John’s Blog

As I wrote up yesterday, my new Jawbone Up bracelet died on me. I was bummed, but customer service did a great job, and got me a replacement in about a day.

But I was surprised about the feeling I had yesterday, which is the first day I hadn’t worn the Up in about 3 weeks: I felt like I was wasting data, since I didn’t have the bracelet on to collect it. Let me say that again: I felt kind of weird, because I was going to all this trouble (you know, by walking around, sleeping, etc) and none of the data I was generating was getting logged. I was just blasting it into the ether.

This is not totally how I expected to feel. So naturally I tweeted about it, and Chris Hogg, founder of 100plus, replied in a surprising way:

And when you think about it, of course this is correct, but my first reaction was: “This is kind of messed up — we need to get credit for living now?” Obviously not a perfect state of being for humans.

But I thought about it some more on my drive in this morning, and it seems to me that a couple of different things are happening. On “getting credit,” we’re talking about taking an intrinsic motivation (living well) and replacing it with an extrinsic motivation (someone giving you credit or status). Now, that happens all the time — it’s a lot of what happens in schools, weight loss programs, sports, etc. But I think it’s not the most durable kind of behavior change mechanism.

And, honestly, I’m pretty sure that wasn’t what was happening with me. Because while I like to think that everyone cares about everything about me, it’s really tough to imagine that anyone reading this (except for maybe my mom – hi mom!) gives a damn about how much I walked yesterday. And it’s even tougher to imagine that I would care much about what you thought about how much I walked.

So for me in this particular case, I think it was my OCD nature kicking in. I didn’t really like imagining the graph (of whatever data it is that I may, some day in the future, create a graph from) having a void spot in it. I want things to be smooth lines, no drop outs. I think that can be a motivator for folks, too — completeness of the data report, but I think it’ll motivate a far smaller set of folks.

I think where you get to in the end is that there is something very important about tracking data, because it changes your relationship to the activities you’re measuring pretty fundamentally. (some ways good, others not good.)

And that collecting data about how you live, in particular, will always have holes in it — for lots of reasons.

Anyhow, was an interesting set of thoughts — some important things going on here.

     

I’ve been wearing my Up bracelet for about 3 weeks now and wanted to write a few things down about the experience. This particular writeup is a little jumbled up — it’s a mix of reflections on the Up specifically with some of the thinking I’m doing about health systems in general. Will follow up with another few posts as I’m able to articulate more and more clearly.

[update: as luck would have it, as I was writing this yesterday, my Up ran out of juice — after a night of recharging, it’s unresponsive, so after a 20 minute call with Jawbone customer service, am getting it replaced with a new one. Making great hardware & software is hard.]

Overall summary: I like the hardware a lot, and think the main breakthrough is that it’s a fashion accessory that you’ll generally want to wear all the time. I think there are some problems with the hardware that will mostly get solved over time. I think the software side of things is not very good, though — very constrained, pretty quirky in operation, very incomplete. Beyond that, I think that it’s not giving me a ton of information that I can actually do anything with, and that’s a huge problem.

But now let me get into some of the details a bit. First, my basic orientation towards health sensors is hugely positive. I’m pretty convinced that, more and more, we will all have sensors that are with us all the time, collecting the information that we generate just by living. Where we are, our activity level, what we eat, our body temperature, how we sleep, etc. They’re pretty nerdy right now, but they’re coming, and coming quickly. That’s the first step. Once that happens, you need to be able to mash all the data from various sensors (and larger data sets such as weather, etc) to get a coherent picture of your own life, and beyond that, we’ll start to be able to build population sensor information as we start to aggregate data from many people. And that will lead to being able to actually derive insight about ourselves and the people around us.

I’ve got another blog post in mind to talk about all that stuff in context, but for now, we’re in the first stages of the story, which is really mostly about collecting the data.

At one time or another, I’ve tried out a lot of different sensors: the Zeo, the FitBit, the SportBrain (a while back), Nike+, and now the Up. I think the Up will be the stickiest for me of the bunch, but it’s got issues.

