Three Months with Fitbit Flex

I am a former owner of a Fitbit One. I thought that the One was an excellent product. It accurately tracked steps, and did fun stuff like measuring floors climbed (elevation change is a cool trick when you run and bike).

My issue with the One was the form factor: the clip-on style works well if you’re male and have a pocket, but many women’s outfits don’t have pockets. Clipping it to the waistband of my leggings or skirt led to one too many incidents where I watched the One sail across the bathroom in a graceful arc. Not what I wanted from my device.

The Flex is a wristband-based system, which is a vast improvement in form factor. And that’s pretty much where the “pros” list ends.

I spent three months living with the Flex. One day, when its battery ran out, I decided not to bother recharging. Why? Because it’s accuracy was questionable, at best, and I didn’t really need to be wearing a device that wasn’t doing what I needed it to do. I felt like I’d spent three months in the Flex beta program while they worked out the kinks with a half-baked device.

I tweeted @FitbitSupport with my frustration, and they suggested that I email customer service to explain my problems in detail. Cool. As I said, I liked the One, so I know that they’re capable of making excellent products, and I wanted to offer my suggestions for how to make it better.

Here are a list of my issues and the responses from Fitbit Customer Service:

Issue 1: Step miscounting. The round-trip walk to my son’s school tracked accurately with the One, but with the Flex, the result was about 24% less. A MapMy Run-tracked route in Germany was about 6.5 miles, yet the Flex gave me credit for roughly half that. One evening, while the device was plugged into my computer for recharging, it “tracked” hundreds of steps, yet tracked precisely zero for the first few hours that I wore it that morning.

Response: “We think that the issue may be fixed by restarting your tracker. To restart your tracker, please do the following…. [Note: I had already restarted the tracker twice in an attempt to fix this on my own. No luck.]

For most people, there may be no difference at all between clip and wrist based trackers or it may be within a few percentage points difference. That said, if you have a lifestyle where you move your hands a lot such as playing the drums everyday, you may see a few extra steps on your Flex as we do want to give you credit for this activity.”

Issue 1: My Follow-up. I explained that no, there was a massive difference between One and Flex counting, and the difference between mileage-tracking apps and Flex tracking.

Follow-up Response: [Note: in essence, “You’re using it wrong.”] “We want to let you know that Flex has been tested extensively against our clip-based devices like the Fitbit One and Fitbit Zip. However, in order to compare your steps to other trackers, the other trackers need to be linked to accounts as well under the same settings, same weight, age, height and gender or stride length.

We recommend you to measure your stride length for a more accurate reading…. 

This is no different than any wrist-based tracker on the market. For most people, there may be no difference at all between clip and wrist based trackers or it may be within a few percentage points difference. Please let us know if changing these settings doesn’t improve the accuracy of your Flex, and we’ll be happy to continue investigating the issue.” [I did let them know that my step length is calibrated correctly, and in fact is calibrated the same as it was with the One, which was more accurate.]

Issue 2: Sleep tracking. To track your sleep, you need to tap-tap-tap to activate the sleep sensor. I can tap that thing all day long and it won’t enter sleep mode. It WILL sleep when I ride my bike over a bumpy patch of road, or bang a spoon on the edge of a pan, or clap my hands at my son’s baseball game, and remain asleep until I notice.

Response: “The Flex can be put into sleep when riding your bike over a bumpy road or when the sensor inside confuses the tapping with the movements you perform when doing an activity. However, the tracker will automatically deactivate the sleep mode after it senses too much activity that can’t be related to a person sleeping. You can also double tap the tracker to place it out of sleep. [Note: it does not. Even as recently as a few days ago, I discovered that my device had been “sleeping” for hours, and I was unable to wake it with the tap-tap method.] We appreciate your understanding with this matter. We’ve passed your request on to our Flex team for consideration and hope that we can improve this in the near future.”

Issue 3: Very active minutes. Fitbit credits you with very active minutes for high intensity activities. They weight step-related activities more heavily than manually logged activities, like cycling or strength training. Yet there seems to be no rhyme or reason to how many “active minutes” you earn. An hour-long bike ride earned 4 minutes, while a moderately paced walk to school earned 30.

Response: “The Very Active minutes can be affected by the activities you log in. You earn “Very Active Minutes” when you wear your Fitbit while doing cardio workouts and high intensity activities like jogging, running, aerobics, biking, rowing, or anything where you are working up a sweat. [Note: I don’t work up a sweat sauntering to school, but I do when out for a long bike ride.]

By default, you are given a starting goal of 30 Very Active Minutes a day based on the Center for Disease Control’s recommendation of 20-30 minutes of daily moderate-to-intense activity. You can always change your Very Active Minutes goal to meet your personal needs.”

Issue 3: My Follow-up. Yes, I get the CDC recommendations, I’m just confused by how it allocates these minutes. Why would 60 minutes of activity earn 4 Very Active Minutes?

Follow-up Response: “In regards of the Very Active minutes situation, we recommend you to log any activities and the duration regarding any active minutes on your Dashboard. We were able to access your account and no Activities were logged for the past few days. [Note: this is because, as I told them, I had stopped using their device.] To log your activities manually please do the following….”

At the end of this exchange, I was still convinced that the Flex device was inferior to the One, but I decided that I’d do one last reset and firmware update. Since then, I would add the following updates:

• It appears to be counting more accurately on long walks. However, on shorter activities, the step count is still inaccurate. For example, I can easily count the 25 steps from my kitchen to my grill, and 25 steps back. That should log 50 steps. A few nights ago, those 50 back-and forth steps logged as 17, 21, 18 and 24 steps on my four passes from kitchen to grill. Yes, I was syncing and tracking between each pass, just out of curiosity.
• It still puts itself to sleep. My afternoon bike ride set it snoozing and I was unable to wake it until the following morning when I banged the oatmeal spoon against the pot. As I type this, I realize that it’s in sleep mode once again.
• Very Active Minutes are still issued on a seemingly random basis. I biked 6 miles from school to the supermarket, 5.5 miles home, and got credit for approximately 40% of my bike time. I don’t understand how it allocates portions of the activity as being more active than others.

Overall Assessment

Buy a Fitbit One. If you’re planning to buy a Fitbit Flex… well, just don’t. Not until they release the next version, anyway.

Disclaimer: I paid for all Fitbit products with my own hard-earned money, and they did not ask me for a review of any kind. Since writing this draft, I went back to using the Flex with its new firmware update, but saw no measurable improvement. I no longer use the Flex because I lost the charger on vacation, and don’t think it’s worth sinking another $20 into a replacement charger.

Tags: accuracy, customer service, data, fitbit, fitbit flex, fitbit one, Fitness, undercount

« Previous Entry Next Entry »