# ⚡Data Cloud Deep Dive — Data Types in Data Cloud

## Table of Contents

1.  Automatic data-type detection
    *   1a. Supported data types
2.  Why data types matter
3.  Date-type mismatch example
4.  The one mapping exception
5.  DLO data types are difficult to change
6.  Recommended planning process
7.  Key terms
8.  Key takeaway

## 1. Automatic data-type detection

When a data stream ingests data from an external system, Data Cloud automatically detects the data type of each field.

**Examples:**

*   Name → Text
*   Age → Number
*   Date of birth → Date

Automatic detection is convenient, but it's not the end of the process — **automatic detection must be reviewed and confirmed.** Data Cloud is making a best guess, not issuing a guarantee.

## 1a. Supported data types

When Data Cloud (Data 360) detects or assigns a data type for a field, it uses one of these ten:

*   Boolean
*   Date
*   Datetime
*   Text
*   Number
*   Currency
*   Email
*   Percent
*   Phone
*   URL

Note that Date and Datetime are separate types. This is why a timestamp from S3 can't be mapped to a Date field in a DMO without standardizing it first.

## 2. Why data types matter

Data types become critical during **harmonization**, when Data Lake Objects (DLOs) are mapped to Data Model Objects (DMOs).

For a mapping to succeed:

*   DLO and DMO fields should have matching data types.
*   The fields should also have compatible business meanings.
*   Incorrect types can cause mapping or harmonization failures.

Matching data types alone isn't sufficient either — a field can technically be the right type and still mean something different. Both conditions need to hold.

## 3. Date-type mismatch example

The same underlying data can arrive in genuinely different formats depending on the source:

| Source | Example |
| --- | --- |
| CRM | Date of birth in DDMMYY format |
| S3 | Date of birth as a date-time or timestamp |

Before mapping these fields to a DMO, determine the target data type and standardize the source data accordingly.

Here's the trap worth calling out explicitly: **a date format such as DDMMYY is not automatically the same as a Date data type.** Unless it's correctly interpreted or transformed, it may simply be stored as text — which looks fine at a glance and breaks quietly the moment something downstream expects an actual date.

## 4. The one mapping exception

Data Cloud allows one specific type conversion during mapping: **Number → Text**.

Other data-type mappings generally require an exact match. If the DLO field type and the target DMO field type don't already match, the mapping isn't going to work by default — you need to standardize the source data first, not rely on Data Cloud to quietly convert it for you.

## 5. DLO data types are difficult to change

This is the section that turns a data-type mistake into a real setback. **After a DLO is created, its field data types cannot generally be changed directly.**

If the wrong type gets selected, fixing it may require:

1.  Recreating the DLO.
2.  Reconfiguring the data stream.
3.  Redoing field mappings.
4.  Repeating harmonization and testing.

That's not a five-minute fix. That's redoing a meaningful chunk of the ingestion pipeline — which is exactly why the planning step below matters more than it might seem to at first glance.

## 6. Recommended planning process

Before configuring a data stream:

*   Inventory all source fields.
*   Document each field's business meaning.
*   Record sample values and source formats.
*   Confirm the automatically detected type.
*   Identify date, date-time, number, and text inconsistencies.
*   Compare DLO fields with the target DMO fields.
*   Define required transformations.
*   Validate the design.
*   Confirm the design with the data architect or product owner.

None of these steps are individually hard. The value is in doing them *before* the DLO exists, not after — because that's the point where a mistake is cheap to fix instead of expensive.

## 7. Key terms

| Concept | Meaning |
| --- | --- |
| DLO | Stores ingested source data |
| DMO | Stores harmonized business data |
| Harmonization | Mapping and standardizing DLO data to DMOs |
| Data type | Defines how a field's value is stored and interpreted |

## 8. Key takeaway

Data Cloud detects data types automatically, but the implementation team is responsible for validating them. **Plan and confirm data types before configuration**, because correcting an incorrect DLO data type can require significant rework.

## Quick knowledge check

1.  What happens automatically when a data stream ingests data?
2.  Why must automatic type detection still be reviewed?
3.  Why do data types matter during harmonization?
4.  What's the one data-type conversion Data Cloud allows during mapping?
5.  Why isn't a DDMMYY date field automatically treated as a Date type?
6.  What has to happen if a DLO's data type turns out to be wrong?
7.  What should be done before configuring a data stream in the first place?
8.  Name the data types Data Cloud supports, and which two are easy to confuse?