The physical bracelet is terrific, I think. Comfortable to wear, not too heavy, doesn’t really bother me when I’m typing or working out. I generally wear it all the time — it’s waterproof, which helps a bunch. The only time I really take it off is to charge it for an hour or two every week or so. (It actually feels a little weird to take it off to charge, since it means that there’s a period that I’m not collecting data — a small thing, but  interesting emotional reaction.) It’s not wireless, which is disappointing — you’ve got to plug it into your iPhone’s headphone in order to pull the data off. I assume that’ll get addressed in a future version by including low power Bluetooth or something like that.

I think that the physical form factor is the most significant breakthrough here. In terms of sensor functionality, I think it’s essentially the same as the Fitbit — a pedometer, sleep monitor, vibrating alarm clock — I can’t think of anything the Up can do that the Fitbit can’t. But the main difference is that I actually have the Up with me all the time. I like wearing it — it helps me to be somewhat more mindful of what I eat and how I act. With my Fitbit, I always forgot to keep it charged, forgot to wear it, had it knocked off my clothes or broken.

So that’s a significant breakthrough — you can’t derive insights from data that you don’t collect. With Up, I’m collecting more consistently, and that’s huge. It’s because it’s fashion (in a nerdy sense, but I think that will change over time) and because it’s well-built.

The software, though, is a pretty big challenge at this point — it’s just not very good. There are lots of UI issues — 3 clicks to do a sync, hidden functionality, lots of bugs in modes — but that stuff will get fixed over time, so it’s not a huge concern. My bigger problem with the software, which is essentially a combination of pedometer, sleep tracker, and foodspotting app, is threefold. The first problem is that the social stuff is all very thin — both hidden in the UI, but also just not very deep in terms of changing the way you interact with others and how you can help each other be healthier. The second problem is that it’s very limited functionality and is a closed system — I’ll expand on that in a minute.

But the biggest problem I have with it is just that it doesn’t help me figure out what to do at all. If you look at the screen shots at the top of this post, the one on the left tells me I walked a lot that day (was the Oregon @ Stanford football game day). The one on the right tells me some surface level statistics about how I slept. So what? Is that good? Bad? How’s it relate to what I usually do? Should I be doing something different? Am I doing it wrong? My essential issue with the Up software right now is that it’s just data porn. Lots of data points, but no behavior change enablement.

The lack of a Web UI is also a head scratcher. My developing view is that we use different size screens for different contexts and types of tasks. I use my phone when I’m on the go, but also because it’s more personal/intimate in terms of collecting data and checking out status of friends (and myself in the case of sensors). But when I’m trying to do more analysis/thinking about the big picture, I really want to do it on a screen like a tablet or my laptop. Going through the UI day by day, not being able to correlate anything with anything else, without transcribing it all in my notebook — just seems ridiculous to me, and essentially unhelpful.

Which brings me back to the open/closed problem. My basic belief is that building great hardware is really hard. And building great software is really hard. And building great social systems is really hard. If you look hard, you can find many companies that have been great at one of those things. And you can find a few that have been great at 2 of them. But I think it’s nearly impossible to find organizations that have been able to build great hardware + software + social systems — building the systems is just very very tricky, and requires a broad set of knowledge, and a surgical set of decisions about how to make tradeoffs between vastly different sets of concerns.

So I conclude from that starting point that it’s going to be tough — very tough — for any one company to build a great experience overall — and it seems to me obvious that there are zero examples of superior experiences overall in the market today.

But in the emerging space of sensors & monitoring, there’s another problem: there are lots of different sensors around, the number is growing, and to really figure out what’s happening, you need to correlate among several. Nobody makes the sensor to rule them all yet. You need your weight (withings scale), diet (lots of apps), sleep (up, fitbit, zeo), etc etc.

So you’re left with the conclusion that what you really need is an open system. Or, rather, several open systems — you really want great devices from a lot of different folks feeding into some data collection system that allows you to do some data analysis, all fronted by user interfaces that help you actually understand things and (hopefully) change the way you live your life.

Right now, I think that (save for the hardware glitches) Jawbone makes probably the most useful sensor around, and certainly the one that’s most attractive that you’re likely to wear all the time. The software on the iPhone is limited and frustrating at present, for me and most folks I’ve talked with, and it’s missing a major component by not being available for larger screens. I like my Up, and plan to wear it, but am going to have to figure out how to get it to play better within the overall personal health/sensor ecosystem that I’m building up over time.