Type 1 diabetic here. Maybe I can offer some unique insight. Or maybe, also being a hacker, that makes me overly critical, so sorry if this is too harsh.

There are tons of tools like this out there (including software from Medtronic, the leading maker of insulin pumps, that they've clinically proven increases diabetes control, and integrate directly with glucose meters). I don't use them because:

* My meter stores that stuff for me and does charts and stats. If I want those charts on my computer, I can just upload the data with the software that comes with the meter.

* Having all the information graphed in one place on the web isn't more useful than looking at my meter.

* If I did want more than the meter does in one place, I would just use a spreadsheet.

* These tools don't add any major additional insight into what I can do to manage my diabetes better, at least compared to the work in using them.

Your pitch -- and the center of your development thought process -- needs to be "use this and you will live longer and have less stress in your life, for X, Y and Z reasons". Don't write another line of code until you have that, is my advice.

The most general problem is that there is no "stickiness" to the app -- no compelling reason to keep using it day after day. (Some ideas to get you going: make it get smarter the more I use it. Send me actionable info -- "you've only gone to the gym once this week -- on weeks when you go at least 2x, your avg. BG is 20 points lower". Make me want to show off when I'm doing well, and indirectly apply social pressure to do better when I'm not.)

And you don't even have some basic details right. You have nothing for exercise! That's huge! And the glycemic load, fat content, and amount of fiber in a food is as important as how much carbohydrate is in it. Knowing how many carbs are in a piece of pizza isn't that useful (and there are a million other tools to do that already --including ones built directly into insulin pumps and blood glucose meters, and mobile apps). If you don't know that you should take very different insulin amounts and patterns when eating a serving of Skittles, pizza or brown rice even though they have the same amount of CHO, you don't know enough to help.

There really isn't anything in the app that shows you've even read the wikipedia page about diabetes. Sorry if that sounds harsh but every part of the app reflects it, and so as a diabetic it's hard not to find the whole thing condescending. In addition to nothing on exercise, your app doesn't do anything with A1C/fructosamine, C-peptide tests, cholesterol, T4, etc., all of which are more useful to compare to blood glucose readings than BP, and as important to track over time. Have you even heard of A1C? They run commercials on TV reminding diabetics to get it checked regularly all the time.

Maybe my expectations are too high, or maybe yours are. If you're going for super-simple, it needs to be a lot more simple, a bit more useful, and a lot more memorable. If you're going for sophisticated, you've got a whole lot of work to do and probably need to get diabetes and/or a medical degree first. But you need to make a choice: are you going for power users, or casual users? Even if you got the medical details right, there's a basic market fit problem here, and a lack of stickyness, as mentioned above. There are way too many competitors not to have the market fit be solid and have some very unique twist.

In general, if you're selling a specialty product, you need to know more about the specialty than the average user. I've created some very successful software for the specialty $foo market. But I was a $foo-ographer for 10 years, everybody in the $foo world knows my name, I've been blogging about $foo for 5 years, etc.

As a white guy, would you start a line of hair-care products for black people? Open up a Thai restaurant even though you've never been to Thailand, don't even eat Thai food, and there are already 3 good Thai restaurants in the neighborhood -- but you heard that lemongrass is somehow involved? Probably not, so why do this? DON'T FAKE THE FUNK.

Oh, final pet peeve. For the love of Pete don't say stuff like "CareLogger makes it possible to share your logged information with your physician simply by printing off the desired records." No. Being able to print a web page is not a feature. I can just take my glucose meter to the doctor and they can pull my data directly from it into their computer! Don't mention stuff that makes you look bad, even to downplay it. If you can't compete on a feature, don't mention it. Sell the features you do have, preferably the ones competitors don't.

user:	liminalist
created:	February 14, 2009
karma:	12