A table is a labeled drawer, and each record inside it is a folder. A column is one of the labeled blanks on an index card filed inside that folder — "Name," "Phone Number," "Status," and so on. In Dataverse, columns are also called "fields" — you'll hear both terms used interchangeably.
- A column is one labeled blank on a record's index card
- "Column" and "field" mean the same thing
- Every column has a data type — the kind of answer it will accept
- Common types: Text, Number, Choice, Date, Lookup, Yes/No
- Changing a data type after real data exists can be difficult or destructive
Every blank expects a certain kind of answer
An index card's blanks aren't all the same kind. "Phone Number" expects digits, "Date of Birth" expects a date, "Are you a member?" expects yes or no. Write a paragraph into a date blank and the card stops making sense. Dataverse columns work the same way — each has a data type defining what kind of answer belongs.
| Data type | What it stores | When you'd use it |
|---|---|---|
| Text | Free-form words and characters | A name, an email address, a short description |
| Number | A whole number or decimal value | A quantity, an age, a score |
| Choice (option set) | One value picked from a fixed list | A status like "Active / Inactive / On Hold" |
| Date | A calendar date, optionally with a time | A booking date, a follow-up reminder |
| Lookup | A reference/link to a record in another table | Pointing a Contact to the Account it belongs to |
| Yes/No | A simple two-value switch | "Is this customer active?" |
Why the data type you pick up front matters
A data type isn't just cosmetic. It controls how a value is stored, validated, searched, and reported on. Number columns can be totaled and averaged; Text columns can't. Choice columns keep answers consistent — no "Active," "active," and "ACTIVE" typed three different ways. Free-text columns don't offer that safety net.
Once a column has real data in it, changing its data type isn't always simple — sometimes it isn't possible without side effects. Converting a Text column full of loosely typed numbers into a real Number column can fail on rows that don't cleanly convert, or lose formatting. Some changes are supported with caveats; others mean deleting the column and rebuilding it, losing whatever it held.
A useful habit: before adding a column, ask "what's the smallest, most specific data type that fits every valid answer?" A status is almost always a Choice, not free Text. A "how many" is almost always a Number, not a sentence.