Each card has some text printed on the top row. The bulk of the card is a grid of pre-printed numbers. Punching out holes on the card lets you represent data in a standardised character format, where successive columns represent individual characters (from top to bottom). For example, in the IBM 80-column format [0]. The text printed on the top of a card makes individual cards human-readable.
When I first encountered computers, in approx 1979, this was how we programmed them. Now get off my lawn, with your new-fangled glass-teletypes.
[0] https://en.wikipedia.org/wiki/Punched_card#IBM_80-column_for...
Like the telephone with a dial disk or the monitor with a CRT screen, this is an unfamiliar object for not too old computer programmers.
Nevertheless, there was a time when writing any computer program required it to be punched on such cards, one 80-character line per card, then you could read your program like with this Jane Austen example and you had to be exceedingly careful to not spill down the cards on the floor, or you would have to do a lot of work to sort them again in the correct order of the program lines.
Actually, there were machines called card sorters that could sort cards back into the right order.[1] In the card images shown on this page, you can see that there are sequential serial numbers in columns 73 - 80, which could be used to sort these cards.
Many programming languages, such as Fortran, would ignore anything punched after column 72, allowing columns 73 - 80 to be used for this purpose.