Here’s what I’ve done…
I started with the created timestamp…
DATETIME_FORMAT(CREATED_TIME(),'X')
This produces a string like this:
1535390717
People may guess that it’s a timestamp and so it’s sequential and be able to guess the ID of other records, so we add a random number to it to make it look less like a timestamp:
DATETIME_FORMAT(CREATED_TIME(),'X')+3647225632
It’s important that the random number be at least as many digits as the timestamp (10 digits). This yields something like this:
5182611245
Ok great, as long as we have this, each of our IDs will always be unique.
The problem now is that they’re still sequential and so people can guess other record IDs easily enough. So lets create a semi-random string of digits.
RIGHT(DATETIME_FORMAT(CREATED_TIME(),'X'), 4) + LEN({Column name}) + 4321
Here I’m getting the last 4 digits of the timestamp adding the length of the contents of another column and the number 4321. Assuming the contents of the column you choose can vary, this should be a pretty different number for each record.
Now we put the two numbers together separated by a hyphen:
CONCATENATE(
RIGHT(DATETIME_FORMAT(CREATED_TIME(),'X'), 4) + LEN({Column name}) + 4321,
"-",
DATETIME_FORMAT(CREATED_TIME(),'X')+3647225632
)
You should end up with something like this:
9963-5182611245
Now we can dress this up some more if you like, sprinkling in some letters for example:
CONCATENATE(
"rb",
RIGHT(DATETIME_FORMAT(CREATED_TIME(),'X'), 4) + LEN({Column name}) + 4321,
"qka",
DATETIME_FORMAT(CREATED_TIME(),'X')+3647225632
)
Then you end up with IDs that look like this:
rb9963qka5182611245
rb12828qka5182614001
rb5777qka5182616349
