flex it too hard!”

Sadly, since the Up is a bracelet with a pair of square ends that don’t clasp together, it’s quite easy to flex it too hard just by snagging it on something like a coat sleeve or a bag strap. I broke one tangling it up as I put on a backpack.

So, the first generation of Ups were pretty much a design catastrophe. Lots of people had problems with them breaking or just dying. Jawbone launched a trade-in program that it handled about as well as you can when everyone on the Internet is reporting that your product breaks when worn. I returned mine, got a replacement, and it broke, too. I returned the replacement and got another one, and it lasted a week before also breaking, at which point Jawbone just gave up and offered a refund to the entire world.

The first generation of Ups were so bad that Jawbone scuttled the product and went back to the drawing board, returning a while later with a new version that looked the same but promised to be more reliable. I broke two of those in the space of three weeks (I swear to god I wasn’t trying), and Jawbone — exhausted from its previous period of apologetic refunds — told me I could keep returning them during the warranty period or get nothing.

When the new Ups with Bluetooth syncing came out, I ignored them because a cursory glance around the ‘net showed people were still complaining about them breaking. It probably wouldn’t have mattered: Fool me once, shame on you. Fool me four more times after swearing to god you’ve figured it all out, shame on me.

Update: Just read the 1-star Amazon reviews to get an idea of how often they still seem to be failing.

Lesson: Cramming a bunch of fragile electronics into a container that’s comparable to a strip of thick garden hose in flexibility is pretty much a recipe for running an exchange program in perpetuity.

Pushback: Reviewers complain about the Up’s lack of a display or other feedback on where you’re at with your daily step goal. It didn’t really bother me. I would just do a sync at the end of the day and see where I came out.

FitBit Flex

After the Up, I got a FitBit Flex. Unlike the Up, the Flex puts all of its electronics in a non-flexible piece of plastic that slips into the pocket of a rubber wristband. It syncs via Bluetooth.

Where the aesthetics of the Up are sleek and somewhat fashionable, those of the FitBit are utilitarian and sort of dowdy. That’s the FitBit aesthetic in general, and if the life I got out of my FitBit is any indication, that’s as it should be. Sticking to a simpler design with less emphasis on making a pretty bangle means the things it would take to break a FitBit will generally involve taking it off your wrist and smashing it between large rocks or suffering an injury to your wrist that’s so severe you will not be wondering how many steps it took you to get into that situation.

Where the Up relies on button presses for input, the FitBit Flex takes taps. Sometimes that works o.k. For instance, it can tell you how far along you are to your daily step goal with a single tap, lighting up from one to five tiny LEDs. For sleep tracking, it’s not as good: You have to sort of tap-tap-tap-tap-tap the device, and it doesn’t offer the most meaningful feedback.

Since sleep and sleep tracking matter a lot to me, the FitBit fell a little short. Like the Up, you can tell it you’re going to sleep and it’ll track how well you rested. Unlike the Up, it doesn’t have a gradual wake alarm (though it can vibrate on your wrist at a set time).

Like the Up, the FitBit Flex is also a pretty good citizen of the wider fitness device/service ecosystem. It has an open API and it talks to smart scales, calorie counters and more. If you have a preferred app for tracking activities besides walking, the FitBit API can take that input and fold it into its reckoning of your daily activity.

The one complaint I have is that I managed to tear the wristband popping the little lozenge out of its pocket. That didn’t affect it too badly, but it did mean water had an easier time finding its way into that pocket. Digging a sweaty lozenge out of a rubber pouch doesn’t feel like the future at all.

Still, I had a positive enough FitBit experience that I really wanted to trade my FitBit Flex in for the next model up (the FitBit Force), which could display time. Sadly, people reported that the wrist band for the Force caused them to break out in terrible rashes. All that’s left of the Force today is artfully worded press releases that promise to get to the bottom of the matter.

One other Fitbit highlight worth noting: Lots of people I know have them compared to the other two trackers I’m covering in this entry. That makes checking leaderboards more fun.

Lesson: An open API goes a long way, even if the only thing you thought you wanted at first was a step and sleep tracker. So does a sense of reliability. I never really trusted my Up after the first two broke, and even though I loved the software and general UX, it really bothered me to think that my data would have gaps during periods where the replacement was in the mail.

