One kilobyte is 1,000 bytes decimally, so a byte is 0.001 kB. This is the smallest step in the storage ladder and the one where the binary alternative of 1,024 originally took hold.
About the units
The byte is the fundamental unit: eight bits, one addressable memory location, one character in most legacy text encodings. The kilobyte was the first prefix applied to it, and the decision to treat 1,024 as "kilo" was made informally in the 1960s because 2¹⁰ happens to sit within 2.4 per cent of a thousand and fell naturally out of binary addressing. That approximation was harmless at the time and became progressively less so with each additional prefix. The IEC's kibibyte, KiB, names the binary value properly, and the difference at this step is 24 bytes.
The exact factor
The factor is 1,000 decimal, 1,024 binary. In practice at this scale the choice rarely changes a decision, since a file of 2,048 bytes is either 2.048 kB or 2 KiB and no workflow depends on which. What does matter here is the relationship to characters: in ASCII or Latin-1, one byte is one character, so a kilobyte is a thousand characters — roughly 150 words of English. In UTF-8, ASCII characters still take one byte but accented letters take two and most CJK characters take three, so the same kilobyte holds considerably less text in some languages than others.
Where you meet this conversion
File system metadata, network packet sizes, and anything where individual bytes are counted. A standard Ethernet frame carries up to 1,500 bytes of payload, which is 1.5 kB, and that number shapes how larger transfers are chopped up. Cryptographic keys, certificates and hashes are all sized in bytes. Embedded systems work at this scale throughout: a microcontroller may have 2 kB of RAM, and every byte is accounted for.