Fitbit’s first two days of trading were good ones: on Thursday, the activity-tracker company’s shares soared $9.68, or 48 percent, to $29.68, and on Friday, they climbed another $2.82, or 9.5 percent, to end the week at $32.50.

Yet while investors are evidently enthusiastic about Fitbit, which focuses on selling a range of wristbands and clip-on gadgets for monitoring metrics like steps, sleep, calories, and heart rate, there are still major challenges to making these kinds of wearable devices work for all kinds of people and using them to glean medically useful data (see “The Struggle for Accurate Measurements on Your Wrist”).

For instance, the wrist–which most popular activity trackers, including most of Fitbit’s products, are made to clasp–isn’t always a great spot for taking measurements. When it comes to using optical heart rate sensors, which Fitbit includes in its higher-end wristbands, arms that are too hairy, fat, skinny, or sweaty, can result in inaccurate measurements.

And while activity tracking has, for the most part, already been figured out across the wearable market, this potential for innacuracy makes it hard for devices to move from gym companions to medical assistants that actually help monitor your health. For that to happen, we’ll need to make devices that can take more accurate measurements on a wider range of people, and we’ll also probably need devices that can take even more types of measurements. These measurements could include blood pressure, for instance, which a company called Quanttus is working on by using a wristband to track the tiny body movements that result from your heart pumping blood, and skin conductance, which is already starting to show up in some wristbands such as Microsoft’s Band and measures the skin’s ability to conduct electricity (this tends to climb along with stress).