I found sources that describe the time as:
"for as long as necessary."
Was that minutes, hours, or did it change dependent on light conditions?
It depends both on light conditions and the specifics of the daguerreotype process used, which advanced over time. At the public announcement of the invention in 1839, even in outdoor conditions of bright sunlight a daguerreotype would take from 15-30 minutes to properly expose. The few portraits made at this time were taken with the subjects' eyes shut! However modifications to the chemistry and lenses came quickly. By early 1840, exposure times could be down to 1-3 minutes in bright light, either outdoors or in a studio with abundant windows, skylights, and white walls. Further chemical innovation brought the average (portrait in bright light, the studio standard) exposure time down to 3-5 seconds by the end of 1840, where it stayed for the rest of its time. As the majority of daguerreotypes made were studio portraits made 1840-1860, it's reasonable to say the average time for a daguerreotype exposure was 3-5 seconds.
If one wanted to make an exposure on a heavily clouded day, or at dusk, or in other conditions than bright light, it was certainly possible, but exposure times would vary (and usually be too long for portraiture). The ability to estimate exposure times based on conditions would largely depend on the experience of the photographer. Luckily, though the preparation of a daguerreotype plate is labor-intensive, it is quick to develop, and before it is fixed, the image can simply be rubbed off if it is a bad exposure, and the plate re-user.
Sources:
Bertrand Lavendrine, Photographs of the Past. Process and Preservation. Los Angeles: 2009: The Getty Conservation Institute.
"Manner of Taking Portraits by the Daguerreotype" The Boston Daily Advertiser and Patriot, March 26, 1840.