Garmin Vívofit

Oh, man. The Vívofit. I’m wearing one right now and it’s a real love-hate thing.

Compared to the Up and Flex:

  • It takes a lozenge-in-a-rubber-pouch approach, like the Flex.
  • It displays the time.
  • It displays your actual steps and tells you how far you are from completing your daily goal.
  • It tracks sleep quality.
  • It uses button presses for input: Short press to switch between modes (time, date, steps, goal) and long press to activate sync or put it into sleep-tracking mode.
  • It syncs via Bluetooth or via a little dongle you plug into your computer.

One final, huge difference: It takes a standard watch battery that lasts for a year. No recharging every 5-7 days.

I really like the Vívofit as a step tracker and wrist watch. I kind of like it as a sleep tracker. Where it falls down is its API, which Garmin isn’t super interested in opening up. Garmin has partnered with MyFitnessPal, which acts as a sort of Switzerland to the many activity trackers out there by pulling in their calorie data to go with the food data you’re entering into it.

Other than that MyFitnessPal integration, the Vívofit doesn’t talk to the rest of the world much. There are services that try to sync it up with Runkeeper Pro and Strava, but they don’t work reliably. It seems to be willing to talk to my scale. In the end, if you want to get data from a bicycle ride or swim into Garmin’s Connect app, you have to enter it by hand.

The reasons for this are pretty clear: Garmin got into the activity tracker business to supplement its more lucrative GPS devices, which are under threat from GPS-enabled smartphones. Talking to a smartphone app like Runkeeper or Strava just means you aren’t going to be as interested in buying a Forerunner GPS tracker or Edge bicycle computer.

I completely get why Garmin’s got to do this, and if I were already heavily invested in their product line I’d probably be really happy with the Vívofit. As it is, I don’t need their other products, which makes their closed off Connect dashboard pretty irritating.

The things keeping me happy with the Vívofit are the watch function (I hate pulling out my phone during meetings to check the time … it feels rude) and the fact that the people sitting on either side of me at work are also using Vívofits, which means we can see each others’ step activity in our dashboards (though none of my biking activity — 75 miles a week at this point — can appear unless I key it in manually). It’s also nice to see my daily steps on the display.

Quantified Me

In the end, I pretty much live in MyFitnessPal to keep track of what my walking and biking activities mean in combination with my diet:

  • I use my Garmin for step tracking
  • I use Strava for ride tracking
  • I use a Withings scale to track my weight
  • I enter my food in MyFitnessPal’s diary

The Vívofit, Strava app and scale send data to MyFitnessPal, which reconciles their calorie information with its food journal. At the end of the day, I sync the Vívofit and close out my food diary, and MyFitnessPal tells me what my weight would be in five weeks if every day was like the one I just logged.

MyFitnessPal doesn’t care about sleep patterns, so if I want to see that data I need to use the Garmin app. Since I use a gradual wake alarm clock app (highly recommended), I can get the data from there, too.


  1. Clanton McClandish says:

    July 12th, 2014 at 12:17 pm (#)

    Interesting write up Mike. I’ve toyed with the idea of a monitor but it seems the market is just now out of its infancy. I understand Apple is going to make a play in this area sometime soon, so I’ll likely wait until it has the bugs shaken out of it.

  2. m. hall says:

    July 12th, 2014 at 12:29 pm (#)

    Yeah, I’m wondering about the iThing, too. I understand it’s going to track quite a bit of stuff, and I’d guess it makes the Pebble (which I have one of but have given up on) look pretty weak.

    The thing I’ll be curious about there is how open Apple makes it. If it’s able to pull in my Strava, Runkeeper and scale data plus tracking steps and sleep, I’m pretty sure I’ll just continue down the path through Mr. Jobs’ walled garden.

    If they close it off, I’ll be less interested.

    iOS 8 changes suggest Apple’s more interested in opening stuff up, so I’m feeling hopeful.

Leave a Response

© Michael Hall, licensed under a Creative Commons Attribution-ShareAlike 3.0 United States license.